ABSTRACT Title of Document: COMMUNICANT ACTIVENESS, COGNITIVE ENTREPRENEURSHIP, AND A SITUATIONAL THEORY OF PROBLEM SOLVING Jeong-Nam Kim, Doctor of Philosophy, 2006 Directed By: Professor James E. Grunig Department of Communication This dissertation presents a situational theory of problem solving that highlights distinctive communicative and cognitive features in human problem solving. Its purpose is to provide a simple and useful, but not atheoretical, account of communication behavior and the cognitive approaches that we adopt during problematic situations. In the conceptualization, I introduce a new concept, communicant activeness in problem solving (CAPS), which has three domains in communicant activeness to explain not only when people voluntarily learn and share information but also how they choose certain information as more relevant than other information. The three domains are information selection (information forefending and information permitting), information transmission (information forwarding and information sharing), and information acquisition (information seeking and information processing). I then use the focal construct, communicant activeness in problem solving, as a dependent variable in the new situational theory of problem solving. I also propose another new concept, cognitive entrepreneurship in problem solving (CEPS). It describes cognitive strategies that we take to reason about a solution in some problematic situations. Depending on the situation, we adopt a more or less entrepreneurial mindset. This construct contains four distinct but correlated dimensions: cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspension. For conceptual convenience, I named the more entrepreneurial approach the cognitive alpha strategy and the less entrepreneurial approach the cognitive omega strategy. The construct of cognitive entrepreneurship becomes another dependent variable to be accounted for by the independent variables in the situational theory. To explain the cognitive and communicative dependent variables in problem solving, I use four situational antecedent conditions from the situational theory of publics: problem recognition, constraint recognition, level of involvement, and referent criterion (J. Grunig, 1968, 1997). I refine these antecedent concepts to accommodate several conceptual issues found from the past research of the situational theory of publics (e.g., the multicollinearity issue among independent variables). I also introduce the concept of situational motivation in problem solving that explains motivational effects on subsequent cognitive approaches and communicative behaviors. These revised situational antecedent variables jointly explain 1) how and why people communicate and 2) how people use unique cognitive strategies when they approach problem resolution. I called this emerging theory the situational theory of problem solving (STOPS). This dissertation elaborates 1) a conceptual model of communicant activeness in problem solving; 2) another conceptual model of cognitive entrepreneurship in problem solving; 3) a situational and motivational account for when, why, and how people communicate and are cognitively unique in a problematic situation. It then empirically tests a set of hypotheses and propositions that pertain to new concepts and the situational theory of problem solving. This dissertation advances conceptual understanding about how communication behavior and cognitive approaches affect our problem-solving efforts (descriptive theory building). It also contributes to finding a way to improve our adaptability in dealing with life problems (normative theory building). The new concepts and theory, CAPS, CEPS, and STOPS, offer some solutions for theoretical and practical problems in communication and several communication subfields. COMMUNICANT ACTIVENESS, COGNITIVE ENTREPRENEURSHIP, AND A SITUATIONAL THEORY OF PROBLEM SOLVING By Jeong-Nam Kim Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2006 Advisory Committee: Professor James E. Grunig, Chair Professor Linda Aldoory Professor Shannon A. Bowen Professor Larissa A. Grunig Professor Gregory R. Hancock ? Copyright by Jeong-Nam Kim 2006 Acknowledgements I thank God, who gave me an intriguing ?problem.? To me, it is still a mystery why God led me to Maryland and to this problem. I could have gone somewhere else and done something else. Kierkegaard said, ?Life is lived forward, but understood backward.? If he is right, I will understand the reasons later. I am curious. I would like to acknowledge the people who helped me in my years at Maryland. My dissertation committee members, Dr. James E. Grunig, Dr. Larissa A. Grunig, Dr. Linda Aldoory, Dr. Gregory R. Hancock, Dr. Shannon A. Bowen, and Dr. Elizabeth L. Toth were great supporters and mentors. Their insights, patience, and cheering helped me survive many challenges. Jim and Lauri have become my intellectual father and mother. I could be reborn by their teaching and caring. I am one of their brainchildren. For that reason, I have an obligation to continue their tradition of excellence. Outside of the committee, I would like to thank Dr. Arie W. Kruglanski, Dr. Woo Young Chun, Dr. Edward L. Fink, Dr. Deborah A. Cai, Dr, Leah Waks, and Dr. Michael Meffert. Through seminars, Arie and Woo Young trained me to be a better thinker. Ed not only built my methodological competence but also built my character as a researcher. And, Deb, Leah, and Michael shared their knowledge and wisdom with me whenever I asked. I would like to specially thank Dr. Randy Patnode, my colleague and mentor at Xavier University. He selflessly shared his time and knowledge to read and comment on my dissertation draft even during his busiest time. Without Randy?s help, I could not have finished this dissertation on time. ii In addition, other colleagues at Xavier, Dr. Indra de Silva, Dr. Tara Michels, Dr. Thomas Wagner, Mr. Tom Schick, Dr. Seung Bach, and Ms. Ryssa Kemper, helped me in many ways. I am also indebted to my friends in the University of Maryland. Seong-Hun Yun took his precious time to run my dissertation defense smoothly. Also, my brother Sungeun Chung, Miia Jaatinen, Brecken Chinn Swartz, Martie Downie, and Harrison De Stefano offered valuable input on my research. I also learned a lot from these friends and colleagues: Mark A. Van Dyke, Hyuk Yi, Lan Ni, Bey-Ling Sha, Yunna Rhee, Flora Hung, Sung-Un Yang, Hyo-Sook Kim, Min Jung Kang, Min Jung Sung, Nayoung Lee, Qi Wang, Bing Han, Yi Luo, Yi-Ru Chen, Natalie Tindall, Neil Mansharamani, Michael Phillips, Bj?rn and Belinda Stillion Southard, David A. Payne, Tiffany Derville, Ahnlee Jang, Shuo Yao, Younggu Kim, and Jai Hwa Choi. Their friendships made my graduate life more enjoyable and rich. Finally, I would like to share my joy with my family. My mother, Bok-Ja Hwang, and father, Seon-Il Kim, sacrificed themselves for my education. I will never be able to repay their support and love. My other parents, Mi-Ja Kang and Duk-Young Jung, were great supporters of my education. They listened to me and shared joys and sadness. I also thank my sisters and brothers, Hee-Jin, Hee-Kyung, Wook, Zyeon, Jaeho, Jaenho, Sang- Jun, and Hyun-Jin for their prayers and smiles. My beloved boys, James (Min) and Edward (Yeon) were great cheerleaders. They made me hopeful every day and night. And, my girl, Yu-Ra. The only choice I never regretted. Without her, I could not be here. She was the land, and I was a tree standing on her. iii Table of Contents ACKNOWLEDGEMENTS.......................................................................................... ii TABLE OF CONTENTS............................................................................................. iv LIST OF TABLES.....................................................................................................viii LIST OF FIGURES ...................................................................................................... x Chapter I: INTRODUCTION ....................................................................................... 1 Situational Theory of Publics.................................................................................... 4 Problems of Communication Theories ..................................................................... 7 Communicant Activeness and Cognitive Entrepreneurship in Communication Research ................................................................................................................................. 11 Methodology........................................................................................................... 15 Significance............................................................................................................. 16 Chapter II: CONCEPTUALIZATION ....................................................................... 20 A MODEL OF COMMUNICANT ACTIVENESS IN PROBLEM SOLVING........ 21 A Need for General Look for Communication Behavior ................................... 23 Situational Commuincant Activeness......................................................................... 25 Data, Information, and Knowledge..................................................................... 25 Conceptualizing Communicant Activeness ................................................................ 27 Information Acquisition: Information Processing and Information Seeking ............. 28 Information Processing and Information Seeking .............................................. 29 Information Transmission: Information Sharing and Information Forwarding.......... 31 Reasons for Information Transmitting................................................................ 33 Information Sharing and Information Forwarding ............................................. 35 Individual Problem Solver vs. Collective Problem Solver ................................. 36 Information Selection: Information Permitting and Information Forefending........... 37 Problems in Studying Selective Communication ............................................... 42 Emerging Communicant Selectiveness during Problem Solving ....................... 51 Information Permitting and Infomraiton Forefending ........................................ 58 Integration of Information Selection, Transmission, and Acquisition: A Catchall Conceptual Model of Communicant Activeness in Problem Solving ........................ 61 Two Phases of Situational Communicant Activeness: Inquiring and Effectuating Phases.................................................................................................................. 64 Selectivity in the Inquiring and Effectuating Phases .......................................... 68 Summary............................................................................................................. 70 iv A MODEL OF COGNITIVE ENTREPRENEURSHIP IN PROBLEM SOLVING . 72 Inferential Order in Problem Solving ......................................................................... 73 A Syllogistic Illustration of Cognitive Working................................................. 75 Cognitive Strategies and Behaviorla Molecules................................................. 84 Temporal Order between Will and Action.......................................................... 87 Delimitation ........................................................................................................ 90 Conceptualizing Cognitive Entrepreneurship in Problem Solving............................. 92 Conceptual Dimensions of Cognitive Entrepreneurship..................................... 94 Cognitive Retrogression.......................................................................................... 95 Cognitive Multilateralism ....................................................................................... 99 Cognitive Commitment......................................................................................... 102 Cognitive Suspension............................................................................................ 104 Cognitive Alpha and Omega Strategies................................................................ 107 Summary............................................................................................................... 112 ANTECEDENTS OF COMMUNICANT ACTIVENESS AND COGNITIVE ENTREPRENEURSHIP: AN INTRA-INDIVIDUAL ACCOUNT OF PROBLEM SOLVING................................................................................................................. 114 Human Perception..................................................................................................... 115 A Hierarchical Model of Problem Detection............................................................ 117 Problem Detection ............................................................................................ 118 Role of Emotion in Problem Detection............................................................. 123 Definition of Problem and Situation......................................................................... 132 Antecedent Parameters of Communicant Activeness and Cognitive Entrepreneurship in Problem Solving........................................................................................................ 140 Problem Recognition ........................................................................................ 140 Level of Involvement........................................................................................ 142 Constraint Recognition ..................................................................................... 144 Referent Criterion ............................................................................................. 145 Situational Motivation in Problem Solving ...................................................... 148 Summary........................................................................................................... 152 Integration of Dependent and Independent Variables .............................................. 154 The Situational Theory of Problem Solving ..................................................... 154 Situational Theory of Communicant Activeness in Problem Sovling and Situational Theory of Cognitive Entrepreneurship in Problem Solving ............................. 157 Chapter III: METHODOLOGY................................................................................ 165 Survey Research........................................................................................................ 166 Strength and Weakness of Survey Research..................................................... 166 Rationale for Choosing Survey Research ......................................................... 167 Research Design........................................................................................................ 169 Design of Study................................................................................................. 169 Scaling Procedure ............................................................................................. 170 Sampling Strategy............................................................................................. 174 Sample Size Estimation .................................................................................... 175 Instruments of Communicant Activeness in Problem Solving................................. 177 Instruments of Cognitive Entrepreneurship in Problem Solving.............................. 180 Instruments for Situational Antecedent Variables .................................................... 181 v Data Collection Procedure ........................................................................................ 184 Data Reduction and Data Analysis ........................................................................... 186 Model Testing ................................................................................................... 186 Model Identification.......................................................................................... 189 Ethical Consideration........................................................................................ 190 Chapter IV: RESULTS ............................................................................................. 192 Description of Survey Participants ........................................................................... 192 Preliminary Analysis................................................................................................. 192 Missing Value Analysis and Normality............................................................ 192 Reliability Tests and Exploratory Factor Analysis of the Key Constructs ............... 193 Communicant Activeness in Problem Solving ................................................. 194 Cognitive Entrepreneurship in Problem Solving .............................................. 204 Antecedent Variables of the CAPS and CEPS ................................................. 208 Structural Equation Modeling Analysis.................................................................... 217 Communicant Activeness in Problem Solving ................................................. 217 ................................................................ 217Confirmatory Factor Analysis ................................................................................. 225Hypothesis Testing Cognitive Entrepreneurship in Problem Solving .............................................. 230 ................................................................ 230Confirmatory Factor Analysis ................................................................................. 231Hypothesis Testing Structural Analysis............................................................................................ 240 .......................................................................................... 240SITCAPS Model Hypothesis Testing................................................................................ 241 .......................................................................................... 248SITCEPS Model Hypothesis Testing................................................................................ 248 ....................................................... 254Relationship between CAPS and CEPS Hypothesis Testing................................................................................ 255 ..................................................................................... 261Additional Analysis CHAPTER V: CONCLUSIONS AND IMPLICATIONS ....................................... 269 Summary and Discussion.......................................................................................... 270 Reliability and Validity of the Measurement Instruments................................ 270 Tests of the New Constructs and Hypotheses................................................... 271 ............................................................................................................ 271CAPS ............................................................................................................ 273CEPS ...................................................................................................... 277SITCAPS ...................................................................................................... 281SITCEPS ....................................................... 283Relationship between CAPS and CEPS ..................................................................................... 285Additional Analysis ....................................................................................... 287Overall Discussion Implications............................................................................................................... 289 Implication of CAPS......................................................................................... 290 ........................................................................................................................... 291 CAPS, Communicant Network, and Model of Meso-Level Intercommunication vi ......................................................... 294CAPS and Problem-Solving Potential ................................................................... 299CAPS and Typology of Publics ......................... 303CAPS and Criticisms of the Situational Theory of Publics ............................. 308CAPS and Active Publics without Information Seeking ................................................................................... 309CAPS and Campaign .................................................................... 310CAPS and Opinion Leadership .............................................................. 312CAPS and Diffusion of Innovation .............................................................. 312CAPS and Agenda Setting Theory ...................... 314CAPS and Coorientation: From Competitors to Cooperators Implication of CEPS ......................................................................................... 316 .................................................................... 317CAOS and Dissonance Theory .............................................................. 317CAOS and Information Selectivity ............................................... 319CAOS and Theories of Behavioral Intention ................................ 323Cognitive Multilateralism and Hedging and Wedging ............................ 325CEPS and Characterisitcs of An Excellent Organization Implication of STOPS....................................................................................... 325 ....................................................................... 326STOPS and Public Relations ............................................. 331STOPS and Health and Risk Communication .......................................................... 333STOPS and Political Communication . 334STOPS and Communicaiton of Science Limitations ................................................................................................................ 336 Validity of Findings .......................................................................................... 336 Generalizability of Findings ............................................................................. 337 Cross-validation with Single Sample................................................................ 338 Nonrecursive Model Identification................................................................... 338 Similarity between CAPS and CEPS Survey Instruments................................ 339 Demographic Variables and Problem Sensitivity............................................ 339 Future Research ........................................................................................................ 339 APPENDIX A........................................................................................................... 344 REFERENCES ......................................................................................................... 359 vii List of Tables Table 1 Phases of Communicant Activeness and Active and Activist Publics??.....?68 Table 2 Cognitive Alpha and Omega Strategies and Cognitive Entrepreneurship in Problem Solving????????????????????...??...?107 Table 3 Reliability for Communicant Activeness in Problem Solving (n = 1380)..........195 Table 4 Items, Factor Loadings, and Reliability for Information Forefending (n = 1380) ??????????????????????????.....?..196 Table 5 Items, Factor Loadings, and Reliability for Information Permitting (n = 1380)?????????????????????????????197 Table 6 Items, Factor Loadings, and Reliability for Information Forwarding (n = 1380)?????????????????????????..???..198 Table 7 Items, Factor Loadings, and Reliability for Information Sharing (n = 1380)? ??????????????????????????.......199 Table 8 Items, Factor Loadings, and Reliability for Information Seeking (n = 1380)?????.?????????????????????..?.... 200 Table 9 Items, Factor Loadings, and Reliability for Information Processing (n = 1380)?????????????..?????????????.??.201 Table 10 Reliability for Cognitive Entrepreneurship in Problem Solving (n = 1380) ?????????????????????.?????..??205 Table 11 Items, Factor Loadings, and Reliability for Cognitive Retrogression (n = 1380) ????????????????????????..????.206 Table 12 Items, Factor Loadings, and Reliability for Cognitive Suspension (n = 1380) ????????????????...???????..????..207 Table 13 Items, Factor Loadings, and Reliability for Problem Recognition (n = 1380) ?????..??????????????????..????.. 212 Table 14 Items, Factor Loadings, and Reliability for Constraint Recognition (n = 1380) ?????..??????????????????..????.. 213 Table 15 Items, Factor Loadings, and Reliability for Level of Involvement (n = 1380) ?????..??????????????????..????.. 214 viii Table 16 Items, Factor Loadings, and Reliability for Referent Criterion (n = 1380) ?????..???????????????????..???.. 215 Table 17 Items, Factor Loadings, and Reliability for Situational Motivation in Problem Solving (n = 1380) ??????????????.??????.??....216 Table 18 Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (War in Iraq) ???????..????????????????????..219 Table 19 CAPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (Losing Weight)??..??????????????????....220 Table 20 CAPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (Eliminating Affirmative Action)?????????.?????..221 Table 21 Variance Explained and Coefficient H in CAPS???.??????.?.. 228 Table 22 CEPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (War in Iraq)????..???????..???????????...??233 Table 23 CEPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (Eliminating Affirmative Action)?????.??????.??..???...234 Table 24 Variance Explained and Coefficient H in CEPS??.??.??????.. 235 Table 25 SITCAPS Model with Situational Motivation in Problem Solving: Chi-square Differences and Goodness-of-Fit indices?????..??????.???244 Table 26 SITCEPS Model with Situational Motivation in Problem Solving: Chi-square Differences and Goodness-of-Fit indices?????????..?????.250 Table 27 Nonrecursive Model: Chi-square Differences and Goodness-of-Fit Indices ?????.????????????????????.??..256 Table 28 Additional Analysis of CAPS (Cognitive Alpha versus Cognitive Omega Groups): Chi-square Differences and Goodness-of-Fit Indices?.??.??.. 265 Table 29 Dimensions of Communicant Activeness and Types of Publics????.... 300 ix List of Figures Figure 1: Conceptual relationship between information acquisition and communicant activeness. ????..????????????????.??????29 Figure 2: Conceptual relationship between information transmission and communicant activeness...???????...?????????.???????...?.. 33 Figure 3: Conceptual relationship between information selection and communicant activeness. ????..???????.???????????.???.. 43 Figure 4: Three-dimensional representation communicant activeness?..????? 62 Figure 5: A model of communicant activeness in problem solving. ???.????.63 Figure 6: Sequence model of communicant activeness..????..????...??...65 Figure 7: Cognitive strategies in problem solving..????..?????.???.....81 Figure 8: A model of cognitive entrepreneurship in problem solving??..????106 Figure 9: Situational motivation and perceptual antecedent conditions.?????..151 Figure 10: Sequence illustration of the situational theory of problem solving?.??156 Figure 11: A model of the situational theory of communicant activeness in problem solving (SITCAPS)??????.???.???.???.???.???....157 Figure 12: A model of the situational theory of cognitive entrepreneurship in problem solving (SITCEPS)???....???????????????.????.158 Figure 13: A model of nonrecursive relationship between CAPS and CEPS (Model I).160 Figure 14: A model nonrecursive relationship between CAPS and CEPS (Model II)?161 Figure 15: CAPS (War in Iraq)...????..????????????.??...?..226 Figure 16: CAPS (Losing Weight)????..????..???.?????......?.. 226 Figure 17: CAPS (Eliminating Affirmative Action)??.????..????..??..227 Figure 18: CEPS (War in Iraq)?..????????????????.?..??.. 236 Figure 19: CEPS (Eliminating Affirmative Action)..??????????....?.?.236 x Figure 20: SITCAPS (War in Iraq).??????????????.?.?..???243 Figure 21: SITCAPS (Losing Weight)?.??????????????.???..244 Figure 22: SITCAPS (Eliminating Affirmative Action)..?????.??..??...?.245 Figure 23: SITCEPS (War in Iraq)?..????..???????..??.?.???..251 Figure 24: SITCEPS (Eliminating Affirmative Action).??.????...?..???.. 251 Figure 25: Nonrecursive model between CEPS and CAPS (War in Iraq).?????.258 Figure 26: Nonrecursive model between CEPS and CAPS (Eliminating Affirmative Action)???.????..????????????..????.??.?..259 Figure 27: CAPS comparisons between cognitive alpha and omega groups (War in Iraq)?????????????????????????????..263 Figure 28: CAPS comparisons between cognitive alpha and omega groups (Eliminating Affirmative Action)?????????????.??????.???...264 Figure 29: Standardized structural path coefficients between cognitive alpha and cognitive omega group (War in Iraq)?????????????.???..267 Figure 30: Standardized structural path coefficients between cognitive alpha and cognitive omega group (Eliminating Affirmative Action)??..?.?.???..267 Figure 31: Standardized structural path coefficients between cognitive alpha and cognitive omega group (War in Iraq)???????..?.????.???..268 Figure 32: Standardized structural path coefficients between cognitive alpha and cognitive omega group (Eliminating Affirmative Action)..??..?????..268 Figure 33: Meso-level intercommunication process among communicants??.?.?..293 Figure 34: Communication behaviors and problem-solving potential in controversial social issues??????????????????????????..295 Figure 35: Three dimensions of communicant activeness and eight types of publics.....302 Figure 36: Effects of communicant activeness to perceptual and communicative behavioral coorientation among problem solvers????????.???.. 306 Figure 37: Conceptual model of the theory of reasoned action.?..????.???.. 320 Figure 38: Conceptual model of the theory of planned behavior?.????....??.. 321 xi CHAPTER I: INTRODUCTION The purpose of this dissertation is to develop a situational theory of problem solving. The situational theory of problem solving inherits, refines, and extends theoretical virtues of the situational theory of publics (STP) (J. Grunig, 1968, 1989, 1997, 2005) to become a more general theory of human communication and cognition during problematic situations. The situational theory of publics has contributed not only to the communication field in general but also to the public relations field in particular, and it has a potential to become a general communication theory. Specifically, I propose to broaden the situational theory of publics from simply information taking to a more general conception of communicative behavior, namely, communicant activeness in problem solving, which identifies such qualities as information taking, information giving, and information selecting. I differentiate six subdimensions in communicative activeness, which include information forefending and information permitting (information selection), information forwarding and information sharing (information transmission), and information seeking and information processing (information acquisition). I make this dichotomous subdivision of six variables in terms of proactiveness and reactiveness in communication action. In addition, I propose another new concept, cognitive entrepreneurship in problem solving, explaining cognitive strategies in problem solving. It captures a varying entrepreneurial mindset in approaching problem resolution. This new concept consists of four subconstructs: cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspension. 1 People respond to problems with different cognitive approaches corresponding to their situational perceptions. When people recognize a problem to which they feel closely connected and have sufficient cognitive resources, they are likely to use a forward reasoning strategy. In a forward reasoning strategy, people tend to look for evidence before making a choice, and thus the evidence directs a certain, optimal, conclusion. In contrast, when people recognize a problem as important but are severely constrained by lack of cognitive resources or a strong desire for a specific outcome, they tend to use a backward reasoning strategy. The backward reasoning strategy is a decision-making strategy that flows from a conclusion to certain evidence that best optimizes the hastily drawn a priori conclusion. Thus, the backward reasoning strategy refers to an optimization process for an a priori conclusion. I hypothesize that as one?s cognitive entrepreneurship increases, the person is less likely to adopt a backward reasoning strategy. I refer to the backward reasoning as cognitive retrogression. In addition, people tend to consider more ideas and options as they become more entrepreneurial in problem solving. As a result, they display more cognitive breadth, requiring more tolerance for even those competing ideas and rather incompatible perspectives. I refer to such breadth and tolerance as cognitive multilateralism. Next, as one becomes a more entrepreneurial problem solver, one tends to be more enthusiastic and patronizing toward the proposed solutions. I call this cognitive commitment. Entrepreneurial problem solvers are typically excited by new ideas. They have a voracious appetite for information that will help them solve a problem. At the same time, entrepreneurs are energetic and enthusiastic about a wide range of ideas, even wild ones, as long as they increase the potential to solve a problem. Thus, their 2 excitement with the possible ideas and options will grow as their cognitive entrepreneurship in problem solving increases. Finally, I hypothesize that as one?s cognitive entrepreneurship increases, one becomes more heedful and takes all steps before finalizing a solution. I name this cognitive suspension. En route to a final solution, a problem solver with heightened cognitive entrepreneurship will invest more discriminatory efforts in evaluating options and reevaluating a selected option before finalizing it. This dissertation attempts to move the situational theory of publics to a more general level by incorporating two new dependent variables in its theoretical formulation. After inclusion of a generalized dependent variable of communication behavior, the situational theory should be able to explain not only when and why people communicate but also how they communicate. If publics are active in problem solving, they are more likely to seek, give, and select information. Likewise, after inclusion of differential cognitive strategies that publics might adopt, the situational theory should explain better how people mentally practice their problem-solving task in a given problematic situation. At some times, publics may be more entrepreneurial in problem solving, and thus they use a forward reasoning strategy (i.e., an optimal conclusion search process). At other times, they may be less entrepreneurial and thus adopt a backward reasoning strategy (i.e., an optimization process of a priori conclusion). Scientific progress is possible not only when theorists introduce new theory and concepts but also when theorists increase the abstraction??generalization??of available concepts and theory. As Popper (1963) said, a theory is ?preferable?which tells us more?which contains the greater amount of empirical information or content? (p. 217). 3 Theory should tell more by abstracting and ?replacing the particular by the general? (Kruglanski, in press, p. 3), and we need to ?relentlessly? seek a general theory. Similarly, J. Grunig, the father of the situational theory of publics, said that no good theory ever stagnates (J. Grunig & Childers, 1988). In this vein, this dissertation aims at a good and general theory about human problem solving. The new situational theory of problem solving replaces ?the particular? (information seeking and processing) ?by the general? (communicant activeness in problem solving). As a result, it brings a ?greater amount of empirical information and content? (theoretical predictions and accounts). What follows is a brief history of the situational theory of publics. Situational Theory of Publics The situational theory of publics was built to explain why and how people communicate (J. Grunig, 1968, 1989, 1997, 2005). The situational theory is a purposeful, teleological theory that predicts the communicative behaviors of publics that most matter to public relations practitioners. This theory has helped define the field of public relations by spelling out who are publics in public relations. It refines, improves, and formalizes two classic theories of public opinion, that of John Dewey (1927) and Herbert Blumer (1966), so as to identify publics and measure their opinions. According to Dewey (1927) and Blumer (1966), publics are critical components of the democratic process who find problems affecting them and organize and act similarly for problem resolution. J. Grunig?s (1968, 1989, 1997, 2005) situational theory of publics provides a means to categorize varying compositions of publics in terms of responsiveness to problems; amount of and nature of communicative behavior; the effects of communication on cognition, attitudes, and behavior; and the potential to participate in 4 collective behavior in problem resolution (J. Grunig, 1989, 1997, 2005). Because the situational theory has the power to explain and predict who is most likely to communicate actively about social or individual problems, it has been a most heavily used applied communication theory, not only by public relations theorists but also by public relations practitioners. For public relations theorists, the STP provides a critical means to build a body of strategic management of public relations (e.g., the IABC Excellence Study, L. Grunig, J. Grunig, & Dozier, 2002). For practitioners, the STP guides the preparation of communication programs and makes them more strategic (e.g., by identifying publics, choosing realistic short and long-term communication objectives, and evaluating the outcomes of program effectiveness). Above all, the most important lesson from the situational theory is that audience information consumption is, in general, random. Information consumption becomes non- random or systematic only when people find that information matches their subjective life problems. The problems in people?s minds contain a limited subset of many possible problems. Only problems that have relevance to their lives will enter into publics? minds. Therefore, in their communication behavior, people are selectively systematic to meet their internal priorities that are influenced by situational perceptions. Notably, problems are situational?they come and go?in publics? minds. Publics actively communicate only when they experience problematic situations, and thus problems come and go. As their perceptions of problematic situations change, their communication behaviors change situationally. Just as communication behaviors are situational, therefore, publics arise and disappear situationally. 5 Message senders would like to define the list of problem priorities for publics. Organizations want their publics to behave in a way that the organization wishes (e.g., accept new business policy as it is). However, this is not an easy pursuit. Disappointingly to message senders, publics identify problems themselves. Likewise, publics define the priorities of their problems themselves. In the strategic management of public relations (L. Grunig, J. Grunig, Dozier, 2002), the situational theory of publics shows why blind pursuit of a maximum number of people in a general public?opposite to specific publics defined by the STP?fails. Because people selectively invest their cognitive and communicative resources only when they perceive it as necessary and relevant, massive audience campaigns, no matter how well intended, often frustrate organizational communicators (i.e., message senders) with poor success. Thus, public relations scholars use the implications of the situational theory to advise public relations practitioners first to identify who is likely to communicate with their organization and next to suggest that only those active publics, the subset of population who are interdependent and interpenetrating with the organization, will have strategic potential for the organization (J. Grunig & L. Grunig, 1997). Only these critical segments of the environment possess the potential to maximize opportunities and to minimize threats for the organization (L. Grunig, 1992). Thus, the situational theory of publics logically leads public relations practitioners to selectively identify and invest resources in communicating with active segments of their environment and in building 6 long-term, quality relationships with the strategic publics without cost-ineffective random communication (e.g., mass 1 -oriented campaigns). Situational variables. The prime interest of the theory is to account for the dependent variables, information processing and information seeking. The two dependent variables address the question of who is more or less likely to communicate about some problems of interest. Three independent variables are problem recognition, constraint recognition, and level of involvement. In brief, one becomes active in learning some new information (information seeking) when she or he identifies something as missing in life situations and stops to think about it (problem recognition). When the problem is perceived to be closely connected to oneself (level of involvement), she or he would do something to resolve the problem. However, she or he would feel more or less capable in taking remedial action across different problems (constraint recognition). In case she or he perceives there is an unmanageable obstacle to do something or feels less connected to the problem, even if the given situation were perceived to be serious, she or he would remain passive in communication (information processing) or not communicate at all. Problems of Communication Theories While communication theories have advanced our understanding of communication behavior, a majority of communication theorists have focused on communication as information learning or consumption by the audience. This trend has limited the scope of communication research in a number of ways. 1 Blumer (1966) defined mass as heterogeneous and public as homogenous. In brief, members of a mass become a heterogeneous collectivity because they turn to the same mass medium or because they share the same demographic characteristics (e.g., living in the same region). 7 Sender-bias in communication research and its consequences. Paul F. Lazarsfeld coined the term administrative research in communication to refer to a series of research programs whose main purpose is to enhance the effectiveness of message senders (e.g., persuading an audience). It addresses questions meeting the sender?s communication needs, such as: By which channel, can we most effectively reach our audience? To what extent will a mass-media campaign be counteracted by personal networks among audiences? How can we isolate the effects of a campaign from the influences of audiences? interpersonal communication? (Chaffee, 1982). McQuail (1997) classified the most common research goals in communication as ?measuring actual and potential reach for purposes of advertising,? ?manipulating and channeling audience choice behavior,? ?looking for audience market opportunities,? and ?product testing and improving communication effectiveness? (p. 15). He found the most fundamental division of research purposes exists between audience control and audience autonomy. McQuail wrote: By far the greatest quantity of audience research belongs at the control end of the spectrum, since this is what the industry wants and pays for (Beniger, 1986). Few of the results of industry research appear in the public domain and are consequently neglected in academic accounts of the audience. (McQuail, 1997, p. 16) In the same vein, most communication literature considers communication something that a source does to a receiver, as something that always originates from the source (J. Grunig & Hunt, 1984). Researchers conceptualize communication behavior mainly as information consumption, so that message senders can predict when people are more likely to buy or learn new information (e.g., information seeking or processing). 8 Few theorists, especially those who are empirically oriented, delve into communication behavior beyond information taking. According to Reeves, Chaffee, and Tims (1982), "Only recently have mass communication researchers begun to question the sender orientation embodied in the set of questions [Lasswell's (1948) who, says what, in which channel, to whom, with what effect?] and to propose new ones." (pp. 287-288). Applied communication fields such as public relations have also shown little interest in the need for studying communicators? voluntary information transmission and selectivity. Recently, there exists a burgeoning interest to better understand information dispersion through personal networks in advertising and marketing communication (i.e., word-of- mouth campaigns, Rosen, 2000; Richins, 1983). Yet, its purpose is originated from a sender?s perspective to enhance information taking of new product and service among potential consumers. Most studies about information giving are often fragmented or subsidiary to studies advancing knowledge of effective information learning among communicators. Why, then, is such a conceptual limitation persistent in the field? This is largely due to the legacy from social learning theory that considers human beings to be targets whose passiveness and ignorance require education and reformation. It looks for a best way to enlighten and educate ignorant masses in a way that the sender defines desirable. From such an asymmetric and limited perspective, we see trees but not the forest in communication research. This dissertation is meant to restore our conceptual orientation from a fragmented part (information acquisition) to an integrated whole (information acquisition as well as information transmission). The prior fragmentary view of communication behavior 9 deemphasizes communicants? voluntary acts of information selecting, producing, sharing, and transmitting. It is a consequence of sender bias and adoption of a media perspective. In contrast, the integral view of communication behavior accounts for information taking and giving to explain how communication roles (sender vs. receiver) and communication action (information giving vs. information learning) have been confounded. From audience behavior to communicant behavior. Historically, communication researchers preferred the term audience in referring to people engaged in communicative action. However, equating audience behavior exclusively with communication behavior is an overgeneralization. McQuail (1997) reviewed the historical evolution of the concept of audience and concluded, ?we keep the familiar word [audience], but the thing itself is disappearing? (McQuail, 1997, p. 2). He pointed out: Beyond commonsense usage [of the term, audience], there is much room for differences of meaning, misunderstandings, and theoretical conflicts. The problems surrounding the concept stem mainly from the fact that a single and simple word is being applied to an increasingly diverse and complex reality, open to alternative and competing theoretical formulation??what is occurring is the breakdown of the referent for the word audience in communication research from both the humanities and the social sciences? (Biocca, 1988, p. 103) [italics added]. (McQuail, 1997, pp. 1-2) This is a problem of trying to do too much with one term. The theories adopting the audience concept produce merely phenotypic knowledge to distinguish communicators by ?place? (e.g., local media), ?people? (e.g., age or gender group), ?particular type of medium or channel? (e.g., Internet or cable), ?content of its messages? (e.g., subject matter), and ?time? (e.g., prime-time vs. daytime audience) (McQuail, 1997, p. 2). Thus, audience and audience behaviors, the central concepts in studies of mass communication, become misnomers that indicate the ?breakdown of the referent? (McQuail, 1997). 10 Because of the limitation of this conceptual misnomer, communication research remains in a ?flatlander thinking? (Abbot, 1952)--failure to think in all dimensions of communication behavior. Without a proper concept, we fail to think outside one narrow area of experience or interest. We cannot perceive the other aspects of communicators beyond the information taker implicit in the concept of ?audience.? In addition, with the limited concept of audience, we miss valuable research questions other than a communicator?s receptiveness. We end with knowledge, at best, about how and why audiences behave, not how and why communicators behave. Hence, I propose a term, communicant, as a general name for encompassing both audience and sender of messages. In this newer perspective, communicant behavior subsumes audience as well as sender behaviors and recipient behaviors. In next section, I will discuss needs and reasons for bringing new concepts of communicant activeness and cognitive entrepreneurship in problem solving. These concepts try to overcome the limited assumption in the studies of communicant behaviors and problem solving. Communicant Activeness and Cognitive Entrepreneurship in Communication Research Why do we need to study communicant activeness in problem solving? I selectively concentrate on communicant activeness, not passiveness, in studying communicant behavior. Such a delimitation for the concept of communicant behavior is necessary because the human default in communication behavior is apathy or passiveness, not ardor or activeness. Irrespective of our awareness, human beings are constantly encountering and being affected by life problems. For example, an abrupt hike in tuition fees exhausts my money in the savings account for spring break. Global warming has 11 gradually increased electricity usage, which affects my utility bills. Election of a certain political figure may increase my taxes. Or, a tsunami may threaten my new house at the beach. The lists of our life problems painstakingly grow and continue. Thus, enlisting and tracking the whole list of problems that affect us is simply impossible. We know that we are less likely to recognize the consequences of some problems until they emerge. It is one thing to be connected with something and another to recognize its presence. When we realize that an almost infinite number of things defines our current state of being, it becomes clear that the human default communicative characteristic is passiveness, not activeness. We become active in communication only when we face a life problem that has significant consequence for us. Communication is a purposeful act, a ?tool for solving problems? (J. Grunig, 1997, p. 11). Just as we cannot recognize every problem that exerts influence on our current state of being, we cannot communicate about everything to which we are connected. Hence, the notion that we are lethargic for most of problems is not shameful, but a modus vivendi or sustenance mechanism. We have learned this from our evolutionary process. Thus, it is not clever to study about ?not doing,? which is uninteresting and hard to observe. Instead, the promise lies in studying about ?doing.? Logically, a better way to inquire about communicant behavior is to delimit the scope to how, why, and when we communicate, instead of how, why, and when we do not communicate. Therefore, the central concept in this dissertation is the narrow concept of communication activeness in problem solving. Communicant activeness is the behavioral alpha and omega of information traffic among social actors. It triggers a social process of problem solving by generating movements of words or symbols among communicants? 12 minds. I choose to study such movements rather than immovable silence so that we can learn better about social process of problem solving. Next, why is it necessary to study a cognitive approach in problem solving? And why does a cognitive entrepreneurial mindset matter? To answer, I should note that people take different mental approaches corresponding to the kind of problems they have. When they have a problem with which they are not connected or well-trained from past experiences, they enjoy a mental idleness in tackling the problem. In other words, we have little need to make extraordinary efforts to solve trivial or solution-ready problems. In contrast, people take a more considered and risk-taking approach when a problem is very important or lacks an easy solution. They find a strong need to be entrepreneurial? extraordinarily hardworking to build a new solution?to return to a default mental idleness. In such instance, we are cognitively effortful in problem solving. Depending on problem types and our readiness for solution, our cognitive strategy moves from more entrepreneurial to less entrepreneurial. Therefore, a problem solver?s cognitive strategy is a variable, not a constant across varying problems. Although we would invest the same amount of cognitive and communicative resources to deal with a problem, our choice of mental approach (e.g., backward or forward reasoning strategy) will result in different problem-solving potentials. In other words, a problem solver?s selection of a cognitive strategy becomes a strong predictor for how satisfactory our problem solving will be. In the earlier version of situational theory, J. Grunig (1968) introduced the term ?entrepreneur? to characterize people with extraordinary problem-solving efforts who are 13 most successful in problem solving. He defined the ?entrepreneur? as someone who is actively seeking a solution: Entrepreneur is defined as ?strategic decision maker? who ?skillfully manages the resources at his command, which means he is more than a routine manager; he is always looking for the most efficient way of doing things. The entrepreneur is ?rational? not in the sense that he is always a profit maximizer or seeks always to maximize a pre-set goal, but rather in the sense that he recognizes alternative solutions to his problematic situation, evaluates these alternatives, and chooses one of them [italics added]. (J. Grunig, 1968, p. 4) Entrepreneurs are the innovators who devise new ways of working out their problems or who adopt new and strange methods in tackling a problem. In general, they are more successful in dealing with their life problems. Hence, theorizing about what an entrepreneurial mindset is and under what conditions people would have more (less) entrepreneurial mindset will enrich our understanding of the human problem-solving phenomenon. I note that communicant activeness or passiveness is an extrinsic behavioral blossom of the intrinsic cognitive strategy that one takes within a life situation. That is, communicant activeness is a phonotypical phenomenon reflecting underlying genotypic internal cognitive strategies one adopts in a given situation. For this reason, to study when a communicant takes a more entrepreneurial mindset can also deepen our understanding of communicant behavior. At the same time, the way we deal with information over the course of problem solving should also affect our cognitive approaches in thinking about the problem. Therefore, such bidirectional causal influences should be studied to better understand how we approach as well as how we should approach problem solving. 14 To conclude, I delimit the scope of this dissertational study to our cognition and communication. Studying our cognitive entrepreneurship and communicant activeness in problem solving promises a better understanding of how and why questions in problem solving. Improved understanding of cognition and communication, in turn, will advance the situational theory of publics to a more general theory of problem solving. Methodology Theoretically, this study aims to develop new theoretical constructs about our cognitive and communicative features in problem solving and to propose a refined situational theory. The two emerging concepts are set as dependent variables explained by independent variables mostly from the situational theory of publics (J. Grunig, 1997). Methodologically, this study has two main goals: first to develop reliable and valid measurement systems and second to test the new theory. I thus used a survey to collect data, and this required a relatively large number of respondents who were willing to answer many questions. To meet the need for a large sample size, I adopted a snowball sampling technique, a non-probability sampling strategy, recruiting student participants and their acquaintances in the University of Maryland with exchange of extra-credit. Generally speaking, non-probability sampling is less desirable because it has severe limitations in generalizing the findings. However, I aimed at theoretical generalizability rather than statistical generalizability. Considering the goal of this study (developing a theory), such choice of data collection can be allowed. Because this study requires using human subjects, I submitted appropriate documentation to the Institutional Review Board of the University of Maryland. Upon approval, I proceeded to data collection. Throughout the data collection, all the 15 participants received a full explanation about the title and purpose of the study, their volunteer participation, time commitment, their freedom of withdrawal and right to skip questions, and potential harm and benefits from participation in the study. Participants were neither forced to participate nor to disclose information. Confidentiality was carefully secured. If those contacted student participants refused to participate, they were offered an alternative opportunity for extra-credit. Significance This dissertation presents a situational theory of problem solving that highlights distinctive communicative and cognitive features in human problem solving. Its purpose is to provide a simple and useful, but not atheoretical, account of communication behavior and the cognitive approaches that we adopt during problematic situations. This is a significant effort in that it expands and generalizes the situational theory of publics (J. Grunig, 1997, 2003). First, it is hoped that the emerging theory in this study will contribute to the communication field in general. I theorize communicative action as a purposive and instrumental act in dealing with life problems. The scope of the resulting theory is to explain human communicative characteristics in any life problems. The theory conceptually links communication action with problem-solving efforts we make. Thus, the resulting theory captures not only cognitive and communicative aspects, but also individual and social processes of problem solving. Those in the communication field have made few attempts to introduce a general theory of communication behavior. This situational theory of problem solving will address that deficiency. 16 In addition, as discussed earlier, theories and research in communication have often confined themselves to information taking or ?processing? in the name of audience research (M. A. Hamilton & Nowak, 2005; McQuail, 1997; Chaffee, 1982). I have attempted to overcome such a limited conception in theorizing the new situational theory. Over the course of problem solving, a communicant actively seeks, forwards, and shares, and selects information to be a more effective problem solver. Thereby, unlike audience behavior research, communication behavior in the situational theory is more general?a purposive action by communicant in information transmission, selection, and acquisition dimensions in dealing with problems. Second, in addition to communicant activeness in problem solving, I theorize cognitive entrepreneurship in problem solving to describe our distinct cognitive approaches toward problem resolution. The theory contains four distinct but correlated dimensions: cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspension. At times, we adopt a more entrepreneurial mindset. Cognitive entrepreneurship should have implications in applied communication. For example, in risk and health communication, the concept can explain how some interest groups would behave in a particular way and how they cognitively approach their problems. Thus, it is possible to improve public health intervention practice by modifying cognitive characteristics rather than difficult behavioral changes. In political communication, the new concept of cognitive entrepreneurship explains how voters make decisions with electoral information. In public relations, the concept will improve our understanding about how managerial cognitive characteristics in problem solving would affect communication and managerial excellence (L. Grunig, J. Grunig, & Dozier, 2002). 17 Third, this study addresses several criticisms of the situational theory of publics. For example, Vasquez and Taylor (2001) attacked the STP because it explains little of the ?nature, role, and influence of communication? and ?marginalized the role of communication process and dynamics? in the emergence of publics and social issues (pp. 149-150). Hallahan (2001) called for more research attention to ?issue processes? and ?issue dynamics? and communication processes in public relations research. Similarly, Cozier and Witmer (2001) requested a ?framework that shifts the locus of analysis to the public?s communicative practices in interactional settings? (p. 618). The new situational theory addresses such calls for new theorizing efforts regarding communication process and dynamics among problem solvers. Fourth, the new concepts and the situational theory offer several conceptual tools, such as a new typology of publics, predicting problem-solving potential in social conflicts, predicting ethical decision-making styles with a cognitive approach in problem solving, and a more comprehensive segmentation of publics. These are not only practical advancements in public relations practice but also theoretical advancements in that these new conceptual tools are derived from a continuation of theorizing with the situational theory of publics. Lastly, the situational theory of publics, the parent theory of the situational theory of problem solving, has posed and answered questions such as: What are publics and how do they arise?, With which publics is it possible to communicate and how can one communicate most effectively with each kind of public?, When and why do members of active publics join activist groups?, What communication effects are possible with each kind of publics?, How do activist publics differ from publics that have an intellectual 18 interest in an issue but do not get actively involved with the issue? (J. Grunig, 1997). The theory emerging from this study complements and refines the answers to these questions. Further, it continues to provide an answer for the ?new challenges? that have emerged from the theory-building process (J. Grunig, 1997). In the next chapter, I will explain the new concepts of communicant activeness in problem solving and cognitive entrepreneurship in problem solving. After the explication, I will introduce the refined situational antecedent variables that will be integrated with the two new dependent variables about cognitive and communicative consequences. 19 CHAPTER II: CONCEPTUALIZATION In this chapter, I first introduce two variables, communicant activeness in problem solving (CAPS) and cognitive entrepreneurship in problem solving (CEPS). The communication field has treated communicants mainly as information takers, and has not studied their information providing and selecting behaviors. The first new variable, communicant activeness in problem solving, addresses this paucity of research by adopting an integrated framework?that is, conceptualizing communicants not only as information takers but also as information givers and selectors?of how people address their life situations. This new variable deepens our understanding about when, how, and why people communicate as addressed by the situational theory of publics (J. Grunig, 1968, 1997, 2005)?i.e., how communication is used as a coping mechanism in problematic situations. Another variable, cognitive entrepreneurship in problem solving, also inherits the conceptual premise?i.e., entrepreneurial decision-making?from the earliest version of the situational theory of publics (J. Grunig, 1968). People in problematic situations take different cognitive approaches toward problem resolution corresponding to their subjective perceptions of a problematic situation. People are more entrepreneurial when they make a decision under some conditions and less entrepreneurial in other situations (J. Grunig, 1968). Because the chosen mental approach results in different outcomes, we need to theorize how people differ in their mindset in dealing with life problems. Although a good descriptive conceptual framework will promise a good prescriptive knowledge to guide better problem solving, there is also a paucity of research to describe different cognitive features in problematic situations. Thus, the new variable, cognitive 20 entrepreneurship in problem solving, is a worthy theoretical venture to extend our body of knowledge in problem solving. I next review under what conditions people take a certain communicative and cognitive mode of problem solving. For that purpose, the situational theory of publics (J. Grunig, 1997) provides a conceptual and measurement framework. The situational theory of publics proposed four independent variables (problem recognition, constraint recognition, level of involvement, and referent criterion). In conjunction with these four independent variables with the two new communicative and cognitive variables that features different problem-solving approaches, I propose a situational theory of problem solving (STOPS). The new theory generates a set of conceptual hypotheses that highlights the theory?s utility to the field of communication and public relations. A MODEL OF COMMUNICANT ACTIVENESS IN PROBLEM SOLVING Words are also actions, and actions are a kind of words. Essays: Second Series, ?The Poet? Emerson, Ralph Waldo 2 (1844) Generally speaking, two generations of researchers have studied communication behavior. The first generation focused on the sender. In market terms, the first generation focused on information supply and the information supplier?s interest about how information should flow (e.g., how consumers behave in responding to different supplying conditions.). Consequently, these researchers conceptualized communicants mainly as a target with varying degrees of receptiveness (information consuming potential) for the message that senders promote. Generation 1 asked ?Who says what to whom with which channel with what effect?? (Lasswell, 1948). 2 I thank Miia Jaatinen for finding this quote. Jaatinen (1999) cited Emerson?s essay in her book, Lobbying Political Issues: A Contingency Model of Effective Lobbying Strategies. 21 In contrast, the next generation emphasized the receiver?s point-of-view. Again using the market analogy, this generation attempted to explain communication in terms of information demand and demanders? interest. It described communicants as active information shoppers who consume information to meet their needs. These researchers asked ?Who hears what from whom via what channels for what purpose?? (Chaffee, 1982). However, neither generation identified communicants as active in information giving. They shared a common view of communicants as information takers. The two only differ in how active people are in information taking. Both approaches take information giving for granted and thus exclude it in theoretical explanations. Few researchers question why we have to limit our query only to the ?audience?s learning potentials.? Communication theorists by default took the sender?s side. They looked curiously at message recipients to increase their receptiveness. However, we can legitimately ask other questions: What would be a general theory of communication behavior that allows communicators as both, not either, sender and receiver under a single theoretical framework? Is it possible to integrate the sender?s as well as the receiver?s communicative behavior within a single theory? What features are common to sender and receiver that could fit into a single theory? A key conceptual link to incorporating these two approaches under a single conceptual roof is purpose of communication. In both approaches, the communicants, both givers and takers, use communication as a coping mechanism. A communicant seeks or disperses certain information for the purpose of problem solving. The contents that drive communicative action can vary, but the use of communication as a coping 22 mechanism is a constant. The role of communication and the purpose of communicant behaviors are identical in information taking and information giving. We can conceptually link both sender and receiver by their common purpose. They use communication instrumentally to deal with life problems. People generate and deliver messages (information) to others to solve their problems (e.g., promotional advertising). Likewise, people search for information to solve their problems when they find an absence of relevant information (e.g., reading medical journals). Senders communicate to solve their problems such as persuading people to adopt new ideas, practices, life styles and commercial products, whereas receivers communicate to solve problems such as remedial treatment or about a device to build a solution for their problematic life situations. Both message senders and message receivers share a common identity as communicants who consider communication as an instrumental and facilitative mechanism to cope with their life problems. In a nutshell, communication behavior, either giving or taking, becomes functionally identical. Hence, it should be possible and useful for us to build a theoretical framework that describes communicants as information givers as well as information takers simultaneously. This should lead us to a third generation of communication theories. A Need for General Look for Communication Behavior I follow Carter (1973) to define communication behavior as a movement of words or symbols by a person within a life situation. Communication differs from other behaviors in that a person may use it to plan other behaviors but not necessarily connect to other behaviors. Communication is a behavior in itself (J. Grunig, 1976). I began this chapter by criticizing the paucity of communication theories wherein simultaneity of 23 communicant roles can be housed. A prior approach (J. Grunig, 1989, 2005) segmented publics using the publics? differential responses, such as their differential nature and amount of communication behavior. Previously, the nature of communication behavior has been solely about the nature of one?s information taking. I question now why we should limit ourselves to information seeking and processing in defining communication behavior. Whereas communication behavior could vary in terms of learning of new facts, ideas, opinions, and attitudes from others, it also could vary in giving facts, ideas, opinions, and attitudes from one?s own knowledge storage. We can describe communicative behavior via as many dimensions as we want. Then, an important question is what kind of dimensions can help us for the problems we face. In the following section, I will propose a more general model about communication activeness. The significance claim of that model will be based on the ?empirical content of a theory? (Popper, 1999, p. 19). Empirical content refers to the class of empirical propositions that can be ruled out by a theory. The empirical propositions should be empirically falsifiable and subject to empirical testing. Thus, a theory contains more empirical propositions?i.e., a theory asserts more?takes more potential falsifiability and thus takes more risks. In other words, it is more subject to falsification. Hence, if a theory has passed tests of falsification, it contains a greater amount of empirical content because it has ruled out more empirical observations. A general theory says more, thus, it ?can clear up more problems???Its explanatory potential or its potential explanatory power is greater? (Popper, 1999, p. 20). It is desirable in this sense to advance a theory with a greater empirical content?a high- content theory (Popper, 1999). By proposing more propositions about communicant 24 activeness such as information giving and selecting, the new model of communicant activeness should have more empirical content than its predecessor, the situational theory of publics and other communication models of information consumption. Specifically, the propositions in the model of communicant activeness describe not only when people learn or consume information, but also when people produce and give and how they selectively take and give information. Situational Communicant Activeness Data, Information, and Knowledge According to Shannon and Weaver (1949), information is anything that reduces entropy and uncertainty. J. Grunig (1968) conceived of their definition as useful but not valid in that a definition of one construct should not be what it does but what it is. Instead, he advocated McDonough?s (1963) definition as having more merit. According to McDonough (1963), the common term information is composed of three components: data (unevaluated messages), information (data evaluated to apply in a specific problem situation), and knowledge (data evaluated for future use in general). Following these definitions, J. Grunig (1968) said that only information and knowledge can reduce uncertainty in a judgmental situation. Information refers to certain data that are judged to be specific and relevant to a given problem situation. All data are candidates for information (or knowledge), but not all data become information unless they prove their applicability and relevance to specific problem solving. In this dissertation, I adopt McDonough?s (1963) and J. Grunig?s (1968) conceptual explication of information. A person who recognizes a problem explores the sea of data ?i.e., the sea of unevaluated messages?inwardly and outwardly to palliate the perceived discrepancy. 25 The more one is capable of narrowing the perceived discrepancy by any means, the shorter the situation (i.e., the psychological time period of a problematic state) will be. Typically, one initiates an internal search for knowledge that has relevance to a current problem. Knowledge carries over from prior situations to apply to a similar kind of problems. This is ?knowledge activation? (Higgins, 1996). When attempted knowledge activation cannot yield an adequate solution, one turns to an external source of knowledge?i.e., information seeking. This is a knowledge action. Knowledge should first be available to the problem solver. Next, the available knowledge is evaluated for its relevance and applicability to a given problem. After it demonstrates sufficient relevance and applicability, it becomes information to be applied to a current problem. Likewise, data should be available and then evaluated whether relevant and applicable to a given problem. I distinguish data, information, and knowledge so that information is the central concept in a problematic situation. Neither knowledge nor data can be used in itself without an evaluative process for the current specific problem state. Problem solvers judge it by their own subjective criteria, although their judgmental competencies vary. Communicant Activeness Communicant activeness in problem solving (CAPS) is a conflated construct to measure communicator?s heightened communicative behaviors by a trichotomic model. I conceptualize the nature of communicant activeness in terms of three domains ?the communicative behavioral aspects of information connoisseurship, information outflow, and information inflow. I call them information selection, information transmission, and information acquisition, respectively. I subdivide the tripartite domains of communicant 26 activeness further into six subdimensions by an activeness dichotomy, reactive and proactive. This results in the variables of information permitting, information forefending, information sharing, information forwarding, information processing, and information seeking. I hypothesize that the increased level of communicant activeness will result in increased communicative proactiveness in each dimension. CAPS is a key component of the situational theory of problem solving (STOPS), which provides a set of endogenous variables to be accounted for. Each dimension captures some unique characteristics of communicative activeness that a person possesses when one encounters a problematic situation. The theoretical assumption of CAPS is that we use communication behaviors to cope with problematic situations. To adapt into our never stable environments, communication behavior becomes a way we live as well as operate. Conceptualizing Communicant Activeness: Information Selection, Information Transmission, and Information Acquisition Main Postulate I delimit the communicant?s behavioral aspects to one?s information inflow and outflow and one?s selectiveness. Such a delimitation does not mean there is no other communicative behavioral aspect, nor does it suggest that other communicatory aspects are uninteresting. Rather, I purposefully select these three dimensions to fill the void in communication research that I feel is most problematic. I refrain from theoretical monism, that there is one best theory to represent the phenomenon of interest. Thus, I advocate ongoing theorizing efforts to construe communicant activeness better. 27 I first propose a guiding premise about communicant activeness. People use communication instrumentally and purposefully to solve their life problems. 3 Thus, their instrumental use of communicative acts increases when confronting important problematic situations. The general postulate is: The more one wants problem resolution, the more one?s communicative actions will increase. Further: The more one commits to problem resolution, the more one becomes selective in dealing with information, the more one becomes transmissive, and the more one becomes acquisitive about information pertaining to the problem. In the following I will elaborate on each dimension of communicant activeness. This will lead to empirical operationalization to build a set of testable measures. Information Acquisition: Information Processing and Information Seeking When one faces a problem, he or she starts a process to find a solution. If the problem is recurring, the person would have a transferable solution from prior experience and endeavors to solve it. Hence, once available, one starts an internal retrieval of a prior solution. If the transferred solution fits well into the new problem, the problematic situation will end soon with application of the readymade solution to the problem. In contrast, one may confront a totally new problem with no applicable prior solution. Such a void of applicable knowledge can produce a meta-problem, a problem about a problem. If one cannot find an applicable solution, a problem solver will take a longer time and have a harder time closing a problematic state. Then, the problem holder must make extraordinary efforts to build a de novo solution. This requires ?building blocks? 3 I temporarily define problem as a perceived discrepancy between expected and observed states regarding a domain of experience. This is a provisional definition. I will fully elaborate the constructs, problem, situation, and problem recognition, in the following section of situational antecedent variables of this chapter. 28 (Kruglanski, 1989), that is, information relevant to constructing a new solution. Consequently, one turns to external sources to forage for potentially applicable data and knowledge. Information acquisition refers to the varying extent of information-collecting efforts for problem solving. In general, the more a communicant becomes active in problem solving, the more one exerts efforts for information acquisition (Figure 1). Figure 1: Conceptual relationship between information acquisition and communicant activeness. Specifically, when one feels an urge to deal with a problematic situation, one is more likely to initiate information collecting proactively?information seeking. Otherwise, a ing information?information ased ility r Information Processing and Information Seeking communicant tends to be remain passive and reactive in tak processing. The closure of information acquisition efforts corresponds to the incre competence regarding subjective relevance of information acquired in constructing a solution. Once a communicant has built a solution and successfully tested its workab to a given problem, one?s competence increases and subsequently decreases his or he need for information. 29 J. Grunig (1997) developed a situational theory of publics to ?predict the differential responses? such as ?responsiveness to issues,? ?amount of and nature of communication behavior,? ?effects of communication on cognitions, attitudes, and behavior,? and ?the likelihood of participating in collective behavior? (pp. 8-9). The amount and nature of communication behavior has special importance in theory and practice in that it solves a critical problem among public communication professionals, that is, who is likely to learn and pick up, not to ignore, the information provided to In the situational theory of publics, thus, information processing and information seek them. ing r, s, m g eeking ot e and less active communicants because active communicants are eq t, are two dependent variables. The former represents a passive communication behavio which is ?unplanned discovery of a message followed by continued processing of it? (Clarke & Kline, 1974). The latter represents an active communication behavior, that i ?the planned scanning of the environment for messages about a specified topic? (J. Grunig, 1997, p. 9), or ?premeditated information seeking? (Clarke & Kline, 1974). The use of information in problem solving is functionally crucial in proble solving because availability and applicability of the information one possesses determines the likelihood of successful problem solving. As a problem solver has a more serious perception about a given problem, one?s need for more quantity and better quality of information increases situationally. Therefore, the less one perceives a situation as bein problematic, the less the person is acquisitive for information about the problem. Information processing characterizes the less active communicants; information s characterizes the more active communicants. However, information processing cann distinguish between activ ually likely to discover some message by continued processing of it. In contras 30 informa e ct up of n to refrain from using the STP to think about activeness beyond informa y knowledge? people begin to look for external sources, that is, information seeking in the tion seeking is a communicative characteristic exclusively representing active communicants? behavior. For that reason, active communicants are both high in information processing and seeking, but less active communicants are only high in information processing. Information Transmission: Information Sharing and Information Forwarding Two dependent variables in the STP--information seeking and processing--hav solved many practical problems of public relations practitioners. By the two distin communication behaviors, the STP helps public relations practitioners identify a gro individuals who are willing to be readers or partners in communicative interactions. However, the STP delimits its scope only to the learning aspect of communication behavior. It discusses public activeness more as an information consumer. However, there is no reaso tion taking. Indeed, with the limited scope of thinking about communicant activeness solely as information taking, we delay advancing a better theory about who publics are and what their characteristics are. One promising communicative dimension to extend our theoretical purview in studying publics is information transmission by communicants. To cope with their problematic situations, people make instant judgments (Bargh, 1996). Making a good judgment requires reliable and applicable judgmental clues and rules for problem solving. Such clues and rules are first sought by the internal inventor check of whether relevant (applicable) information or rule(s) are available in memory. When requests for guiding knowledge or referent rule(s) are unmet?absence of 31 STP. In reality, however, during this process, problem holders are not only ?learning? of what should be done but also educating others what would happen if something is not done an et es ople to eople It a larger tial ork for activist events. At the heart of such coorientation transmission. In summary, the more the communicant becomes active in problem solving, the more the communicant is likely to transmit information pertaining to the problem (Figure 2). d how much its consequences are as close enough to have direct effects on them. Such an effort is often captured in some social movement theories as how people g organized or form a group around a common problem. Yet, few communication theori directly conceptualize how it occurs. Communicative behavior has a central place in the process that connects pe each other not only by collecting and learning related information or solutions for problem but also by providing, provoking, or seeking sympathy for problems. P intuitively and unconsciously realize that a problem becomes easier to solve when it became a problem for others. The communication act of educating others about a problem is a coping strategy. It forms a collective around the problem. By talking about a problem with others, individuals free their secluded problem. For that purpose, information giving (talking about one?s problem) costs little, but the payoffs are huge. is a thrifty way to solve a problem by activating information traffic that makes body of people perceive a problem as similarly problematic. It also increases the poten for collective behaviors, such as participating in pressure campaigns, donating money, and engaging in voluntary w toward a problem is information exchange among problem holders, especially information 32 Figure 2: Conceptual relationship between information transmission and communicant activeness. Reasons for Information Transmitting People use communication not only to learn relevant information on their w finding a solution but to share the burden of a problem collectively. They also educate others about potential harms of and treatments for some problems. Why, then, do communicants give information to others? People use communication to build a solutio for a problem as w ay to n ell as to diffuse negative burdens (costs) of the problem (1/n, divide the bur nd solution publics den into a smaller pieces) and translate the attempted solution into action in the most active case. We can break down information transmission as problem forwarding a sharing. Active and activist communicants make efforts to spread their perceived problems widely as being worthy of attention from other people. At times, an organization that has negative consequences on its publics has to compete for publics? attention against activist publics (J. Grunig & L. Grunig, 1997; L. Grunig, 1992). Maltreatment of these negative consequences creates angry publics. Those active 33 disperse their problem perception and a self-serving solution to other publics (e.g., medi government officials, courts, other activist groups). Against the activist publics? communicative efforts, the organization finds it difficult to represent its perspective on the problem. Then the organization becomes active in problem solving. It not onl information to better explain its perspective, but also gives inf a, y seeks ormation it believes best represe ). irs who n ho ing the ion is nts its perspective to other publics. We commonly find such information transmission efforts both from publics and organizations. We initiate and forward the presence of a problem, our own diagnosis of the source of a problem, and a resolution method to stakeholders. Frequently, information giving such as problem forwarding or solution forwarding becomes the act of problem resolution. For example, you may be annoyed by noise every night (problem recognition Your investigation reveals that the source of the noise is your neighbors upsta fight almost every night. A solution would be to ask the neighbors to stop fighting or to call the police to stop the rude behavior. As such, the very act of information transmission becomes a problem-solving behavior. For another example, you may find that smoking smells bad and that your children are coughing as a result. You begin to look for the source of problem and find your neighbor is violating the no smoking rule i your apartment building. You would like to share your trouble with other neighbors w face the same problem by asking whether they are also affected by the smoking. Then, you may put up a ?no smoking? sign in the corridor to politely correct the neighbor?s misbehavior. You give information of the problem as well as a solution by inform affected others as well as those who are causing the problem. Information transmiss 34 a comm ng. to a les erly wever, information sharers are less likely to initiate their informa m g to olver a solution. Thus, at the earlier n is primarily about on resort among problem holders. As they learn and build a solution by information seeking, they educate and apply a solution when they face a problem. Information Sharing and Information Forwarding CAPS distinguishes two levels of activeness in information giving. A passive information giver shares when someone else requests their expertise in problem solvi The information sharer possesses relevant knowledge and decision rules applicable problem at hand. The information sharer has acquired knowledge and decisional ru from past problematic situations. In other words, an information sharer is a form active problem solver. Ho tion giving themselves. Rather, they proffer information only when being solicited by some information seekers. Information sharing is, thus, a reactive communicant behavior. In contrast, a more active information giver forwards information about a proble even if no one solicited it. An information forwarder is proactive in propagating his or her problem perception and preferred way of problem solving. Information forwarders are thus active communicants who exploit communication instrumentally for problem solving by reproducing a similar problem perception and a solution. Thus, information outflow from an information forwarder is voluntary and self-propelled by heightened problem perception. Notably, information transmission evolves from problem givin solution giving as a problematic situation continues. At early stages, a problem s communicates to obtain necessary information to build phase of a situation, a communicant?s information transmissio 35 problem sharing and forwarding. At the later phase, however, a problem solver communicates both the problem as well as a solution. Individual Problem Solver versus Collective Problem Solver Publics are disconnected systems of individuals experiencing common problems, and they can evolve into organized and powerful activist groups (J. Grunig, 1997; Grunig, 1992; J. Grunig & Hunt, 1984). Problem solvers can maximize their problem solving potential and minimize its costs when they are better connected and coordi in action. In essence, information giving increases connectedness among isolated individual problem solvers. Cross awareness about an extant problem and subseque behavioral coord L. nated nt ination is only feasible through information exchanges regarding problem rs. . She e same problem. In other words, coorienting among s communicants commonly face. Therefore, we can meaningfully distinguish problem solvers by the extent of activeness in information transmission (active vs. passive giving). Chwe (2001) explained that whether members of a collectivity have ?common knowledge? is the prime mark differentiating between disconnected and connected social groups. Cross meta-perceptions about whether members know about their common problem leads a transition from individual problem holder to collective problem solve The cross meta-perception of a problem demarcates between ?disconnected systems? and connected systems: e.g., I know I have a problem. I know she has the same problem knows I have the same problem. I know that she knows I have the same problem. She knows that I know that she knows I have th disconnected individual problem solvers toward a problem will transform 36 individuals toward connected collective problem solvers if they are aware of common perceptions about the problem they face. In terms of the coorientation model (J. Grunig & Hunt, 1984; McLeod & Cha 1973; Newcomb, 1953), the levels of individual problem solvers? heightened congrue ffee, ncy (i.e., th y ally, - if a solving ted to problem resolution, a comm ation sharing e extent to which each person thinks the other persons? idea or evaluation is similar to one?s own) and accuracy (i.e., the extent to which one person?s perception of the other persons? idea or evaluation approximates the other person?s actual idea or evaluation) is a necessary condition to form a social collectivity in problem solving. Importantly, communication behaviors among problem solvers are the only wa to increase common knowledge (congruency and accuracy) among publics. Specific voluntary information transmission about a problematic state and the following cross awareness of knowing each other?s meta-perceptions regarding the common problem is the primary mechanism that allows separated individual problem solvers to transform into interconnected and coordinated collective problem solvers. To summarize, communicant becomes active in the task of problem resolution, he or she is likely to make more effort in information transmission. Specifically, when one commits to a problem, he or she is more likely to initiate information transmission proactively? information forwarding. In contrast, if less motiva unicant tends to be remain passive and reactive in giving information?inform . The heightened likelihood of each problem holder to share and forward information about a given problem increases the collective problem-solving potential among disconnected individual problem solvers. Information Selection: Information Permitting and Information Forefending 37 Communicants tend to develop certain preferences in their ?usual diet of information? (Case, 2002, p. 93). Festinger (1957) and Hyman and Sheatsley (1947 were forefathers of research on communicator selectivity. Since then, social psychologists and communication researchers have done ample studies about selective exposure. Although communication selectivity has attracted many researchers, findin are often controversial (Frey, 1986; Sears & Freedman, 1967). Putting aside all the hubbub around selective exposure, I conceptualize information selection as a situationally evolving human communicative feature, not only to cope with the problem of ?cognitive discrepancy? reduction (Carter, Pyszka, & Guerrero, 1969) but also to cope with the problem of ?inf ) gs ormation reduction? (Evans, 1989, p. 112) that is a meta-problem to a s ) ists n are two lem recognizers?who find a discrepancy between expecta t communicant. Past researchers have explained that selective attention or exposure occur when people attempt to solve a problem of cognitive dissonance (Festinger, 1957, 1964 by taking information in a selective manner (i.e., avoid dissonant but approach consonant information). I am dissatisfied with the default dissonance reduction mechanism that cons solely of avoiding dissonant information or seeking consonant information (Festinger, 1957, 1964). Avoiding dissonant information and seeking consonant informatio ways, but not the only ways, to reduce an internal state of dissonance. In general, through information behavior, communicants can reduce dissonance in two ways. One is a revising expectation strategy. Prob tion and current observation?may modify their expectation state (e.g., ?conjectural knowledge? Popper, 1963) to reduce the perceptual distance (e.g., the exten 38 of departure from an anticipated state) by approaching to relevant information and by avoiding irrelevant information. The other is reinforcing expectation strategy. Problem recognizers may modify a current observation state by avoiding dissonant and/or seeking consonant information. Here, in any strategy, a communicant?s information selection will occur given that the communicant is highly active in problem solving. To resolve dissonance, one has to discriminate between irrelevant and r elevant information. This requires communicants to weigh ) , e communication did not use such a general conception of f information (i.e., specific interpretation pursuit). At times, communicants have low control in processing certain available information. Specifically, one could reduce a dissonant state either by weighing reinforcing information to modify a problematic expectation state (i.e., dissonance reduction by deflecting observation that is a pseudo problem-solving strategy or by weighing reforming information to modify a problematic observational state (i.e. genuine problem-solving strategy). Past studies in selectiv dissonance reduction strategies. In the former reinforcing expectation strategy, communicants tend to seek and forward any information that is subjectively relevant to reinforce prior expectation. In the latter, revising expectation strategy, communicants tend to seek and forward any information that is relevant to revision or refinement of content of prior expectation. Also, I put emphasis on selective interpretation and selective production of information in addition to selective access, processing, or exposure. I conceptualize information selection to include not only selective information taking and giving (i.e., specific information pursuit) but also selective interpretation o 39 unpalat ormation. a st e ving g elevant to the problem solving) in their communication behavior. Traffic ve able information (e.g., a TV news program). At other times, communicants have sufficient knowledge or confidence enough to face and tolerate distasteful inf Further, communicants may systematically seek dissonance-arousing information (e.g., communicant who anticipates a discussion with opponents). Finally, selective communication often occurs among communicants who need to reduce information during a problematic situation. As a result of heightened communicant activeness in problem solving, communicants tend to accumulate a va amount of potentially relevant information, more than they were able to manage. At som point, communicants face a problem about problem solving?i.e., managing an adequate information pool while considering the tradeoff between quality and quantity. Whereas communicants will take any information related to the problem at an early stage of problem-solving efforts, they begin to collect information discriminately at later stages. In sum, communicants situationally evolve from relation seeker (e.g., taking and gi any information related with the problem) to relevance seeker (e.g., taking and givin only information r king information?i.e., information inflow and outflow?is more and more under the guidance of certain discriminatory judging criteria as communicant activeness increases. The more one becomes an active communicant, the more one becomes a relevance seeker. In summary, information selection occurs either to solve the problem of cogniti discrepancy reduction (i.e., problem) or to solve the problem of information reduction (i.e., a meta-problem in dealing with a problem). In both cases, communicants use information selectively for problem solving (i.e., to reduce cognitive discrepancy or to 40 reduce information). In any instance, a communicant becomes increasingly more of a relevance pursuer whose perceived relevance is subjectively defined corresponding to changes in one?s situational perceptions. At one tim e, communicants may avoid some inform ation ful to revising old beliefs) and ignore or avoid information confirming prior belief a rmation lop sk ation selection results from a self-defined notion of e discrepancy reduction) or a meta-p cate ective ation-causing discrepancy as irrelevant (i.e., because one feels that information conflicting with a prior belief is wrong and unworthy) and approach agreeing inform as relevant (e.g., because information bolsters one?s confidence that a prior belief is appropriate and valuable) to the discrepant state. At another time, the same communicants may approach discrepant information as relevant (e.g., because one feels that information conflicting with a prior or present belief is use s irrelevant (e.g., because one feels that information supporting prior belief has little use in rescripting an old belief). In all cases, the communicant becomes a relevance pursuer, corresponding to the communicant activeness and to the amount of info available. In taking, interpreting, and giving information, communicants tend to deve some discriminatory rules about what kind of information is relevant in the current ta of problem solving. Thus, communicants solve a meta-problem (i.e., a need for information reduction) by situationally evolving from a relatedness pursuit to relevance pursuit. In either case, inform relevance in solving a problem (i.e., a need for cognitiv roblem (i.e., a need for information reduction). Before I conceptually expli what I mean by information selection, a brief review of those issues in studying sel communication is required. 41 Problems in Studying Selective Communication In the literature, almost without an exception, researchers have discussed communicant selectivity only in terms of selectivity in information collection and recollection?i.e., selective retrieval of internal memory or selective search for an external pool of information. There are few, if any, attempts to extend the scope human information selection beyond information acquisition. The common terms referring to studies in selective communication are selective exposure, selective perception, or selective communication. T of hese studies narrowly focus on passive informa people he lants, previous studies regarding ose ter communicant activeness, the more the communicant becomes selective in information acquisition as well as in information transmission (Figure 3). In the next section, I will discuss the issues in selectivity research. tion consumption. Selective communication studies have their roots in dissonance theory and Heider?s (1958) balance theory. Its main assumption is that tend to avoid dissonance-arousing information while looking for information reinforcing current beliefs, opinions, and decisions. Putting aside whether selective information processing or seeking is empirically tenable, I question why we limit selective communication to selective exposure. I alternatively suggest that we should study selective communication in terms of selective information giving as well. Such selectivity in information giving can introduce important knowledge to explain individual and collective problem-solving processes (e.g., via what mechanism an extreme view reproduces itself among active publics). Thus, t conception I propose here supplements, rather than supp communicative and cognitive selectivity with a general theoretical frame. I here prop that the grea 42 Figure 3: Conceptual relationship between information selection and communicant activeness. Selective processing, selective production, and selective propagation. As briefly mentioned, there is a paucity of literature on the phenomenon of selective communication as both acquisition selectivity and transmission selectivity. Thus, I raise the question of why we do not see selectiveness in terms of information production and propagation. Past research has invariably focused on the learning aspects of human information selectivity, namely, selective exposure. In his classic review of selective exposure to information, Frey (1986) surveyed the scope and history of selective exposure research up to that date. However, even his exhaustive review of informational selectivity remained exclusively within a notion of ?selective exposure to information??i.e., information inflow?with no consideration of information outflow (Frey, 1986). Such a narrow focus in communication research programs has its origin from the sender-oriented research paradigm (e.g., to increase audience susceptibility to educate or persuade target recipients 4 4 Chaffee, Stamm, Guerrero, and Tipton (1969) noted that selective exposure has some problems: ?at any rate, it is impossible to distinguish between these two kinds of behavior in that study, so the more inclusive term selective exposure is used, to indicate that selective behavior of either type is under observation? (p. 17). 43 as a message sender intends). Thus, conceptual balance in a subsequent review of selectivity is necessary to get generalized lessons from past studies of selectivity. In addition, the inconsistent findings in selective exposure studies have suggested that res (Sears & a ey V is ify, and translate information they are processing in a wa eness selective earchers should consider selectivity in terms of ?information evaluation? Freedman, 1967; Feather, 1963) or selective interpretation. Often, information medi allow little control for communicants in processing messages. Chaffee et al. (2001) noted that selective attention is not empirically observable from the less controllable media such as newspapers and television (news program and television political ads). As th summarized well, these media sources ?offer balanced coverage? and ?certainly some exposure to both sides is unavoidable when watching or reading news? (p. 263). Thus, audiences have ?little opportunity to avoid counter-attitudinal information? (Chaffee et al., 2001, p. 264). Instead of viewing selectivity solely as selective processing, it would be theoretically more promising to take selective communication as not only selective processing under some conditions (e.g., more constrained or presence of strong decision rules) but also selective interpretation. Indeed, people processing identical information reach starkly different conclusions. For example, voters who watched the same T presidential debate often conclude that their preferred candidate has won the debate. Th suggests that communicants resist, mod y consistent with their internal rules and beliefs. Hence, information selectiv not only comes in the way of selective processing (de facto selectivity) (Sears & Freedman, 1967) but also takes the form of selective interpretation or selective production. In addition, selective interpretation of some information results in 44 propagation through communicants? networks (e.g., if an active communicant processe and interprets information in a selective manner, she or he would also discharge it to others in that way). In the present study, I develop such a general conception of information selectiveness t s o include selective processing, selective interpretation, and selectiv dissona e. The mo alternative(s) are missed; and thus dissonance is easier to observe. To reduce postdecisional internal negativity (i.e., dissonance), one type of selectivity deals with information?i .e., uses a ?confirmatory strategy? by favoring information supporting one?s already made decision and/or by avoiding information challenging that decision. Frey (1986) summarized: The general hypothesis was that the search should differ according to whether it be relatively unbiased in their seeking and evaluation of information. Once the on- supporting (consonant) information and avoid decision-contrary (dissonant) evaluate the information found: Items of information that support the decision are . 44) in ested that there may be a ?biased on of e transmission of information. Postdecisional selectivity versus predecisional selectivity. In his cognitive nce theory, Festinger (1957, 1964) identified different effects from dissonanc st dissonance is aroused after a decision is made because counterfactual decision occurs before or after the commitment is made. Prior to a decision, people should decision has been made, however, selectivity sets in: People search for decisi information. This same bias is evident as well in the manner in which people often considered to be more credible and reliable than contrary information. (p Festinger (1957, 1964) and Frey (1986) said that selectivity could be observed ?postdecisional? processing, assuming that it cannot be observed in predecisional processing. However, Brownstein (2003) examined many studies conducted since Festinger?s (1964) dissonance theory was developed and sugg predecision processing.? According to Brownstein (2003), biased predecision processing may happen ?when decision makers restructure their mental representati 45 the decision environment to favor one alternative before making a choice? (p. 545). Specifically, this bias related to a selective information search that ?favors one alternative, or reevaluation of alternatives, in which one alternative is bolstered and/or the others denigrated until the choice becomes obvious? (Brownstein, 2003, p. 545). Review relevant research on theories of cognitive dissonance and selectivity (e.g., studies abou choice certainty or motivated reasoning) led him to favor, despite some disfavoring findings, the empirical presence of predecisional selectivity. I take a position that we can observe sel of t ective information behavior either in predeci e state s exposure studies resulted in a serious pessimism about cogniti o meet its theoret e results n equivoc sional or postdecisional processing. We frequently encounter a problematic situation wherein we hold a strong preference for or avoidance of a certain outcom (e.g., curing my beloved one?s disease). It often leads us to selectively pursue information that can enhance our subjective confidence and sustain hope for the desired end. At time we selectively forage for information to increase self-fulfilling confidence while suspending a final judgment (predecisional selectivity). At other times, we selectively collect information to reinforce a previous conclusion (postdecisional selectivity). Communicant selectiveness regarding information behavior, therefore could occur, either in a predecisional or in a postdecisional way. Reinforcement account versus relevance account. Freedman and Sears? (1965, 1967) reviews of selective ve dissonance theory because it had failed to generate empirical support t ical claims. Sears and Freedman (1967) pointed out: ?It is enough to say that th [from testing hypotheses derived from cognitive dissonance theory] are agai al? (p. 208) and ?there is no consistent result in this research? (p. 207). Feather 46 (1962, information than were nonsmokers. Brock (1965) failed to find clear support for selective exposure in his partial replication of what Feather (1962, 1963) had found. Freedman (1965) even found a strong preference for non-supporting information among those who experienced dissonance. Sears and Freedman (1967) reviewed this literature and concluded that: ? a considerable amount of experimental research has uncovered no general nces, people seem to prefer information that supports their opinions; under other ns. In no way can the available evidence be said to support the contention that people rmation. (p. 212) However, Sears and Freedman (1967) also noted the existence of some ?de facto selectivity.? ?Most audiences for mass communications apparently tend to over-represent persons already sympathetic to the views being propounded, and most persons seem to be exposed disproportionately to communications that support their opinions? (p. 212). They took it as ?paradoxical? and questioned: ?How can it be that people are in fact selective, yet display no trace of a general preference for supportive information?? (Sears & Freedman, 1967, p. 212). More recently, Chaffee et al. (2001) found that selectivity increases both in ?counter-attitudinal? and ?attitude-consistent? directions if a communicant becomes active for a problematic situation. The prediction of the direction in which one?s selectivity moves is inconsistent in their views. In brief, Chaffee et al. (2001) tested and found some support for the conclusion that the more people are involved, the more information they seek in both directions. 1963) found that smokers were more interested in both consonant and dissonant psychological preference for supportive information. Under some circumsta circumstances, people seem to prefer information that contradicts their opinio generally seek out supportive information and avoid nonsupportive info 47 Earlier, Carter, Pyszka, and Guerrero (1969) presented strong experimental evidence that selectivity can be in an opposite direction. They found from three experiments that whereas there was equal selectivity under a control group, two-thirds of the participants in a ?dissonance? condition chose to read an essay countering their favored position as the personal relevance increased. Carter et al. (1969) interpreted tha result as suggesting that dissonance should be reformulated ?simply as perceived cognitive discrepancy? so that avoidance would just be one possible communicative reaction en route t to discrepancy reduction. This finding tells us the concomitant presence of ?rev ef? is ent problem- solving rent e ersal? selectivity with confirmatory selectivity, unlike what dissonance theory had suggested. Dissonance theory adheres to the notion that ?preservation of current beli the only communicative goal in a dissonant situation (i.e., a pseudo problem solving). In contrast, what Carter et al. (1969) suggested was more general. At times, preservation of current beliefs can become irrelevant if one?s perceived involvement with a problematic state is high. Under such a circumstance, one would perceive the revision of curr beliefs as more relevant. At other times, preservation of current belief would be more relevant for situational reasons (e.g., high constraints against embarking on efforts) and thus looking for information increases confidence in one?s cur beliefs. Unlike the prior reinforcement account, Carter et al.?s (1969) relevance account better explained why some selective communication happens. Its major premise is: Communicants select some information over other information because of its subjectively defined relevance. Subjectively defined relevance in using information suggests that one?s judgment about information relevance depends on whether the communicant finds som 48 utility after due consideration of problem-solving conditions (e.g., constraints). Subjectively defined suggests that the perceived relevance by the communicant could be fairly different among communicants even if they have the same problem. One would attribute such results to personal bias, perceptual distortion, or overest n to t, elt it is more re unt xplains a a communicant can reduce discrepancy via any method? i.e., by r by imation of situational constraints. However, a given communicant?s judgment is the most important, no matter how biased it is, because one translates one?s perceptio action. What matters is the value defined by the eyes of beholder. Thus, at one time, a communicant would avoid dissonant information, not because it is unbearably dissonan but because the person felt it is less relevant to the task of problem solving. In other cases, one would approach some dissonance-causing information because he or she f levant to his of her effort to solve a problem. I adopt the relevance account as a more general explanation for selective exposure. In contrast to the reinforcement acco (e.g., dissonance theory), relevance explains not only the old findings but also e other selection types such as reversal selective strategy. The relevance account thus becomes a general explanation. Whereas cognitive dissonance theory adopted a reinforcement explanation, the present study will use a relevance explanation wherein reinforcing and revising strategies can be explained by single account. It allows that pseudo problem solving (e.g., selective exposure to reinforcing information) o genuine problem solving (e.g., selective exposure to reforming information). Whereas selective attention is a more ethereal approach to reducing a dissonant state (e.g., by deflecting one?s perception of environmental conditions), the problem solving approach is a more material approach to reduce perceived discrepancy 49 permanently (e.g., by bending one?s course of action to fit into environmental condit However, both app ions). roaches are functionally equivalent in reducing cognitive discrepancy. That is ful gy is tivation orients one to be more selective?i.e., to pursue a specific subset of infor he one has nce , the information one selects plays the role of augmenting subjective confidence either by strengthening a current belief (attempted solution) or by constructing a new belief (attempted solution). However, the reinforcing approach is a self-fulfilling strategy resulting from situational constraints and/or from a self-serving referent criterion (i.e., wishful or will thinking on the outcome state). On the other hand, the revising approach is a more self- reforming strategy through which communicants are free from situational constraints and have less of a self-serving referent criterion. The choice of a problem-dealing strate closely associated with the extent of constraint recognition and strength and type of referent criterion. In both cases, communicants travel the path from general to specific, from random to systematic, and from related to relevant in dealing with information corresponding to the level of communicant activeness. A communicant?s heightened situational mo toward problem solving mation to fit into the specific characteristic of the given problem content. T content of a referent criterion, its magnitude of influence, and the extent of situational constraints explain the type of information specificity. For example, if one?s problem is to anticipate a political discussion with opponents, he or she would use a revising selectivity to better attack an opponent?s view and defend his or her own view. If unmanageably high situational constraints in a given problematic situation (e.g., prese of another urgent problem), he or she would use a reinforcing selectivity. Or, if one has a 50 strong decisional reference readily applicable to the current problem (e.g., knowledge carried from past problem solving of a similar kind), he or she would again use a ith of dissonance and only preference for consonance, as dissonance theory predict a t us scrutinize the influence of reinforcing selectivity. To summarize, some situational parameters such as referent criterion and constraint recognition jointly lead communicants to be more or less selective as well as to turn to specific directional information in a given problematic situation. Thus, one who is devoted to a position would situationally pursue more countering information regardless of the current stance one favors. The same person could be a partisan pursuer of information who only keeps cognitive poise under serious commitment to a priority. However, at any rate, the strict and exclusive association of selective directionality w only avoidance s, is theoretically and empirically less tenable. Emerging Communicant Selectiveness during Problem Solving It is easy to be open and generous in taking information when we feel little connection (e.g., a small stake) to an issue the information is about. However, we tend to develop some needs and find a certain method to discriminate information when it begins to interrupt one?s routine commitments to other life problems. People without much of stake either become generous and open-minded or less caring and indiscriminant abou information that comes in and out of their minds. They feel little difficulty in permitting any position, opinion, or information to enter unless they perceive something being problematic. In contrast, people with high stakes (e.g., highly motivated people) need to develop a more sophisticated understanding and th 51 informa become become m solving, most communicants find it difficult to allow any rous in blem solver? often d from the want of time and wit to pursue it all. There is just too much material. The stacks? I would drown, or panic, and certainly lose my way. (pp. 29-30; as cited e ?send an assistant to get it so 94). He esearching, despite some potential relevance, which he thinks might be redundant and low in quality. Such ?nonuse? (Wilson, 1996) or ?filtering? (Case, 2002) commonly expedites, rather than distracts, from a given problem solving process. Similarly, Wilson (1995) observed and advocated efficiency and rationality in filtering out information?a ?nonuse policy??that is necessary when one is given more information than one could absorb (Wilson, 1995, pp. 45-46). To avoid being engulfed by tion more carefully (positive or negative, supporting or refuting). Thus, they often selective as a result of increased seriousness of a problem. Notably, when people more active in proble information to enter their minds. Although we know that being open to take, interpret, and give information is more desirable in problem solving, various reasons (e.g., too much seemingly important information) prohibit us from remaining open and gene dealing with information. Thus, selecting a manageable subset of useful information out of would-be useful information is a necessary but challenging task to communicants. Perrow (1989) observed what an active researcher?who is a pro oes in taking information during research: I require libraries to hide most of the literature so that I will not become delirious problem is not access, it is the reverse, containment?.Were I now to browse the in Case, 2002, p. 94) Perrow (1989) confessed that when he needs books and articles, h that he will not be distracted by adjacent materials? (as cited in Case, 2002, p. wants the assistant to keep him away from the literature in the areas he is r 52 a large supply of relevant information we make and apply rules in managing the information we face. Perrow (1989) added a similar observation: ?foreign? contributions, ignore contributions from identifiable schools and styles or with particular approaches (p. 199; as cited in Case, 2002, p. 95) Throughout problem solving, communicants tend to create some discriminatory rules to guide judgments that will be included and excluded. Case (2002) sharply distinguished such information ?filtering? from ?avoidance.? Information is often ?not avoided but rather simply not used? (Case, 2002, p. 95). Although using discriminatory rules could lower the quality of information, it can, at least, solve the problem of information overload?a meta-problem of problem solving. In summary, as a communicant becomes active, information selection or specification is heightened. In the following, I will discuss more closely sources, types, and strategies of communicant selectiveness. Defining information selection. Communicants with some heightened situational need for information soon would face another problem. Because active foraging of data often results in a huge pile of information that would surpass one?s cognitive capacity and cognitive capability (Kruglanski & Thompson, 1999), managing information effectively to build a solution to the problem is a meta-problem one must deal with. As Evans noted (Evans, 1989; Evans & Over, 1996), one often must reduce an overwhelming amount of candidate data in the problem solving task. Indeed, just as the absence of available knowledge is a serious problem, the abundance of available knowledge is an equally challenging problem?i.e., the embarrassment of riches. Hence, I conceptualize the problem of information reduction as another source of communicative selectiveness that Large literatures may be cut down drastically: one may ignore the past, ignore traditions of thought?ignore work done with certain techniques or in particular 53 communicants use in taking, interpreting, and giving information. In summary, during a problematic situation, there is a tradeoff between quantity and quality of would information. As the available poo -be l of data increases, a need for related and more informa y need for y. By information specification, regardless of its valence t n in ts or erion ty tion turns into a need for relevant but less information?a contradictor a narrowing down mechanism. Naturally, we carry out information taking, evaluating, and giving ?in a highly selective manner using some form of heuristic process? (Evans, 1989, p. 112). We adaptively opt for a strategy of selective search to increase the range of viable alternative solutions. In the present model, therefore, I define information selection as the degree of pursuing specificity evolving from random to systematic, from general to specific, and from related to relevant in dealing with data corresponding to a communicant?s activeness. The two general sources of communicant selectiveness arise from the problem of reducing cognitive discrepancy and/or the problem of reducing a surplus of information in problem solving. Selection types. I use the terms, information selection and information specification, interchangeably in this stud (i.e., counter to or consistent with current belief), I mean that a communican becomes selective if one develops some preference for a certain subset of informatio pursuit of problem resolution. Communicants develop more confidence in some subse of information. The direction of information a communicant will pursue (e.g., avoiding seeking dissonant information) is associated with type and magnitude of referent crit as well as the level of constraint recognition. However, in the current model of communicant activeness, predicting direction of information specificity (i.e., selectivi 54 expressed either counter to or consistent with current belief) matters less than predicting the magnitude of information se lectivity. Regardless of the direction of information selectiv tiveness to e g in preference or avoidance is conceptually identica g , rs & ity, we can impose a single conceptual dimension of the extent of selec the direction of information selectivity. In other words, I conceptualize reinforcing selectivity (i.e., what dissonanc theory predicts) and revising selectivity as being functionally equivalent in reducin cognitive discrepancy and the amount of information. Communicants at times prefer supporting information while avoiding nonsupporting information; at other times they prefer nonsupporting while avoiding supporting information corresponding to their choice of problem solving approach. What then is common around such contradictory tendencies is the pursuit of specificity of certain information over other information, which depends on the communicant?s own meaning and definition of relevance for the given task?problem solving. The direction of content material can be attitude-consistent or attitude- inconsistent, but the magnitude of specificity l. Just as pro-life and pro-choice activists can fall into a single variable of activeness or partisanship, the degree of pursuing specificity in taking and givin information can merge the duality of contents (e.g., seeking supporting vs. seeking nonsupporting information) into a singular concept?i.e., information selectivity. Hence we can reinterpret the contradictory tendencies as a common way such that the more active the communicant is the more one becomes selective or specific in consuming and sharing information with others. In other words, there is no reason to view selectivity as patently associated with ?a general preference for supportive information? only (Sea 55 Freedman, 1967, p. 212). Rather, communicants develop a need for pursuing specifi in whatever way they perceive to be subjectively relevant, either to reduce cognitive discrepancy or the amount of information. After divorcing direction of selectivity from magnitude of selectivity, we can develop a more general concept of communicant selectiveness in dealing with information. This is a good way to solve the city paradox: ?How can it be that people are in fact sel or uch ategy. Once a communicant finds he or she has an important problem but doe g ective, yet display no trace of a general preference for supportive information?? (Sears & Freeman, 1967, p. 212). When we stripped the direction of selectivity from the presence of selectivity, the paradoxical two conclusions (i.e., contradictory empirical findings regarding information selectivity) noted by Sears and Freedman (1967) can be reconciled. If a communicant decides to revise a prior expectation, he or she would ignore information consistent with beliefs or pursue information countering beliefs because it is more relevant to the situation. If a communicant decides to reinforce a prior expectation, he or she would pursue specific information reaffirming current belief because it is relevant to meeting the goal of problem resolution. Because such a need f specificity arises situationally, one crucial question is: Under which conditions does s selectivity in communicant behavior arise. Selection str s not have adequate knowledge internally, he or she starts an external search to find a workable solution to the problem. In general, at earlier phases of problem-solvin efforts, communicants adopt an adding strategy. By an adding strategy, communicants attempt to collect any information that is perceived to be related to the problem. By using an inclusive strategy that selects information indiscriminately, communicants can 56 generate copious amounts of potential information in a relatively short period o terms of goal concepts, this suggests that in earlier phases, the most important goal recognizing a serious problem is to build a solution. However, when a person finds a shortage of relevant knowledge, it creates another problem of problem solving. This meta-problem thus creates a subordinate goal of gathering information applicable to finding a solution. f time. In after ill al ter o pass ., s natural consequence of earlier covetousness) to the primary To provide enough information grist for the cognitive mill, communicants w forage for as much potentially useful information as possible. Again, in terms of go concepts, this task is a meta-goal or instrumental goal that facilitates the achievement of the superior goal?problem solving. However, at some point, problem solvers encoun another kind of meta-problem. By using an inclusive strategy, communicants tend t through a threshold of adequate would-be information. Problem solvers find it increasingly difficult to manage the information pool they collect. It then taxes the problem solver?s cognitive capabilities enough to threaten the investment of cognitive resources to the primary goal?problem solving. Hence, at a later phase, communicants begin to adopt a removing strategy that takes information only when it is relevant?i.e an only if strategy. To solve a new meta-problem of information inflation, one starts to reduce cognitive complexity caused by earlier covetousness in acquiring information. To illustrate with goals concepts, two situational goals begin to compete in recruiting necessary cognitive resources for meeting goals at a later phase of problem solving. However, the meta-goal or instrumental goal must yield its demands for cognitive resources (which i 57 situatio ?s by y on itive sidiary to the mmunicants? minds is that iscrim A s a nal goal, problem resolution. Hence, as a compromise, a communicant shifts one informational behavioral strategy from increasing complexity of the cognitive pool? increasing quantity of available information? to reduction of cognitive complexity b being miserly and stringent in selecting information to be added. Communicants thus become more aesthetic and selective?connoisseur-like?as they pick up informati only if it meets certain criteria. In other words, communicants now seek cogn competence in problem solving rather than cognitive complexity that is sub primary goal of a situation. The rule that emerges in co d inating information should change from vaguely alluding to specifically referring, from randomly encountered to systematically located, and from remotely related to closely relevant. In summary, problem solvers pursue available information at the earlier phase, whereas they pursue applicable information at the later phase of problem solving. judging criterion evolves during a problematic situation. It in turn guides subsequent information acquisition and transmission toward resolution of the problem. This explain why information selection occurs in the continuum toward higher communicative activity?i.e., the higher communicant activeness, the greater selectivity in communicative actions (information acquisition and transmission). Information Permitting and Information Forefending An active communicant exhibits selectivity not only through selective intake and outtake of information but also by selectively interpreting and producing information. Specifically, information selection will be observable in selective consumption, production (or interpretation), and propagation of certain information that fits well into 58 referent criterion a communicant holds for a problematic situation. I break down information selection conceptually into two distinct states of communicative activene The higher state of information selection is information forefending, the lower state of information selection is information permitting. Forefending information means advance or prior warding off or resisting information. These are parallel to the proactive (more active) and reactive (less active) information acquisition (information seeking and processing) and information transmission (information forwarding and sharing). In taking and giving information pertaining to a problem, communicants te forefend the scope of information as the level of communicant activeness rises. I d ss. nd to efine informa s to ught es tion forefending as an active communicative feature of information selection. A a problematic situation continues, some forethought or discriminatory referent criterion emerges whereby communicants solve problematic situations by reducing cognitive discrepancy or information. In other words, the more active a communicant within a problematic situation, the more one develops some forethought in evaluation of data to construct a solution (i.e., external data?information) and in evaluation of knowledge give out to other communicants (i.e., internal knowledge?information). The foretho or discriminatory referent criterion helps communicants organize the search process for related information and to sort out the irrelevant from the relevant. It not only provid some preliminary familiarity with data that will be evaluated but also provides an organizing frame wherein a wide but related range of information can be hosted. 5 5 A need for forethought arises from information abundance?i.e., active communicants tend to have more information sources and thus more cognitive knowledge available. An active communicant?s decision and solution building derives from the ?composite of all information? they learn (J. Grunig, 1997, p. 25). To make a decision and to compose a solution, thus, an active communicant develops a strong need for an effective reduction mechanism. Some evaluative lens or referent criteria evolve therefore. 59 Whether a communicant applies a discriminatory criterion in judging the value and utility of available information depends on his or her preconscious and conscious cogniti for an ) . and the fer ve processes. The process could be preconscious if one has brought a strong decision rule, or guiding principle, carried from prior situations, or a strong desire end state (e.g., parent?s willful thinking to cure one?s ill child). In such instances, he or she will use the decision rule to evaluate the relevance of given information without consciously evaluating. At other times, communicants would not have such a cross- situational decision rule or guiding principle?encountering a problem requiring a novel solution. As a result, one has to deal with more information and thus find a need to reduce it to expedite the process of problem solving. Hence, as the problematic situation continues, communicants tend to better train themselves in evaluating the utility of information by developing a set of criteria (e.g., reliable sources, topics, or content areas to distinguish and sort out the influx of candidate information In contrast, communicants tend to not discriminate information if they are not active regarding the problem. They do not commit to any specific judging criteria because of problem irrecognition. Such communicants behave in an ill-guided and ill- focused manner in taking and giving information. In dealing with information, they are random, general, and pursuing of mere relation; they are not specific, systematic, pursuing of relevance. In the absence of information judging criteria, the less active communicant is, the more one is lenient in evaluating information so as to consider merely related information to be potentially useful. Thus, less active communicants of little value to an information provider because they are painfully nonchalant, their minds 60 are crowded by too much irrelevant information, or they are incompetent in sorting out applicable information from that which is simply available. Because I postulate that the heightened perception of a problematic situatio higher problem recognition and a high percepti n (e.g., on of involvement to a problem) increases of being problematic would a newly compos motiva he person to be permissive, which is indifferenc oid being a side. Th urce, nformation Selection, Transmission, and Acquisition: blem Solving the activeness of situational communication, a weak perception elicit a lower need for revising/reforming a current solution or lower confidence in emerging solution. For the former case, the person feels less provoked to e a new solution, meaning that a less active communicant has little internal tion to think about a problem. This leads t e: e.g., ?I don?t care whatever it says.? For the latter, the person can av explicit because of the absence of any committable solution, thereby not taking is leads him or her to be permissive, which is diffidence. Regardless of the so that is, indifference or diffidence, I here postulate that people under lower magnitudes in situational-perceptual parameters will be more permissive in their communication behavior, whereas those higher in situational-perceptual parameters will be more forefending in their communicative behavior. In sum, regarding communicant selectivity I postulate: The greater the communicant activeness, the more the communicative behavior is systematic, specific, and pursuing of relevance. In contrast, the lesser the communicant activeness, the more the communicative behavior is random, general, and pursuing of mere relation. Integration of I A Catchall Conceptual Model of Communicant Activeness in Pro 61 Communicant activeness in problem solving is the central concept that bring together three communicative behavioral dimensions: information selection, transmission, and acquisition. The construct, communicant activeness, here becomes the integrative glue that binds together six variables about information behaviors. The major premise is that the more a communicant becomes active for a problem resolution, the greater the communicant becomes selective, emissive, and acquisitive for information related to the problem. Specifically, as communicant activeness increases, a communicant does more information forefending, forwarding, and seeking. As communicant activeness decreases, a communicant does more information permitting, sharing, and processing. Figure 4 and 5 summarize the major premise of communicant activeness and three dimensions of informational behavior. s. Figure 4: Three-dimensional representation communicant activenes 62 Figure 5: A model of communicant activeness in problem solving. I define six subdimensions of informational behavior that are correlated with each other. In factor analytic terms, the six dimensions of communicant activeness are correlated with the construct of communicant activeness. Information permitting, sharing, and processing indicate reactiveness in communicant activeness, whereas information seeking, forefending, and forwarding indicate proactiveness in communicant activeness. These dimensions conceptually tap different levels of communicative activeness (e.g., liking vs. absence of liking). Therefore, the greater the communicant activeness, the more likely are reactiveness and proactiveness indicators to both be strongly present. In contrast, the less communicant activeness, the more likely it is that only reactive indicators are present. H1: The higher the communicant activeness in problem solving, the higher the H2: The higher the communicant activeness in problem solving, the higher the H3: The higher the communicant activeness in problem solving, the higher the information forefending. information permitting. information forwarding. 63 H4: The higher the communicant activeness in problem solving, the higher the H5: The higher the communicant activeness i information sharing. n problem solving, the higher the information seeking. information processing. Two Phases of Situational Communicant Activeness: Inquiring and Effectuating Phases I sequentially break down communicant behavior during a problematic situation into an inquiring phase and effectuating phase. By an inquiring phase, I refer to communicant?s activeness that is salient in composing?i.e., exploring, investigating, and delving into?a solution. By an effectuating phase, I refer to a communicant?s activeness in carrying into effect a solution that is to be selected. As a communicant?s activeness continues to solve a problem, the communicant shifts his or her focus from information acquisition to information transmission and selection. At the early stage of the on H6: The higher the communicant activeness in problem solving, the higher the communicative behavior in a problematic situation, problem solvers invest their communicative resources primarily to extend the pool of usable knowledge and information. Sooner or later, problem solvers reach a level of subjective confidence in quantity and quality of information and knowledge they have gathered. Once saturated with information, a communicant feels fatigue in increasing his or her information inventory. Hence, one?s wealth of relevant information toward a given situational problem bolsters selectivity in managing?processing, producing, and propagating? information during a problematic situation. Thus, a subjective conception of informati saturation leads one to the effectuating phase?investing communicative resources in applying and carrying out a solution. This is a turning point from the inquiring phase to the effectuating phase. To illustrate the two phases, I offer the sequence model of communicant activeness shown in Figure 6. 64 Figure 6: Sequence model of communicant activeness. Once we detect a problem, we seek a solution internally. Human beings have a general tendency to record and transfer their successful experience of problem solving? i.e., knowledge?to other similar problems. Thus, one?s own memory is the first stop when one recognizes a problem. If one?s effort for knowledge activation results in some solution that is available, applicable, and having judged usability (Higgins, 1996), she or he can immediately move into the effectuating stage?carrying into effect the available, applicable, and usable solution. I call such an initial retrieval effort for knowledge and information recollection from past situations the internal inquiring stage. If the initial internal retrieval efforts are not successful, one turns to external sources. External sources could be any one in one?s communicant networks as well as any medium or database that could provide knowledge and information related to a given problem. I name such external collection stage the external inquiring stage. Should one?s inquiring efforts be successful, the communicant transfers oneself from information consumer to information supplier. That is, the more one is active in 65 taking information about something, the greater the communicant is likely to give information as a result because he or she is now likely to have certain ?solutions? of one?s own (i.e., a new revised referent criterion as the outcome). As noted earlier, information transmission contributes to problem solving in many ways (e.g., resource mobilization Notably, such a new perspective (consisting of a new revised referent criterion or solu for the problem) is likely to drive information seeking and forwarding in a more selectiv way (i.e., as consistent with his or her new revised solution/referent criterion). Figure 6 captures this sequential shift of focus by its differential curves. As seen in the figu information acquisition increases first and information selection and information transmission follow next. To summarize, in the inquiring phase, we can observe more information permitting in processing and seeking possible solutions. The transition from the inqu phase to the effectuating phase is demarcated by the perceived information saturation the communicant who has little confidence that th ). tion e re, iring of e additional collection of information will inc e., t rease his or her ability to deal with the problem. However, this threshold point can be hastily reached?i.e., shortened?by external pressure (e.g., if immediate action is necessary). In the effectuating phase, we can observe more information forefending in processing, seeking, sharing, and forwarding. To do something about a problem (i. reduce a negative consequence from the given problem), one now should endorse a specific solution. That is, the problem solver finalizes a certain inferential rule, solution, or treatment that has developed and emerged from the inquiring phase. Such endorsemen and finalizing of a proposal for a certain solution (at least temporarily until one 66 encounters a new problematic situation in which the chosen solution is not effective) transitory process of wedging (i.e., decisive and committing to one conclusion) from hedging (i.e., indecisive and dividing risks by considering more than one options) (J. Grunig & Stamm, 1979). Now the selectiveness of information acquisition increases? that is, the communicant selectively seeks and forwards information that better supports, reinforces, and elaborates the endorsed solution. A wedged solution becomes an evaluative lens through which any information is interpreted subsequently. That becomes a new referent criterion (decision rule) for subsequent problems of t is a he same kind. Under nquiring), a person?s activeness in communication behavior is mainly expressed er phase II (effectuating), a person?s activeness in communication info mission a refending. Finally, the sequence m ceptua pirically distingu to differentiate between active and in the present model ly, we cou publics active i Phase I (i in terms of information acquisition and more permitting. In contrast, und behavior appears in terms of rmation trans nd more fo odel helps to con lly and em ish between active and activist publics. The easiest way activist publics is to say that activist publics have far stronger beliefs and actions. However, according to CAPS we can meaningfully and usefully distinguish the two, active and activist publics, the most important targets of communication in public relations. The newer dimensions of information transmission and selection conceptually distinguish between an active public and activist publics. Previous ld only distinguish activist publics as more active in problem solving than active . However, now we can conceptually articulate that activist publics are more n that they are more selective and more emissive. Activist publics attempt to transfuse their way of problem definition as well as a solution proposal. In contrast, 67 active publics are selective but relatively less emissive. They tend to approach problems individually rather than collectively. In short, active publics are those communica who are more effortful in information acquisition. Thus, they are in the inquiring phase in the sequential continuum. In contrast, activist publics are those communicants who are more effortful in information transmission. Hence, they are in the effectuating phase w stronger transmission and selectivity than active publics (Table 2). Table 1 Phases of Communicant Activeness and Active and Activist Publics Type of Public (Communicant) nts ith Active Public Activist Public Inquiring Yes No Phases of Acting Solving es for Problem Effectuating No Y Selectivity in the Inquiring and Effectuating Phases We tend to believe that obtaining some information obligates a communicant to believe, feel, or behave in the way the information dictates (Thayer, 1987). However, exposure to some information does not always elicit informed behavior consistent with the information provider?s belief and wish. Case (2001) explained it well: Not on their he ly are people told that taking drugs and smoking are ultimately bad for alth, they can observe this fact in the world around them; those observations, however, often do not result in less consumption of harmful substances. As Sears and Freedman pointed out, failure to act on information is often due less to selective exposure than to a rejection of information with which we disagree: Perhaps resistance to influence is accomplished most often and most successfully at the level of information evaluation, rather than at the level of seeking and avoiding of information. (pp. 93-94). I explain such inconsistency in communication effect as a difference between what information content suggests and what the information acquirer would interpret or how he or she would behave by the communicant?s selectivity?specifically selective 68 evaluation or interpretation. When a problem holder concludes that a solution, regardless of its choice from merits or from effects, she or he is more likely to communicate with clarity (i.e., specific information) than with confusion. A communicant with a chosen problem solving method tends to assign cognitive and communicative resources toward effectuating rather than inquiring about a solution. He or she shifts problem solving efforts from constructing a workable solution to enacting it. Hence, we often find that communicative selectivity is more salient in the dimension of information transmission: unicants who are more forwarding are also more forefending. Communicants attemp n om others. To effectu is ce on doctor given information as consistent with a personal referent criterion (e.g., the patient would interpret his doctor?s serious warning as indicative of a professional clich?.). For example, i.e., comm t to effectuate?forward a problem paired with their preferred solution?by forming a collective around the problem. This happens because communicants can share the cost of problem resolution with other problem holders. In addition, communicants ca facilitate the problem solving process by mobilizing more resources fr ate a solution is in essence to produce and propagate a set of information so as to mobilize other problem solvers? potential resources by information selection and transmission. Relative to salience of communicative selectivity in information transmission, communicant selectivity is less noticeable in information acquisition. A primary reason because information acquirers have little control over providers of content materials. During the information acquisition period, information solicitors have little influen information contents (e.g., a patient cannot hear what he wants to hear from his about his illness.). Instead, information acquirers can control the way of interpreting the 69 audiences have little power when they are watching TV programs even if some contents are distasteful. Thus, active liberal voters and active conservative voters are not differen in the way and extent of information acquisition?e.g., diligently watching the same TV debate. However, they often express their views differently in their interpretations drawing starkly different conclusions about who won a presidential debate. Summary Information selection refers to one?s selectiveness in dealing with?consuming producing, and propagating?information. Specifically, I conceptualize that information acquisition will be more select t ?e.g., , ive during seeking than processing, whereas information transmi ation se ant activeness in problem solving (CAPS) predicts that heightened motivation will lea n) ssion will be more selective during forwarding than sharing. 6 The more a communicant becomes active, the more one would become selective in his or her communicative behaviors. In addition, I postulate that selectiveness should be stronger in information giving than information taking. Next, Information transmission refers to the extent of imparting information about a problem one perceives. At some point of information consumption, one is likely to develop a good deal of inform inventory about a problem. This in turn motivates a communicant to have a greater sen of informative connoisseurship and opinion strength. A communicant?s endeavor increases self-confidence and commitment to the solution one has developed. Communic d a person to transmit information (e.g., about a problem with a loaded solutio willingly and voluntarily to others. Information transmission can be a functional act to 6 Information sharing and forwarding are used here in a slightly different way from common usages of th words. In brief, forwarding is proactive communicant behavior (e.g., voluntary and willful transmission a e nd transfusion to an intended segment of people) regarding a certain problem, whereas sharing is reactive communicant behavior (e.g., simply being responsive when solicited). 70 resolve a problem directly (e.g., mobilization of resources). Or it can be a humanitarian act to reduce negative consequences among neighboring communicants. Finally, information acquisition refers to extent of one?s desire to accumulate information re to a problematic state. A communicant is more likely to satisfy his or her information appetite corresponding to the level of his or her epistemic motivation. The situational theory of publics has set this aspect of communicative behavior as default endogen variables to be accounted for. CAPS postulates two phases of situational communicative activeness as the inquiring phase and effectuating phase. The inquiring phase is highlighted by informa acquisition to carve out a workable solution(s) as its feat of communicative action. Yet, information transmission and information selection are not salient at this period. How such low salience of information tran lated ous tion ever, smission and selection as communicant behavior in the inq out the situ The commun is, a period in which he or she translates preferred beliefs, opinion, or solution into action. At this phase, one experiments with a chosen solution or belief (e.g., an opinion) to the problem and becomes an exponent, although not permanently, to the epistemic conclusion resulting from a prior inquiring phase. One can still amend a favored solution even when it is put in force. However, consuming information for amendment in the effectuating phase is not uiring phase becomes dominant features at the effectuating phase. Early indulgence to relevant information will elate the communicant enough to apply one?s alleged solution to the problem. A certain level of self-confidence from the knowledge accumulation ends one?s need for more information before doing something ab ation. icant now enters into the effectuating phase, that 71 likely t st or the role o be permissive or indiscriminant as in the inquiring phase. Now, one has a preconceived view. Communicants are now prepossessed by a chosen solution and are likely to forefend?i.e., to be selective?in acquiring information until experimentation provides irresistible evidence of the malfunction of the favored solution?another problem recognition. At the effectuating phase, increasing the communicant?s selectiveness is more rewarding to him or her than a balancing neutral effort, in contra to in the exploratory, inquiring stage. It would be easier to find a flaw in a composed solution by being partisan to it, 7 by recruiting helping hands and resources to work out the solution, and magnifying ?acting-out? energy in translating a composed solution within one?s situational constraints. Next, I turn to another new variable, cognitive entrepreneurship in problem solving. With it, we can observe that problem solvers use different mental approaches following different problem perceptions across situations. I offer a model to describe those differential mindsets in problem solving in next section of this chapter. A MODEL OF COGNITIVE ENTREPRENEURSHIP IN PROBLEM SOLVING Life is lived forward but understood backward. S?ren Kierkegaard We believe our will or intention precedes an act. We would be bewildered if someone told us that ?we first did a certain action and next understood what we did.? Why is such a claim so offensive to us? Perhaps it is due to our deep respect f 7 In the effectuating stage, we tend to develop more contents for conjecturing (to be tested) carrying from the inqui ind must con res selectivity. Molding a proper and durable solution to a problem requires being painstakingly persistent in repeating such an ostentatious experimentation until reaching an irrefutable state. ring stage. As a consequence we are likely to practice selective communication. The human m jecture to know the utility or veracity of an idea. This conjecturing procedure inherently requi our minds to commit or feign commitment to an idea as though useful and veracious at least momentarily, even though not assured and unknown. Such a commitment, although momentary or pretended, is 72 of our will in an act. We dwell on the thought that we are the master of our own life and that we take control our own actions. We live so deeply in a rational decision-making tradition that it seems that every important decision we make must be intended by us beforehand. The decision-making process flows unilaterally from a deliberative pro concerning what we will do to a subsequent action (decision), not vice versa. Altho could be extremely short or even unconscious, any given conclusion of judgment ought to be preceded by some degree of will or intention. However, in this section I will theorize our mental process with a counterintuitive assumption that our action or judgmental conclusion made during a problematic situation can precede our intention, volition, or rationale of a given act or choice. Unlike the common concept cess ugh it ion about the unilateral flow of cognitive efforts to decision making, I concep a 4). choice [italics added],? little research effort has been made to study the phenomena of tualize a bilateral sequence between cognitive efforts and decision making (e.g., a decision precedes cognitive efforts about the decision). Even further, I contend not only that intention at times has no place in our cognitive working process, but also that such counterintuitive sequence from an action to cognitive working (e.g., intention) is our default cognitive approach, which we take routinely. In what follows, I will discuss how the sequence of cognitive efforts and decision making can often be reversed. Inferential Order in Problem Solving I conceptualize two directional flows between decision-making and cognitive working in problematic situations as forward reasoning and backward reasoning. Brehm (1956) once raised the issue of understanding ?what happens after the choice? (p. 38 Although much research has been done regarding ?the phenomena that lead up to the 73 reversal (Brehm, 1956, p. 384). More recently, Frey (1986) in his classic review selective exposure to info of rmation summarized as ?seeking out of decision relevant informa g how he esearchers seem to take serious r the oning. For this reason, both reasoning strategies play a functionally equivalent role in the mind. A person who suffers from an tion does not cease once a decision is made. Rather, this search continues durin a postdecisional period during which the person confronts and weighs the various decision alternatives and their respective advantages and disadvantages [italics added]? (pp. 41-42). As Brehm (1956) and Frey (1986) said, research on the phenomena of problem solvers are cognitively working before and after a choice made is a significant area of inquiry. Underlying the common research focus on predecisional cognitive efforts is t normative belief that people should behave in a rational way. R ly the wisdom that ?there is no use crying over spilled milk??i.e., little can be done after making a decision. However, regardless of such normative influences on theorizing about choice situations and cognitive working, we often observe that we are ?crying over spilled milk.? People make cognitive efforts after making a choice. Such postdecisional mental elaboration has no effect on the given choice, especially when a problem solver enacts a chosen solution for the problematic situation. Putting aside the issue of how we can make a better normative theory, in the present section of this chapte I will conceptualize both approaches of predecisional and postdecisional cognitive working. In other words, I build a descriptive theory?i.e., sketch a process?about illative orders between a problem solver?s cognitive labor and the drawing of a judgmental conclusion. Human beings are pragmatic in their reas 74 infestat or her cats as able ces? urces and efforts in relation ognitive effort to construct, define, t among possible solutions with re is the last step after using up mos e s, we e the ental ion of mice at home does not discriminate between the colors of his long as they reduce the number of mice. In a similar sense, the directionality of the reasoning process does not matter to problem solvers as long as it generates a work solution. However, to devise a way to improve problem solving in general, we need to know under what conditions one adopts which reasoning strategy and how well the chosen cognitive strategy supports effective problem solving. What is the major distinction between backward and forward reasoning sequen I answer that it is how a problem solver uses his or her cognitive reso to a conclusion. In the forward strategy, one invests c and compare solutions as broadly as possible, and selec gard to their merits. Thus, one?s selection of a solution t of one?s available cognitive resources. In contrast, in the backward strategy, one invests cognitive effort to construct, define, and select a best justification for an already chosen conclusion. Thus, one?s selection occurs before using up most of one?s cognitiv resources. In other words, a backward reasoner invests most cognitive resources to reinforce an a priori conclusion. To better understand these two reasoning strategie need to understand how we make decisions during problematic situations. A Syllogistic Illustration of Cognitive Working Before I conceptually elaborate the focal construct of the cognitive entrepreneurial mindset in problem solving, I will take some time here to describ m process of cognitively working toward a situational conclusion in problem solving?i.e., how we perform cognitive tasks during a problematic situation. Here, a syllogistic reasoning process is a useful frame within which to explain the human 75 judgmental process (Kruglanski & Thompson, 1999). In brief, a syllogism is a deducti argument consisting of two premises and one conclusion (Hurley, 1997). It takes the form of major premise?minor premise?conclusion. For example: No painters are sculptors. [Major Premise | Evidence] Some sculptors are artists. [Minor Premise | Evidence] Therefore, some artists are not painters. [Conclusion ve ] in a at contains the major term: e.g., ?No painters are sculptors.? The minor p s are ans, syllogism known fact, (Evans, 2002). For example, we stop our car when we see a red light?e.g., if I see a red Depending on their positions in the argument, we distinguish three terms with syllogism. The major term is the predicate of the conclusion (i.e., painters); the minor term is the subject of the conclusion (i.e., artists); and the middle term, which becomes the conceptual bridge between the two premises (i.e., sculptors), is the one that occurs once in each premise and does not occur in the conclusion. The major premise, by definition, is the one th remise is the one that contains the minor term: e.g., ?Some sculptors are artists.? The conclusion is the derived argumentative result from the combination of major and minor premises: e.g., ?Therefore, some artists are not painters? (Hurley, 1997). This formal categorical syllogism provides a baseline to discuss any routinely drawn human judgmental conclusion. However, our everyday reasoning processe more pragmatic and probabilistic than such a rigid framework of logical steps (Ev 2002). Lay thinkers often draw judgmental conclusions using a more basic as the if?then?rule. Lay people who are not trained in formal logic do, in exhibit a rudimentary deductive competence when confronted with judgmental tasks 76 light, then I should stop my car. This does not require us to set up a strictly formal categorical syllogism argument to reason a proper action. For another example, we may routinely use incorrect rudimentary syllogistic reasoning when stereotyping others: e.g., if a person is an Asian, then she or he must be good at mathematics. One may see an Asian student in a math class and predict that or she must do well on exams. We can almost always restate such basic pragmatic a probabilogistic syllogistic reasoning examples into more formal and categorical syllogism arguments. Regardless o he nd f formality or logicality, however, lay thinkers conduct judgmental process en) 8 judgmental conclusion. We not es via a more implicit and simpler syllogistic reasoning process (i.e. if?th (Evans, 2002; Kruglanski & Thompson, 1999; Over & Evans, 1997; Evans & Over, 1996). Therefore, I assume that the human reasoning process can be best illustrated by a pragmatic and probabilogistic syllogism in conceptualizing the focal construct of cognitive entrepreneurship in problem solving. From now on, I will use the term syllogistic reasoning to denote the method that lay thinkers routinely use. Directionality between evidence and conclusion. As mentioned, we make decisions through a simple and pragmatic process of syllogistic reasoning (i.e., if? then?). Within the syllogistic reasoning frame, people recollect, collect, or elaborate information to deploy it as supporting evidence toward a given only use the rules that we are carrying from prior situations by a form of the if? then rule (e.g., [if] children watch violent movies, [then] they behave aggressively); but 8 We do inferential tasks by implicit, not explicit, application of the if?then?syllogism. We ma explicitly use the words, ?if?? and/or ?then?? in performing an inferential tasks. Simple assoc y not iation linkages between ?terms? are often sufficient to making inferences: e.g., ?red light?stop.? However, we can almost always reconstruct the syllogistic reasoning process by using the if?then?format. 77 we also perform current inferential tasks by a deductive processing of an if?then? syllogism (e.g., [if] I saw a very aggressive kid, [then] he must have watched many violent movies in the past). Here the extent of association (the strength of connection) between the if?component and the then?component is called ?relevance? (Kruglans Thompson, 1998). ki & ., [if] id the ?the ch on ns d We confer a certain amount of relevance to the inferential association in corresponding with the relevance we can draw from the decisional referent rule (e.g children watch violent movies, [then] they behave aggressively). That is, the confidence that we have in our judgmental conclusion (e.g., confidence about ?the aggressive k who had watched many violent movies?) is commensurate with the strength of the associative link between if?then?rule we use as a decisional referent frame (e.g., extent of one?s belief that ?watching violent movies causes aggressive behaviors for children?). For example, consider that you observe a very aggressive act of a child. Subsequently, you might take that act as evidence to draw a judgmental conclusion that ild must have watched many violent movies/games.? Evidence intuitively precedes a conclusion. However, initiation and completi of the judgmental process between evidence and a conclusion in problematic situatio can occur in any direction. One may start from a conclusion and proceed to seeking evidence. Or, one may start from seeking evidence to proceed to a conclusion. Sometimes people benefit, consciously or unconsciously, by following the forwar direction (i.e., evidence dictates a certain conclusion). For example, one might think that ?if someone is a Harvard graduate and working in top management for a large business, 78 then she must be smart.? At other times, however, people find merit, conscio unconsciously, usly or in reverse order reasoning (i.e., a conclusion dictates certain evidence). ental ide ungry; he ords, a conclusion directs the individual to seek certain evidence that justifies the hastily drawn conclusion. It is important to understand that the drawing of a conclusion does not exclude active cognitive working or elaboration in a retroactive way. Even though we make a decision, we might feel it is necessary to elaborate our chosen itive fore, lusion and connecting evidence to it is bidirectional. 9 One draws a conclusion first by applying a salient rule?i.e., a prime decisional referent?and next collects evidential information that warrants the predetermined conclusion. For example, a person with a terminal illness might draw a quick judgm conclusion such as ?I am OK? and collect evidential information that indicates and reinforces his physical well-being. Or, a group of political leaders might quickly dec to go to war for a salient reason (e.g., the political regime of that country has been uncooperative with us) and next seek out additional supporting decisional referents and information (e.g., the leader of the country is a dictator; he made his people h have made weapons of mass destruction; he provided support for terrorist groups etc.). In the example, the conclusion (e.g., we are going to war) precedes substantial evidence (e.g., the whys for war) that warrants and justifies the conclusion. In other w conclusion. We conventionally assume that a drawn conclusion completes our cogn working process. However, quite often we go backward in problem solving. There the direction between drawing a conc 9 Although we almost always describe the judgmental process as a forward and unidirectional manner (e.g., I decide so for such reasons), we often go backward but do not explain so because of our reconstruction of a judgmental process to report to others. 79 Forward reasoning vs. backward reasoning. Now I will elaborate two directional flows of syllogistic reasoning. Assuming an equal amount of cognitive resources and motivation in solving a problem, a problem solver can take two contrasting mental approac ng ng frame, he process of a forward reasoning approach, I offer the following: be selected as a best decision. that merit and favor this conclusion among the others. T refo warrants this specific conclusion. In contrast, I define backward reasoning as a cognitive approach when a conclusion directs evidence. The following is an illustration of a backward reasoning approach: I selected option A for an important reason. evidence) for the option A (possible evidences) would be a, b, c, d. conclusion. a good decision. Here, th le and hes in expending cognitive capacity and capability. One is the forward reasoni and the other is the backward reasoning. First, forward reasoning is the commonly conceived way of problem solving. In terms of the syllogistic if?then?reasoni I define forward reasoning as a cognitive approach when evidence directs a solution. To illustrate t If information a, b, c, d (i.e., evidence) tells this, then option A (i.e., conclusion) should I found some antecedent conditions (i.e., evidence) he re, I choose this course of action (or a solution) because preceding evidence (If) I selected option A (i.e., conclusion), (then) the acceptable justifications 10 (i.e., I found some antecedent conditions (i.e., evidence) which fit well with the chosen Therefore, I made e thinker quickly reached a judgmental conclusion by a prime decisional ru then sought out rationales that make the selected option more conclusive and convincing. This is an optimization process for an a priori conclusion. 10 Justifications can be possible consequences (merits or harms) associated with a chosen option. Such a case is forward reasoning rather than backward reasoning. 80 Notably, in both cognitive approaches, a chosen solution for a problem should first contain the observational contents that best fit the major premise chosen within a syllogism model. Then, the chosen solution will produce a level of confidence commensurable to the degree of relevance between ?if? and ?then? contents in the major premise. Figure 7 summarizes the two distinct cognitive reasoning approaches in problem solving. Figure 7: Cognitive strategies in problem solvi ng. define (Krugla ce ) came a Certitude of a given conclusion. By using the syllogistic reasoning frame, we can attitude as a judgmental conclusion drawn about a certain social object or issue nski, 1989). The attitude?an evaluative judgmental conclusion?might be supported by evidential materials. However, the certainty one can draw from eviden toward a conclusion is not determined by the frequency or amount of information connected, but by the subjective ?relevance? of prior belief or decisional referent rule(s in making the given judgment (usually, as another form of an if?then rule that be major premise.) 81 A person under pressure to make a quick judgment would draw a referent criterion (in a form of an if?then?rule that becomes a major premise) that is available hed to s . nt to which the evidence is similar is commensurate with the degree of certitude that wi blic on issues, however, have found some consistency in the cognitive strategies es. Grunig and Ipes (1983) also found that active publics have more organized d Ipes (1983), showed that passive publics are more likely to hold attitudes than Less active publics express attitudes even when they have no cognitions on which and applicable to the given problem. Next, she or he seeks out some analogous evidential material from the current situation (via observation). When new evidence collected is similar to the evidence in the activated referent criterion toward the conclusion, he or she then confers the given certainty (relevance in the major premise or referent criterion) in the old premise to the newly drawn conclusion, which is tantamount certitude attac a fit (relevance) between the old if-then rule (the referent criterion one is now deploying). In other words, when a person is under pressure to make a quick decision, he or she look for evidence similar to that which supports his known experience or referent criterion The exte ll be associated with the new decision. This can solve a puzzle that many public opinion researchers encounter. Pu opinion researchers often have found that people who express a strong attitude about something would lack cognitive knowledge that supports a given evaluative conclusi toward the attitudinal target. J. Grunig and Hon (1988) reported and summarized such affective publics without cognitive counterparts on attitudinal objects: Several studies of publics arising from environmental issues and corporate policy constructed by members of active publics and in the nature of their attitud cognitions than do passive publics. Two studies, Grunig (1982a) and Grunig an cognitions. Active publics are equally likely to hold both attitudes and cognitions. to base them. (italics added, J. Grunig, & Hon, 1988, pp. 5-6). 82 Combining this judgmental process with the directionality of initiation and comple nd ve an internal preference or directional expectation about t terion or se intentio arge ates tion of a judgmental task, an explanation of why passive publics often have unreasonably strong attitudes (conclusions) in the absence of cognitions (evidence). As most dual models of social influence (e.g., the Elaboration Likelihood Model, Petty & Cacioppo, 1986) suggest, under some conditions people with limited judgmental motivation and capability become cognitive economizers. People draw a quick conclusion using an activated previous judgmental rule (referent criterion or schema) a match easily identifiable evidential materials. However, when individuals ha he outcome that the decision might produce, backward reasoning is more likely to happen because the preferred outcome exerts influence in selecting a referent cri prior rule. This is the way wishful thinking happens and why many decisions that a lay person makes are unrealistically biased. People adopt a referent criterion that best warrants the preferred end state regardless of its actual likelihood. This happens becau a preferred outcome state powerfully influences an individual to activate a certain prior judgmental rule that more successfully warrants the preferred outcome state among others. Parallel syllogistic reasoning processes. At the same time, drawing a conclusion first is not necessarily limited to a single conclusion. It is possible for a person to nally (and often thoughtfully) select multiple, conflicting conclusions. Problem solvers may want to be scrupulous or wish to reduce possible errors and risks in the judgmental task. The forward reasoning strategy requires considering a relatively l number of alternative courses of action (i.e., the larger number of solution candid 83 reviewed). In contrast, the backward reasoning strategy would consider relatively fewer alternative courses because of the readymade solution or because of strong prior motivation that leads one to a specific course of action. Although backward reasoning problem solvers can be cognitively hard working enough to construct multiple syllogisms, problem ). A awn, ses nd to To be selected, a solution proposal should us solvers with a forward reasoning approach are more likely to construct and go through a more scrupulous multiple syllogistic reasoning process. Cognitive Strategies and Behavioral Molecules In terms of the syllogistic reasoning framework, I described a backward reasoning process (i.e., a conclusion comes first and seeking information (evidence) follows strong major premise?a prime decisional referent?would compel the lay thinker to draw a syllogistic conclusion pertaining to a problem. Once a hasty conclusion is dr the person looks for information that increases the fit between the observed minor premise and the preferred major premise. The enrichment provided by observational information that increases the relevance of the if?then rule of a major premise increa confidence in the given conclusion. The forward reasoning process, in contrast, suspends drawing a conclusion until reaching a certain level of subjective confidence?i.e., a feeling of information saturation?to make a better decision (i.e., seeking evidential information comes first a drawing a conclusion follows). Here judgmental rules and proposed solutions compete demonstrate their merits over the competing sets. demonstrate superiority by its merits. Problem solvers thus undergo the laborio iterative process of what-if thinking to examine merits and pitfalls associated with given pieces of information until one solution emerges as the best. These two cognitive 84 approaches provide a simple way to summarize the multiple differential decision-mak approaches described in J. Grunig and Hunt?s (1984) behavioral molecule model. Behavioral molecule. Drawing from Richard Carter?s (1973) behavioral molecule, J. Grunig and Hunt (198 ing 4) proposed a behavioral molecule that illustrates how people (e.g., an about what to do in problematic situatio of several segments that capture the processes individ rder: detect, s are describ oretically endless and if followed thoroughly will lead to hink a In this segment, he or she tries to be totally objectiv ere is . ative. Select is the step in the sequence during which one chooses the best plicable prior decision rules (i.e., referent criteria organizational manager) make decisions ns. The molecule consists uals or systems go through to plan and select behaviors. Its segments are in o construct, define, select, confirm, behave, and detect. The segments or step ed as sequential and the more successful problem solving. Detect is the segment in which a person discovers a problem and begins to t about a solution. Construct is the segment in which a person begins to formulate solution to the problem he or she detected. e and abstains from making a judgment about what to do. The major task h to be effortful in cognitive processing to define the problem, choose appropriate objectives pertaining to a problem, and formulate alternative solutions to the problem Define is the step in which a person specifies distinctly how each alternative can be implemented. The define segment ends when a single plan of action has been elaborated for each altern alternative in solving the problem. Here ap ) or one?s values or attitudes exert greater influence to favor (or eliminate) one against the others. Next, confirm is the step in which a person reviews the reasonableness of the selected solution and finalizes it before enacting it. Behave is the segment in 85 which one translates the chosen course of action (solution) into action for problem resolution. Finally, the last segment is, again, detect to evaluate whether the intended effect?problem resolution?has been achieved. J. Grunig and Hunt (1984) suggested that ideally the segments should occ sequence because that maximizes the potential to make the best behavioral decision abou a problem. However, in real life, the full sequential order might be shortened because o situational constraints. The steps of the behavioral molecule provide a useful way to describe some common mistakes in decision making. Often decision makers omit some of the segments in the behavioral molecule or change the sequence from the model. Some common mistakes are: Dogmatism (detect?select?behave--). Habit (detect?behave ur in t f Rationalization (detect?select?behave?construct--). --). Procrastination (detect?construct?construct?construct--). construct --). t? define?select?confirm?construct?define?select?confirm?construct--). (J. With d the C back exam inte cons back Indecision (detect?construct?define?select?construct?define-select? Perfectionism (detect?construct?define?select?confirm?construc Grunig & Hunt, 1984) in the previously described framework of the entrepreneurial cognitive approach an AOS terminology, dogmatism, rationalization, and habit are special cases of ward reasoning, whereas indecision, procrastination, and perfectionism are ples of a forward reasoning strategy. The sequence between information collection and decision-making can be rchanged in some cases. In the forward reasoning approach, information helps to truct and define the alternatives?i.e., a prospective use of information. In the ward reasoning approach, information is used to justify the omitted steps (i.e., 86 construct and define) and to reinforce the selected alternative?i.e., a retrospective use of information. Temporal Order between Will and Action In the present model of cognitive entrepreneurship in problem solving, I postulate that human cognitive strategy in judgmental situations is a variant, rather than a constant (e.g., an enduring personal trait). That is, the temporal order between conclusion and evidence is bidirectional across situations. The problem of discerning the temporal order between our will and an act is analogous to the problem of discerning the temporal order between our evidential reasoning and drawing a conclusion. Because of the similarity of the problem sets, I look to the past half century of research in psychophysiology regarding the problem of discerning the temporal order between intention and action to bett have g em, Libet, Gleason, Wright, and Pearl (1983) found a perplexing pattern that shattered our conventional beliefs about the order between ?will? and ?action.? He found reversal time sequence between one?s will to act and our movement preparation. That is, our subjective will for moving is preceded by brain movement preparation, so called, ?Readiness Potential (RP).? Experimental finding. Obhi and Haggard (2004) summarized the groundbreaking finding from Libet et al., (1983) study on the ?source of control? as follows: ?participants watched a small clock hand that completed one full revolution in 2.56 st at a time of his choosing. After the movement, the clock hand continued to rotate for a ck hand at the time when she first became aware of the will to move?.this subjective n er understand the problem of judgmental sequence. Many cognitive psychologists investigated the temporal order between ?intention? or ?will? and ?action.? Amon th seconds. While fixated on the clock, a participant voluntarily flexed his wri random time and then stopped. Then, a participant reported the position of the clo judgment W, for ?will.? In other parts of the experiment, participants judged whe 87 they actually moved?this judgment M, for ?movement.? The timing of the W and and actually moved. In add the M told?when?subjectively speaking?a participant formulated a will to move ition, Libet?s team measured two objective parameters: the electrical activity over the motor areas of the brain, and the electrical activity of the a well-known psychophysiological correlate of movement preparation called the electrodes placed on the scalp overlying the motor areas of the frontal lobe, and n by about 1 second. By also recording the electrical activity of the muscles involved in d to the RP. (Obhi & Haggard, 2004, pp. 358-359) Libet et al. (1983) studied the temporal order of conscious experience and neural activity by comparing the subjective W (will) and M (movement) judgments with objective RP and muscular activity. Their finding first showed that W came before M. This means that the participants in the experiment ?consciously perceive the intention to move as occurring before a conscious experience of actual moving,? which is consistent with our common conception (Obhi & Haggard, 2004, p. 359). However, Libet et al. found an intriguing temporal order that ?actual neural preparation to move (RP) preceded conscious awareness of the intention to move (W) by 300 to 500 milliseconds? (Obhi & Haggard, 2004, p. 360). Obhi and Haggard (2004) restated the meaning of this surprising finding: Put simply, the brain prepared a movement before a subject consciously decided to motor preparatory activity in the brain rather than a cause?this finding ran directly 2004, p. 360). Libet et al.?s (1983) finding, however, did not totally upset the relationship between intention and action, that is, that conscious processes such as intention cause actions. Subsequent findings suggested that ?conscious processes could still exert some effect over actions by modifying the brain processes already under way? and thus it would be muscles involved in the wrist movement. Over the motor areas, Libet recorded readiness potential (RP)?[RP] is measured using electroencephalographic recording appears as a ramplike buildup of electrical activity that precedes voluntary actio the wrist movement, Libet precisely determined the onset of muscle activity relate move! This result suggests that a person?s feeling of intention may be an effect of contrary to the classical conception of free will [italics added]. (Obhi & Haggard, 88 more accurate to call ?free won?t? rather than ?free will? [italics added] (Obhi & Haggard, 2004, p. 360). The temporal order between intentions and actions can be i.e., either from intention to action or from action to intention. According to Obhi the n e . e gnitive forward t model. As in Libet et al.?s (1983) findings, in many situations, our intention or will to bidirectional? and Haggard (2004), our brain feels the intention of an action when the prediction of movement fits well with the actual movement. Thus, when the fit takes place?e.g., the past examples of the similar action can guide current action well, person might feel a euphoric sense of control. In addition, a strong sense of intention ca script an action subsequently. In addition, a mental illness known as utilization behaviors in which ?patients uncontrollably interact with and use every object that they come across,? provides a piece of interesting evidence that the cognitive backward approach can be found in som cognitive neuroscience studies (Obhi & Haggard, 2004, p. 364). Utilization behavior patients are not aware of what they are going to do until after the action has been made In such a case, there is ?no awareness of intention before the movement;? and thus ?th patient is left to rationalize the behavior afterward? (Obhi & Haggard, 2004, p. 365). The bidirectional reasoning conception (i.e., backward and forward reasoning) in the present model explains that a person flexibly situates oneself on either a co strategy or cognitive backward strategy by one?s situational-perceptual variables. As psycho physiologists have found, at times our actions are followed by our will and vice versa at other times. Similarly, our problem solving acts (a conclusion for a judgmental task) are often done first and followed by a certain intention of why we did it. This is a case of backward reasoning that is one of the cognitive approaches in the presen 89 perform a certain action is reconstructed backwardly. Often we are asked by others (e.g., experimenters) to explain our actions. Quite often in our routine life, the intention has little use until it becomes necessary to explain our acts to others. It is often reconstructed?reasoned backward?to make sense of our action (conclusion) to ourselves and others. Very often, intention is situated within a subjective time sequence as if it ss route (heuris y n HSM or the occurred ahead of an action (e.g., asked to reflect prior action). 11 Delimitation If the main focus in theorizing about the routes that human problem resolution takes toward a decision or chosen solution, the resulting theory would only reiterate cognitive routes already described in popular social psychological theories (e.g., Heuristics Systematic Model, Chaiken & Eagly, 1989; Elaboration-Likelihood Model, Petty & Cacioppo, 1986). Typically, such theories contain either an expre tics?i.e., a decisional shortcut) or an effortful route (elaborative or systematic cognitive working) in reaching a judgmental conclusion. However, in the present theor I deliberately focus on the roles of cognitive efforts occurring before and after a judgmental conclusion (decision) is made. The previous theories implicitly assumed predecisional cognitive working or at least were mute, about postdecisional cognitive working. Because their theoretical goal was narrowly aimed to feature a typology of cognitive efforts (e.g., amount of cognitive elaboration) by the parameters of motivatio and cognitive capacity toward a decision (Kruglanski et al, 2003), those theories (e.g., the ELM) were only interested in predecisional cognitive processes. 11 If Libet et al?s (1983) finding is a robust fact, then the theory of reasoned action and the theory of planned behavior have a problem because of the behavioral intention construct. If intention is often reconstructed, the theories have a serious limitation. 90 In contrast, I am introducing a model that encompasses not just predecision cognitive efforts (how problem holders mentally invest their cognitive resources toward a given conclusion) but also postdecisional cognitive efforts (justification of a previously drawn conclusion). Consequently in the present mo al del, it is unnecessary to assume that a problem nor HS of cogn holders ? the decision even after favoring and finalizi orking t and a cognitive process within a problem es of solver ceases to make cognitive efforts once a decision is made. Neither ELM M, the two most popular theories of cognitive processing, conceives of the notion itive retrogression in their conceptualizations. Yet, we observe often that problem mentally linger on or keep ?elaborating ng solution (e.g., diligently reading about the great features of a product after purchasing it.). Although one may have decided on a solution to a problem, arriving at the solution does not necessarily indicate the end of the problematic situation. Therefore, people under problematic situations could still be cognitively active and effortful even after making a decision. For that reason, we experience that decision making is not the end of our cognitive efforts in problem solving. I have moved the theoretical scope from decision-making to problem solving. Consequently, the reversed order cognitive w (conclusion?evidence seeking) becomes another key cognitive feature. To summarize, current cognitive processing theories describe the decision process and put little theoretical emphasis on post-decisional thinking. In contrast, the presen theoretical model describes a problem-solving process atic situation. Thus, I propose a model of our mental approaches that featur the role of cognitive efforts during a problematic situation. It postulates distinct roles two different sequences in cognitive efforts. One is to reach a better solution?thereby 91 being characterized by predecisional cognitive efforts; the other is to reach subjectiv confidence in the chosen solution?being highlighted by postdecisional cognitive efforts. Conceptualizing Cognitive Entrepreneurship in Problem Solving We respect those who are entrepreneurial in their work. They endeavor to progress and make desirable changes for themselves and for others. Thus, we train ourselves and our children to maintain an entrepreneurial life. However, the term entrepreneur or entrepreneurial mindset, needs to be explicated further. J. Grunig (1968 conceived of entrepreneurs as those who drive individual and social development. He described the entrepreneurial problem solver as a: strategic decision maker who skillfully manages the resources at one?s command looking for the most efficient way of doing things. The entrepreneur is ?ratio a pre-set goal, b e , ) , which means she or he is more than a routine manager; she or he is always nal? not in the sense that he is always a profit maximizer or seeks always to maximize ut rather in the sense that he recognizes alternative solutions to his problematic situation, evaluates these alternatives, and chooses one of them. to least effort, e ha cognitive alpha strategy, wherein we find more cognitive entrepreneurial features, and the [italics added] (J. Grunig, 1968, p. 4) I conceptualize cognitive entrepreneurship in problem solving (CEPS) here as a unidimensional variable. Cognitive entrepreneurship is a human cognitive feature used cope with problematic life situations? in other words a mental approach to reduce a perceived discrepancy. It varies from an extremely entrepreneurial mindset to the entrepreneurial mindset across problems. The absence of entrepreneurial cognitive for convenience, can be called the cognitive omega strategy, whereas the strong presenc of entrepreneurial cognitive effort in dealing with a problem is called the cognitive alp strategy. To make the concept theoretically more useful, I break down the single construct, cognitive entrepreneurship in problem solving, into two dichotomies: the 92 cognitive omega strategy, wherein we find fewer cognitive entrepreneurial features. These are two conceptual faces of a single construct. this concep in erein problem solving has more cognitive features. In cognitive entrepr ) e two stinguished two decisional approaches such that: ?Most ns We typically construe entrepreneurship as a personal trait that varies across individuals and not within an individual mind. However, there is no reason for tual limitation. I conceptualize entrepreneurship here as changing across different life situations, that is, as a variable that fluctuates with changes in one?s situational conditions (e.g., situational constraints). Thus, I define cognitive entrepreneurship problem solving as a situational mindset one opts for in a problematic situation wh one?s mental effort for eneurship 1) one?s cognitive sequence progresses from evidence to a conclusion following the syllogistic reasoning process, 2) one increases cognitive effort by increasing the breadth of evidential knowledge and solution alternatives considered, 3 one is more likely to cognitively commit to ideas and alternatives under review (i.e., on average, entrepreneurs show more enthusiasm to the proposed ideas and solutions that ar allegedly viable than non-entrepreneurs do), and 4) one has more cognitive heedfulness in finalizing a proposed solution?i.e., makes another confirmatory evaluation after selecting a solution. 12 This new construct has a conceptual root in J. Grunig?s (1966) distinction of decisional approaches called habitual and genuine decisions. Following Katona (1951, 1953) and Carter (1965), J. Grunig di economic decisions?are made on the basis of habit, genuinely rational decisio are made in new situations where the decision maker has little previous decision 12 In the behavioral molecule, one would ?stop to confirm that the selected behavior will work and is the best alternative? (J. Grunig & Hunt, 1984, p. 107). This is a double checking effort for the decision-maker to assure the solution selected is the best available for the given problem before finalizing the decision. 93 experience and in important, ?crossroads? decisions? (J. Grunig, 1966, p. 93). The concept of cognitive entrepreneurship in problem solving is analogous to the ?ge decisions? (Katona, 1951, 1953) or ?crossroads decisions? (J. Grunig, 1966), when a habitual decision is unsatisfactory. To make a workable crossroads decision, a problem solver must make an extraordinary mental effort. In the next section, I will concept four subd nuine ualize imensions to further delineate cognitive entrepreneurship in a problematic ip ng and tions, whereas wedging is a one-and-only commit . situation. Conceptual Dimensions of Cognitive Entrepreneurship: Cognitive Retrogression, Cognitive Multilateralism, Cognitive Commitment, and Cognitive Suspension I theorize four conceptual subdimensions to highlight cognitive entrepreneursh under an extraordinary problem situation. They are cognitive retrogression in illation, cognitive multilateralism in considering solutions for problem solving, cognitive commitment to the identified solution proposals, and cognitive suspension before finalizing a solution. The cognitive alpha and omega strategies, which I elaborated earlier, have their conceptual origins in part from what J. Grunig and Stamm (1979) called hedgi wedging. Hedging is a cognitive strategy that reduces risk from devotion to a single option by dividing one?s commitment to the op ment to a certain option. The backward cognitive strategy shares its conceptual root with wedging, whereas the forward cognitive strategy shares its root with hedging Problem solvers who hedge expend more effort during the illative process. Thus, they tend to build and consider more solution options than those using a wedging strategy. In contrast, problem solvers using the wedging strategy take advantage of achieving a 94 cognitive certainty quickly. Thus, they abstain from increasing cognitive complexitie considering fewer soluti s by on options. ?s ore ers. I y rs ached. The referen ents in For these reasons, wedging is close to the cognitive backward strategy, while hedging is close to the cognitive forward strategy. Each strategy either satisfies one desire to be quick and firm or to be deliberate and take risk. In conceptualizing cognitive entrepreneurship in problem solving, I expand the hedging and wedging concepts to understand different cognitive strategies for problem solving. This conceptual specification of hedging and wedging will result in a construct that has m empirical content (Popper, 1999) as well as practical implications for problem solv will now elaborate upon each dimension of cognitive entrepreneurship in problem solving. Cognitive Retrogression Previously, I conceptualized two cognitive strategies in judgmental situations; a forward illation strategy and a backward illation strategy. A backward illation strateg occurs when a problem solver reaches a conclusion quickly and engages in cognitive labor primarily after drawing a conclusion. In contrast, a forward illation strategy occu when a problem solver makes more cognitive effort before a conclusion is re t point to decide backward or forward is when a conclusion is finalized. Human default decision-making is a non-entrepreneurial approach until a person faces an extraordinary, problematic situation (J. Grunig, 1968). Most of our routine judgm non-problematic situations follow a backward reasoning strategy, whereas in extraordinary situations a forwarding reasoning strategy is required. What then is the merit of taking a backward reasoning mode? 95 To every living organism, living means a series of problem findings and certain ways of solving them (Popper, 1999). Some problems are not so difficult for a livin organism when a solution is available from past problem solving experience. Novel problems can be very threatening if one has never experienced them previously. When confronted with a novel probl g em, one must exert extraordinary effort to find a solution. Hence, nt rching for and applyin ing it when we identify a problem, we immediately begin an internal search?i.e., knowledge activation. If there is a solution or decisional referent available and applicable to the curre problem, our problematic situation ends fairly quickly. However, if no applicable knowledge is available, we initiate an external search for knowledge and information? i.e., knowledge action?until we reach a threshold of subjective confidence in dealing with the problem. In the first case, by applying a readymade solution immediately we can take a fast track to closure of a given problem situation. In that case, we take a backward reasoning approach to reach a conclusive solution by sea g a decisional referent immediately. We then need only to check evidence subsequently to confirm its utility for solving a given problem: i.e., cognitive retrogression that goes from conclusion to evidence. In contrast, forward reason requires sacrificing agile adaptation to a judgmental situation to some extent. Although increases one?s ability to reduce risk from potential problems, the forward reasoning approach heavily taxes cognitive resources. Thus, every living organism tends to use a problem-solving mode that allows for cognitive idleness so that one can invest surplus cognitive resources in alternative priorities (c.f., ?cognitive economizer?). 96 Backward reasoning and the problem of sunk costs. Quite often problem solvers are lazy in seeking counter-examples to evaluate the soundness of an alternative solu This happens more commonly when a congenial conclusion can be found that is consistent with the premises (Oakhill & Johnson-Laird, 1985). Problem solvers using a forwarding reasoning approach are more willing to change their mind (e.g., pr a certain solution proposal) in the presence of contradictory evidenc tion. eference on e. Because they move f ther . inger ive less desirable than e was made]? (Brehm, 1956, p. 384). me orward from evidence to conclusion, they can change their course of action ra easily even when they find some countering evidence against a preferred solution However, problem solvers who take a backward reasoning approach are less likely to be flexible because their conclusion makes it difficult to withdraw their commitment. This is the problem of sunk cost. What dissonance studies have found over the last half century is closely related to the problem of sunk costs. Dissonance theory (Festinger, 1957) is centered on the recurrent human tendency to reduce experienced dissonance. According to Fest (1957), people prefer ?reinforcing? cognitive elements (information) that favor the chosen alternative, whereas they avoid reinforcing information that favors the ?unchosen alternative? because of its dissonance arousal against the past choice. Brehm?s (1956) study found that once a choice is made, participants try to reduce dissonance ?by making the chosen alternative more desirable and the unchosen alternat they were before [the choic To be more adaptive, our cognitive commitment should be flexible and capable to shift among tasks to handle a stream of problematic situations. However, at the sa time, our judgmental conclusion, regardless of its enactment status, ought to be 97 unshakable once finalized. To revoke a choice already made or to reverse a co action to which one is already committed is prohibitively costly and laborious, especi when something interrupts our necessary transition to concurrent problems. Because of this, it becomes attractive and less costly to make the drawn conclusion more desirable and make the dropped conclusions less desirable. Under such conditions, we of our cognitive labors to support the already drawn conclusion after a decision has b made. We go cognitively backward. Also, we similarly go backward when we engage strong wishful thinking or willful thinking about the end state within a problematic situation. A strong desire for a certain outcome, despite obvious undesirabil long term, drives one to pick a course of action that fits with the wished or willed a prior conclusion. In this case, we assign our cognitive resources to postdecisional justification?to confirm the wished or willed decision we prefer. urse of ally invest most een in ity in the i the ight ate Although sunk costs ?should not affect decisions about the future,? decision makers are often tempted to favor one alternative over others mainly because of irrevocable prior costs we have paid (Dawes, 1988, p. 22). As Dawes (1988) argued, tendency to honor sunk costs and make a decision that preserves a prior investment m be rational or wise at the time. Yet, it becomes irrational in that it replaces the current and future consequences with past consequences. People honoring sunk costs in making a decision, then, are likely to pursue a backward reasoning approach for a most salient alternative?an option that has the most prior investment. The decision maker cognitively reasons backwardly to honor a favored alternative (conclusion). One searches internally and externally to justify, support, and honor a given solution candid with the highest nonrefundable sunk costs. In sum, the more problem solvers employ an 98 entrepreneurial mindset in problem solving, the less likely they are to use backward reasoning or to honor the nonrefundable deposits (sunk costs). Cognitive Multilateralism From the first day, education preaches to us that we be should be more op willing to tolerate even somewhat distasteful ideas in dealing with important life problems. A mindset that is open and willing to tolerate even distasteful ideas when considering alternative solutions to important life proble en and ms is a core value in entrepr t and one?s son- ed that we are capable of constructing multiple mental models eneurship. J. Grunig (1968) conceptualized ?the second stage of the decision process? in problem solving as ?a process of discrimination? (p. 26). He reasoned that the number of discriminations that problem solvers make ?among competing relevan alternative courses of action? is a critical quality with which to measure the entrepreneurial cognitive process (J. Grunig, 1968, p. 26). Entrepreneurial problem solvers are willing and capable to ?discriminate? available solution proposals by their merits. In doing so, entrepreneurial problem solvers first identify and increase the potential solutions (information) relevant to the problem. Hence, cognitive entrepreneurship in problem solving is closely related to the problem solver?s cognitive breadth. This breadth is measured by the number of alternatives one generates tolerance of rival information during the problem-solving process. Within the syllogistic reasoning framework, we illustrate such cognitive breadth as the number of syllogism models one would bear in judgmental situations. John Laird and Bara (1984) propos to make syllogistic inferences. Johnson-Laird and Bara (1984) reviewed psychological theories of syllogistic inferences and presented the theory of mental 99 models. Their theory proposes that logically competent reasoning is feasible without any use of rules. The way people reason is through attempting to test whether a conclusion must be true given that its premises are true. The specific steps in this process are 1) construct a mental model of the premises; 2) formulate an informative conclusion tha true in all models of the premises that have been constructed; and 3) if able to make a conclusion, try to construct an alternative model(s) of the premises that renders it false there is such a model, abandon the conclusion. Otherwise, the conclusion is valid.) (Johnson-Laird & Bara, 1984). Ordinarily, people attempt to build multiple models and to ?search for counterexamples to putative conclusions.? They have ?no machinery for making the search in a systematic way, and consequently often t is (if lapse into error? Bara, 1984, p. 51). This implies that people vary in their reasoning in er of models constructed and their degree of effort in conducting an aluative search for counterexamples to the premises. we seek out or construct a large number of alternative models, whereas t other times, we do not. More deliberate problem solvers (i.e., entrepreneurs) are more u iple syllogistic models that consist of multiple sets of major and conclusions. By doing so, they increase the potential to produce fective solutions. During the syllogistic inference process, entrepreneurial problem nalized a conclusion at the expense of dropping alternatives. Instead, ir confidence in their finalized solution. This becomes a rite of passage for the selected? the value of which is measured by how many rivals it has encountered and overcome to be crowned as the final one. I postulate here that the more a problem solver is cognitively (Johnson-Laird & terms of the numb ev At times, a likely to set p mult minor premises and ef solvers have not fi they utilize multiple fully constructed syllogistic models to increase the 100 entrepreneurial, the more one becomes effortful to increase the number of solution options (i.e., syllogistic mental models). With breadth of knowledge, information, rules, ed solutions available, individuals greatly enhance their potenti indset is old?i.e., 13 ents, or isagreement) with a set of c ir level statement 2 ( ) and abundance of propos al to solve problems. Measurement strategy. A problem solver with a more entrepreneurial m likely to increase his or her breadth of information, knowledge, decisional referents, and solution proposals. As a consequence, entrepreneurial problem solvers not only h accept ?a greater number of cognitive beliefs (e.g., information, decisional refer solution proposals) but they also tolerate somewhat conflicting and even incompatible beliefs in their cognitive inventory. We can use these tendencies as a ?yardstick? to distinguish non-entrepreneurial problem solvers from entrepreneurial problem solvers for measuring cognitive multilateralism. Thus, to delineate the two types, it is necessary measure the degree to which an individual is familiar with and will tolerate incompatible information. By letting respondents review and express agreement (d ompeting or contradictory statements about a problem we can measure the of cognitive multilateralism regarding that issue. For example, statement 1 ( ) : : statement i ( ) 13 It is important to note that ?even if one is aware of some contradictory view points? if not ?accepting? it, then it is not considered ?to have cognitive breadth? about the issue. J. Grunig and Hunt (1984) provided a useful taxonomy of objectives regarding communication effectiveness (p. 134). Retention of messages or efforts, problem solvers would increase their cognitive inven expect problem solvers with high cognitive entrepreneurship acceptance of cognitions is not necessarily to lead a behavioral intention or behavior. By communication tory to facilitate problem solving. Here, I to have cognitive breadth and tolerance for incompatible ideas. 101 ) statement (Followings are contradictory or competing statements to the former set of statements i+1 ( ) statement i+2 ( ) : : statement j ( ) The following formula computes one?s cognitive multilateralism: Cognitive Multilateralism = ( ) StatementsofNumber |)st st st ( - )st st (st | - )st st st st st (st j 2i1ii2 1j 2i1ii2 1 +?+++?+++?++++?++ ++++ Cognitive Commitment Cognitive multilateralism conceptually taps an entrepreneurial pro tolerance to somewhat incompatible ideas or proposals. In contrast, cognitive commitment conceptually taps an entrepreneurial pr blem solver?s oblem solver?s degree of enthusiasm atronizing the proposed solution for a given problem solving. trepreneurial problem solvers are typically excited by new ideas. They have a e for information that will help them solve a problem they face. At the s they increase the potential to solve a problem. They welcome and courage unconventional thoughts and can leave commonsensical assumptions behind. blem solvers value challenging ideas, fresh approaches, and original at might be considered distasteful to others. and the ex ent of pt En voracious appetit same time, entrepreneurs are energetic and enthusiastic about a wide range of ideas, even wild ones, long as a en Entrepreneurial rop concepts, even those th Entrepreneurial problem solvers are more committing?i.e., enthusiastic, patronizing, and incubating?to any ideas under consideration until they reach a final solution. Although there may be stark contradictions and incompatibility between rival 102 proposed solutions, most share the common quality of viability or feasibility. Thus, at least temporarily, entrepreneurs show a higher level of commitment toward ideas under consideration than do non-entrepreneurs. On average, their level of commitment to the ideas?as if they will be useful?are stronger than non-entrepreneurial problem solver Thus, the more the problem solvers become cognitively entrepreneurial, the more one is, provisionally, committing to the proposed ideas. Measurement strategy. Problem solvers with a more entrepreneurial mindset t to be more cognitively committing to or patronizing of thos s. end e considered ideas while moving als, al statement 2 ( ) : stateme toward problem resolution. They become more motivated to problem solving so that their level of cognitive energy and enthusiasm to the potential solutions increase situationally. In addition, to be deliberate in selecting and evaluating solution propos problem solvers have to patronize any potentially useful pieces of information until they finalize their decision to choose one solution over the others. Hence, a more entrepreneurial problem solver shows a greater level of cognitive commitment to the ideas and thoughts around the problem. Thus, the degree of attachment to pieces of information around a problem regardless of their interrelationships (e.g., incompatibleness among rivalry solution proposals), has to be higher for entrepreneuri problem solvers. We can compute it as following: statement 1 ( ) : nt i ( ) (Followings are contradictory or competing statements to the former set of statements) statement i+1 ( ) statement i+2 ( ) 103 : : statement j ( ) Commitment = StatementsofNumber j 2i1ii2 1 )st st st st st (st +?++++?++ and l can ic t, most routine decisions ?seldom take more than the simple discrim al with thinking). Therefore, some period of extended cognitive labor is required to fully ++ Cognitive Suspension Although an entrepreneurial problem solver is more enthusiastic, incubating, patronizing toward the proposed ideas for problem solving, he or she will withhold fina commitment to any single idea/option until all the necessary steps of scrutiny have been taken. Successful problem solving is closely related to the discriminatory power one exert in a given problem solving period. Cognitive entrepreneurship in problem solving, thus, closely correlates with one?s heedfulness in drawing a final conclusion. In problem solving, it is one thing to welcome and patronize many different ideas and another to discriminate and finalize (i.e., select and confirm in behavioral molecule terms) a specif solution from a set of rivaling options. In successful problem solving and decision- making, ?a relatively large number of discriminations are normally required?except when decision rules are formed? (J. Grunig, 1968, p. 31). In contras ination of recognition of a single alterative? (J. Grunig, 1968, p. 31). To de a ?genuine problem,? not a ?routine problem,? problem solvers should discriminate between available solutions by considering the problem solving potential of each. Their high level of patronage and enthusiasm does not mean a blind commitment. Entrepreneurial problem solvers discriminate the virtue and viability of a given solution by merits, not by any affection they might have incidentally developed (e.g., wishful 104 evaluate and confirm a solution. Even after a solution is isolated, an entrepreneur will take one additional step of scrutiny?i.e., finalizing it by evaluating what can go wrong if I select 4)? ination in that inaction caused by too much procrastination can actually worsen a situation. In other words, sometimes no decision is worse than a less than optimal decision. Inaction brought on by too much cognitive processing is as non-entrepreneurial as backward cognitive processing. I delimit cognitive suspension as focusing on the cognitive process, not to encompass the behavioral process of problem solving. Entrepreneurship requires a problem solver to be agile in translating a selected solution into action (agility in the behave stage in terms of the behavioral molecule or in the effectuating phase in the APS). Thus, behavioral suspension?one?s abeyance to act on a chosen solution, is he steps of ten thus, it is non-entrepreneurial problem solving. In ioral gnitive ch to - ntrepreneurial. this option (checking the possible ?Murphy?s law,? J. Grunig & Hunt, 198 before taking action. However, cognitive heedfulness is different from cognitive indeterm C non-entrepreneurial. However, cognitive suspension?defined as taking all t generating, evaluating, selecting, and finalizing a solution? makes problem solvers reach better problem solving outcomes. Thus, suspension of an action after a conclusion (a confirmed solution) is problematic because it delays problem resolution. This of prohibits early problem solving and contrast, cognitive suspension as mental heedfulness to take all steps of the behav molecule before finalizing a solution increases problem-solving effectiveness. Co suspension en route to a finalized solution is thus part of an entrepreneurial approa problem solving, whereas behavioral suspension en route to problem solving is non e 105 In summary, cognitive suspension refers to a problem solver?s heightened willingness to invest discriminatory efforts in evaluating options and reevaluating a selected option before finalizing it. The behavioral molecule captures this subtlety the confirm step, which occurs after select and before behave. I characterize a problem solver with high cognitive entrepreneurship as having stronger cognitive suspension in problem solving. Entrepreneurial problem solvers give closer consideration in drawing and finalizing a concl with usion; therefore, they are more willing to take extended steps of ing and discriminating before making a final choice. ns of cognitive entrepreneurship in problem solving in Figure 8. Figure 8: model of cogni blem solving scrutiniz Summary. I have constructed a conceptual model that highlights the four dimensio A tive entrepreneurship in pro . m solver becomes cognitively entrepreneu ore one vies for and the more one is heedful in drawing conclusions and finalizing proposed solutions. Thus, I summarize the hypotheses: Figure 8 shows that the more a proble rial in dealing with a problem, the less one reasons backward, the m cognitive breadth and tolerance, the more one commits to potential solutions, 106 H7: Th H8: Th e multilateralism. 9: The higher the cognitive entrepreneurship in problem solving, the higher the H10: T tween ega Strategy e higher the cognitive entrepreneurship in problem solving, the lower the cognitive retrogression e higher the cognitive entrepreneurship in problem solving, the higher the cognitiv H cognitive commitment. he higher the cognitive entrepreneurship in problem solving, the higher the cognitive suspension. Cognitive Alpha and Omega Strategies (CAOS) From the model of cognitive entrepreneurship in problem solving, I conceptually derive the cognitive alpha approach as the heightened state of situational cognitive entrepreneurship and the cognitive omega approach as the diminished situational entrepreneurial mindset toward a given problem. The two conceptual strategies differ only quantitatively but not qualitatively. Table 2 summarizes the relationships be cognitive alpha and omega strategies and four dimensions of cognitive entrepreneurship in problem solving. Table 2 Cognitive Alpha and Omega Strategies and Cognitive Entrepreneurship in Problem Solving Cognitive Alpha Strategy Cognitive Om Retrogression More forward reasoning More backward reasoning Multilateralism More cognitive tolerance Less cognitive tolerance Commitment More committing Less committing Sus a solution pension More heedful in finalizing a solution Less heedful in finalizing Entrepreneurial Approach rial Cognitive More entrepreneurial Less entrepreneu Behavioral molecule and CAOS. J. Grunig & Hunt (1984) used the beha molecule to classify common managerial mistakes in problem solving. CAOS provi simpler typology to explain the differential problem solving approaches. They are: Dogmatism: detect?select?behave [omega] vioral des a 107 Rationalization: detect?select?behave?construct [omega: justification?a more sophisticated omega approach] Procrastination: Detect?construct?construct?construct [alpha] Indecision: detect?construct?define?select?construct?define?select? construct [alpha] define?select?confirm?construct?define?select?confirm?construct [alpha] Default cognitive strategy. I postulate that the human default cognitive strategy i cognitive omega rather than cognitive alpha. People generally only adopt a cognitive alpha approach when they face a problem without a readily available solution?i.e., no routine or Habit: detect?behave?detect [omega] Perfectionism: detect?construct?define?select?confirm?construct? s n- extraordinary problems. In contrast, people more often will take a cognitive omega roach a given s nking from a cognitive omega to a cogni approach when they have a problem with a readymade solution?i.e., routine problems. This gives us an intuitive explanation for why the cognitive omega app becomes the default mental approach. Problems are always fewer than non-problems. The cognitive omega strategy lessens cognitive effort for the present problem to economize problem-solving capacity for other concurrent or more urgent tasks at moment. When encountering familiar problems, the cognitive omega approach increase one?s ability to adapt to other problems by speeding up the problem-solving process. However, when encountering unfamiliar problems, we cannot maintain our non-thi and minimal cognitive investment. Then, we are likely to shift tive alpha strategy to compose a new solution and to restore one?s default cognitive idleness (c.f., Carter, 1965, ?evaluative mode? and ?reinforcement mode?). 108 To summarize, the cognitive omega approach could be described as the shortest path through the behavioral molecule using the fewest steps, that is, ?detect?behave,? whereas the cognitive alpha approach is the longest path using all the steps and completing the full process of the behavioral molecule, that is, detect?construct? define?select?confirm?behave until a problem situation has ceased to be problematic (J. Grunig & Hunt, 1984). Many human motor behaviors (e.g., blinking if a person detects a sudden movement near the face) are done by the cognitive omega strategy. Many t d the the itive cle to n ?i.e., a cognitive ambidexterity in changing one?s mental approac imes, we do not have any intention regarding a certain behavioral decision beyon simply acting itself. However, if we detect an out-of-the-ordinary situation for which motor-behavior-like response will be ill-suited, we are likely to make a transition to cognitive alpha approach to better adapt to the new problem. Here the model of cogn alpha and omega strategies (CAOS) captures the human tendency to establish and recy certain knowledge that allows us to extend the use of the cognitive omega state. Thus, when perpetuation of a readymade solution is difficult, our cognitive working goes in an ?extraordinary cognitive mode? until we have decided upon a novel solution. Such a extraordinary cognitive modus operandi is the cognitive entrepreneurship in problem solving. A normative implication of CAOS is that it is problematic if a person does not have cognitive aptness h from cognitive alpha to omega and vice versa. For instance, many serious health problems become worse because of the problem holder?s cognitive ineptness (e.g., maintaining a cognitive omega approach to a new problem either deliberately or otherwise). At the other extreme, many people also suffer from unnecessary cognitive 109 stress by employing a cognitive alpha approach even though a cognitive omega approach would adequately deal with the problem. Therefore, neither cognitive alpha nor a cognitive omega strategy is invariably superior over the other. Rationality assumption in cognitive alpha and omega strategies. J. Grunig (1968) criticized the rationality assumption in major programs in economics and communication because they consi dered a rational person to be a ?profit maximizer? or one who ?seeks always 968) seemed to equate high entrepreneurial decision-making with high ra itive , em .g., to maximize a pre-set goal? (p. 4). Against such a presumption, he extended the meaning of rationality to be construed as one?s ability to find out and evaluate alternative solutions to a problem and to choose one by its merits. Thus, he studied the conditions under which a person becomes a rational entrepreneur. His conceptualization of the entrepreneur paved the way for decision-makers to become more rational in problem situations. In contrast, I conceptually separate rationality from entrepreneurship in the model of cognitive alpha and omega strategies. J. Grunig?s (1 tionality in tackling a problematic situation. However, the model of cogn alpha and omega strategies considers the highly entrepreneurial approach?i.e., the cognitive alpha strategy?and the low entrepreneurial approach?i.e., the cognitive omega strategy?as orthogonal from judging one?s rationality in problem solving. In other words, to be more entrepreneurial is not always to be rational. For example, with situational constraints in a problematic situation such as low cognitive capacity (i.e. lacking cognitive resources)??hardware aspect??and high cognitive capability (i.e., having a ready solution)??software aspect? (Kruglanski & Thompson, 1999), a probl holder would be considered more rational by adopting a cognitive omega strategy (e 110 backward reasoning strategy) with a well rehearsed conclusion (prior solution applicable to current problem). To determine what is considered rational problem solving requires thinking about unique situational conditions (e.g., constraints) in problem solving contexts. Specifically the cognitive alpha and omega model no longer equates an entrepreneurial approac rationality in problem solving. The less entrepreneurial approach can actually be m rational if it reaps the reward of economizing cognitive capacity for the problem solv With this concept, I identify a key problem from which many decision-makers suffer. This is the problem , h with ore er. holder?s ineptness to make flexible shifts from cognitive alpha to cognitive omega and vice versa when si ntexts demand mental dexterity. The more o n- ion here aims to tuational co ne decision-maker is inept at this changing, the less the person is able to adapt to the environment. CAOS as a descriptive and normative theory. Rational decision-making theory is a normative theory in that it extracts a portion of the phenomenon of the human decisio making process. In contrast, CAOS is a descriptive theory in that it encompasses both notions?i.e., rational and (somewhat) irrational aspects of decision-making. In Carter?s (1972, September) terms, most theories derived from the rationality assumption are used to construct a procedure?to make a practice better?whereas the CAOS concept construct a conceptual narrative toward a process?i.e., to describe a phenomenon better. In the cognitive alpha and omega model, I conceptualize that either the entrepreneurial alpha approach or the less-entrepreneurial omega approach can be rational or irrational. The prime factor demarcating the boundary between rational or 111 irrational in cognitive strategy is the problem holder?s choice made with consideration of his or her situational constraints (e.g., internal and external such as knowledge necessary and resources necessary to deal with the problem). If one keeps procrastinating in making a decision even with an applicable prior solution (i.e., referent criterion), his or her over-deliberativeness (i.e., cognitive alpha approach) should be called irrational. Likewise, if someone hastily completed a problematic situation without creating an applicable solution but used only strong wishful thinking (i.e., cognitive omega approach), his or her lack of due consideration would also be considered irrational. Problem solver should be able to shift from cogni s tive omega to alpha and vice versa in accordance with situatio omega oblem solution; 2) general g nal constraints. We can describe a problem solver?s incapability of changing as lacking meta-rationality?that is, rationality about rationality during the problem resolution period. The lack of meta-rationality causes a meta-problem?that is, a problem in dealing with a problem. Thus, the CAOS dichotomy offers a normative lesson for problem solvers by decreasing meta-irrationality, hence making them better problem solvers. Summary Cognitive entrepreneurship in problem solving and the cognitive alpha and models describe different mental approaches under problematic situations. A problem solver with heightened cognitive entrepreneurship tends to 1) generate a large number of mental syllogistic models before he or she finally selects one for pr ly commits more to proposed solution proposals, as if they are a solution, durin evaluation; 3) is more heedful in finalizing a conclusion; and 4) is more likely to invest 112 cognitive labor prior to finalizing a conclusion (i.e., an evaluation purpose) rather than to spend cognitive efforts after finalizing a conclusion (i.e., justification purpose). To define the situational variations of cognitive entrepreneurship in problem solving, I introduced the alpha and omega cognitive strategies. Sometimes, problem solvers internall dge and evaluate its ?situation e identified evidence. some situations, proble conclusion? reasoning. People make a decision ver then ferret out evidence (reasons) that jus l a of r behavior precedes any cognitive elaboration?i.e., cognitive omega (or even absence of it), such as whe n (action) and subsequently justify the preceding des any overt action?i.e., cognitive alpha. A prob the cognitive alpha strategy is vying f y and externally scrutinize available and applicable knowle al relevance? in warranting a conclusion from th Thus, one follows a process of reasoning? conclusion. However, in m solvers take an alternate approach such as (reasoning)? y quickly and tifies the hastily made decision. In such an instance, external and internal evidence seeking compensates for an ill-conceived prior decision. I distinguish the latter reversal approach called a backward cognitive strategy from the former, which I cal forward cognitive strategy. The backward reasoning strategy is likely to result from willful or wishful thinking to achieve a certain decision outcome (the inclination to take a stand without just grounds or sufficient information) or from premature engagement influential prior decisional rules. At times, ou n we make a decisio action. At other times, our cognitive effort prece lem solver using or a perfect solution selection for a problem, whereas a problem solver using the cognitive omega strategy is vying for a perfect justification for a preceding decision. A cognitive omega problem solver may have lower aspirations for information but has no 113 lower aspiration for problem resolution. However, determining the rationality of problem solvers by their choice of cognitive strategy is futile without considering th situational conditions under which the decision was made. Rationality should be judged only through the eyes of the beholders?i.e., the problem solvers. Evans (1989) noted that human reasoning has many variations. He wrote that one?s ?apparent competence in?reasoning exhibited under one set of circumstances is s frequently absent in others? (p. 7). Across various problematic situations they encounter, problem solvers take different mental approaches in dealing with the problems. Then, w should question how and why such situational variations occur. In the remainder chapter, I will turn to antecedent conditions to account for such situational variations in communicant behavioral patterns and cognitive strategies under problematic situations. ANTECEDENTS OF COMMUNICANT ACTIVENESS e o e of AND COGNITIVE ENTREPRENEURSHIP: AN INTRA-INDIVIDUAL ACCOUNT OF PROBLEM SOLVING All organisms are problem finders and problem solvers. Karl Popper The objective of this chapter is to build a theoretical account of the antecedent conditions that explain two new constructs: CAPS and CEPS. These two communicative and cognitive features are recurrent phenomena that bear important theoretical and practical implications to problem solving. The two previous sections identified the phenomena to be accounted for?i.e., conceptualizing two dependent variables?by the independent variables I am going to elaborate here. In what follows, I describe the 114 conditions that precede communicant activeness and cognitive entrepreneurship in problem solving. Human Perception We see a map of the world, not the world itself. Albert Einstein Our universe dives into a sea of our perceptual experience. Once we perceive th something is problematic, that we are not constrained from doing something about it, that we feel close connection to a problem, we act on our perception. We do not a directly on the actual sources that trigger perception. In the situational theory of proble solving, I delimit the scope of antecedent variables to perception, not to sources of perception. In other words, what prompts our actions to do something about a problem is our subjective perception of feeling problematic, unconstrained, and connected, not the objective things that trigger such senses. Therefore, to gain conceptual coherence, we need to theorize from our internal perceptions to our action rather than to theori the external or internal sources of those perceptions to our action. In the latter theorizing approach, external object variables as independent variables (e at or ct m ze from .g., expert estimation of problem s) seriousness) explain the dependent variables of action (e.g., a public?s information seeking about a problem). In contrast, in the former approach, subjective variables as independent variables (e.g., a public?s perception of problem seriousnes explain the dependent variables of action (e.g., a public?s information seeking about a problem). Problem solvers will not initiate a communicative behavior or perform any cognitive labors unless they perceive a reason for it. For that reason, a message (a source 115 of perception) cannot trigger action unless it triggers a sense of problem recognition (perception). Our perception is often biased and inaccurate. However, regardless o seriously our perceptions are flawed, people consistently respond to their own subjective perceptions. Therefore, knowing about people?s perceptions can help in predicting their subsequent behaviors. For example, some people mistakenly overestimate current obstacles against them to do something about their perceived problem. Even if a b is simply exaggerated or illusory, one will not do anything because his percei (constraint recognition) lowe f how arrier ved obstacle rs his motivation to act. Thus, he would not take any action. Misperceptions are illusor t do on nt d ?the eyes of the beholders? perceive the situations that confront them. y, but their consequences are substantial. Suppose that a doctor knows that her patient misperceives and overestimates barriers to improve his health. The doctor may not understand why her patient canno something to change a problematic situation. To use a doctor?s more objective evaluati of the actual level of constraints cannot necessarily predict the likelihood of the patie taking action to do something about his problem. Rather, despite being pathetically inaccurate, people?s own misperceptions will become stronger predictors of their subsequent behavior in dealing with the problem, even if less than optimal. What this suggests to us is that many theoretical efforts in communication behaviors adopt, more or less, the objective criteria of non-perceptual variables (e.g., education level, income, an psychographics) rather than subjective criteria dealing with matters of perception (e.g., problem recognition, constraint recognition, level of involvement). We perceive and construct our own reality, and then behave on that subjective reality. Thus, to understand how people approach their problems, we need to explain better how 116 Therefore, I delimit the following conceptual efforts to apply to the situational perception found in ?the eyes of problem beholders.? Drawing a conceptual boundary in because, without such a delimitation, external versus internal distinct e By us. cedes our evaluation of its importance to our l percept commu someth Specifi crowd model? hierarchical model of problem detection that provides a conceptual basis for concepts and this way is necessary ions can complicate the resulting account. For example, in a prior study of th situational theory of publics, the temporal order of involvement and problem recognition questioned whether problem recognition precedes involvement or involvement precedes problem recognition (J. Grunig & Childers, 1988). It remains a chicken-and-egg debate unless we can delimit our conceptual boundary to only the perceptual variables. demarcation of independent variables to perceptual ones, we can solve the debate. Without a perceived problem state, there is nothing to evaluate how it is connected to We never evaluate the importance of something until we know its existence. Therefore, our perceptual awareness of a problematic state pre ife. Next, I will construct a perceptual account of what kind of situational ions would prompt our problem-solving efforts within cognitive and nicative dimensions. A Hierarchical Model of Problem Detection How do we learn that a problem exists? Or, when do we notice that there is ing so problematic that it subsequently initiates our problem-solving efforts? cally, how do we begin to pay attention to something as being problematic in a of alternate candidates vying for our attention? Here, I propose a conceptual a verbal theory 14 ?that answers these questions. This theoretical account is a 14 In terms of Popper?s (1999) terminology, the verbal theory I introduce in this section is a theory with high ?logical content? rather than ?empirical content? that lead me to a testable study (Popper, 1999, p. 19). 117 propositions of antecedent parameters within the situational theory of problem solving (STOPS). A hierarchical model of problem detection originates from what Karl Poppe called ?conjectural knowledge? (1963). Popper (1963, 1999) said in his evolutionary epistemics that all perceptual knowledge presupposes a priori knowledge for an organism. Conjectural knowledge may or may not be valid. Regardless of its validity, it becomes a referent frame to judge what is presently wrong perceptually. Human bei r y living ngs become d person perception comes ahead of some cognitive needs. Earlier schools percept Moskowitz, 1996, p challen le seek no 1996, p. 377). t dissatisfied state from which we struggle to free ourselves and pass into a state of do not wish to avoid, or change into a belief in anything else.? The struggle to d closure. According to Dewey (1938), this process of turning indeterminate ate ones, of turning a state of inconstancy to one of constancy, ?like all activity is stimulated by discomfort, and the particular cognizant of something problematic when they perceive a perceptual discrepancy??the clash between hypotheses and reality? (Popper, 1999, p. 47). In the following, I will discuss in detail how we come to perceive something as problematic. Problem Detection How do we know when something is problematic? Most psychologists an philosophers agree that of social psychology (e.g., Associationists) believed in ?data as strictly driving ion and humans as seeking accurate knowledge (truth)? (Gollwitzer & . 377). However, Instrumentalists such as C. S. Peirce and John Dewey ged the assumption of accurate knowledge-seeking by pointing out that ?peop t truth, but simply an end to doubt? (Gollwitzer & Moskowitz, Gollwitzer and Moskowitz (1996) made a comprehensive historical review about the shif of the assumptions. They succinctly summarized: Peirce states (1877, p. 66) that the irritation of ?doubt is an unhappy and belief, while [the feeling of believing] is a calm and satisfactory state which we end doubt was labeled as a process of Inquiry that produced what Gestalists calle situations to determin 118 discomfort concerned is called ?doubt,? just as hunger is the discomfort that stimulates eating and thirst is the discomfort that stimulates drinking.? Thus, the seeking meaning and reducing doubt (similar to what Festinger, 1957, labeled oubt, sets the person off on to what Dewey (1929) called a quest for certainty. This any knowledge that will end doubt quickly and produce closure (so long as it is (Gollwitzer & Moskowitz, 1996, p. 377) A need to do something arises when a living organism perceives a doubtful situation, because an indeterminate situation is problematic to a living organism. Unless the ?doubt? or indeterminacy turns into a determinate state (i.e., ?closure?), the living organism will avoid this situation of ?discomfort.? This imposes a most fundamental problem?to live on. Thus, a living organism initiates a ?quest for certainty,? maybe through ?accurate knowledge? or through ?any knowledge? that can achieve a satisfactory degree of closure. Naturally, any living organism evolves to have an internal detection mechanism to anticipate a potentially doubtful or discomforting situation. Related to this detection mechanism, Simon (1967) proposed a model of dual mechanisms?goal-terminating and interrupt mechanisms?in the human nervous system. He reasoned that the human mind needs an interrupt mechanism to redirect cognitive resources to the most important tasks. Marcus, Neuman, and Mackuen (2000) reviewed Simon?s model of the working of the nervous system in proposing their model of affective intelligence. The human nervous system?is primarily a serial processor of information. He emory to support this contention. Such serial processors require two support redirect attention when goal-oriented behavior has reached a satisfactory state in environment requires an interrupt mechanism to redirect human attention to processing system operates in the service of needs to gain a sense of control, avoidance of dissonance). An upset or imbalanced system, one beset by d quest can occur through perusing accurate knowledge or through the pursuit of experienced as being a good enough or sufficient conclusion). [italics added] [Simon] reviews research on attention, temporal response intervals, and m mechanisms, first a goal-terminating mechanism (Simon?s term is satisficing) to terms of an initial goal. Second, the human organism living in a demanding 119 higher priority real-time needs, no matter the ongoing effort to secure some antecedent goal. [italics added]. (Marcus et al., 2000, pp. 6-7) To answer for why we need a detecting mechanism, thus, we must consider that it helps living o commu Specifi us to know when we turn how much of problem-solving resources in what direction for how lo w the , 63). f . Observ out rm such s achine with a hope to get a jackpot, thinking the time for a jackpot has come after a series of losses. Or it could be even genetically-coded inborn rganisms mobilize their limited problem-solving capacity (i.e., cognitive and nicative resources) to generate a better fit with their living conditions. cally, a perceptual detecting mechanism for impending problematic states allows ng of a time period. Expectation, Observation, and Unbearable Badness-of-fit How do we detect a problem? To understand this process, we need to kno basic mechanism of perception. Two key concepts that comprise the perceiving mechanism are expectation and observation (c.f., ?conjecture?) and their joint interaction the perceived degree of goodness-of-fits between two (c.f., ?refutation?) (Popper, 19 We continuously expect something out of living. At the same time, within the limits o our perceptual capacity, we continuously test the validity of our expectations ation tests the reasonableness of our expectations. By expecting something ab what we are about to experience and by testing its fit with experiential objects or events (i.e., perceived reality), we get an internal sense of experiencing. The content of what we expect can be anything. It could be transferred from experiential knowledge?i.e., our past experience or learning in a very specific fo as ?after a cold winter, there comes a warm spring.? It could be derived from simple inferential knowledge?i.e., a guess or a wish about something to experience such a pulling the arm of a slot m 120 knowledge in a living organism such as a newborn baby who expects to be fed by its mother (Popper, 1963). However, without expecting something or some state of conjectural a prior knowledge, no matter how crude its content, we cannot experience anything (Popper, 1963, 1999). A biological research finding suggests that the role of expectation- observation-refutation would be essential in our perception of the world (Radnitzky & Bartley, 1987). According to McCulloch (1965), frogs register only four kinds of visual effects because only four types of signals can be sent to their brains. McCulloch (1965) summarized: The frog does not seem to see or, at any rate, is not concerned with the detail i of stationary parts of the world around him. He will starve to death surrounded by movement. He will leap to capture any object the size of an insect or worm meat but by any moving small object?. His choice of paths in escaping enemies is darker. (p. 231) To live, frogs only need four kinds of visual effects that enable them to accomplish such tasks as catching small moving objects (e.g., flies) and leaping towards dark spaces if they encounter a predator. To frogs, the world consists only of the contrasts, the small dark objects, the moving shadows and sudden dimming of light that they perceive. A frog?s knowledge of the world is not given (i.e., data drive perception like the Associationists? claim), but is the product of the evolved sense organs and conjectural knowledge that only reflect some, but not all, aspects of the world (i.e., a priori conjectural knowledge drives us to perceive some aspects of world). Why cannot frogs perceive things other than four objects in this world? Our experiential world, obviously, consists of many more than four things. Irrespective of food if it is not moving. His choice of food is determined only by size and providing it moves like one. He can be fooled easily, not only by a bit of dangled does not seem to be governed by anything more devious than leaning to where it 121 their presence, frogs have difficulties in perceiving them because they have no matchin internal theory to be used for expecting the immediate environment they face. Like fro other living organisms cannot perceive without the available content material of expectations and the very act of expecting toward the immediate world. Unlike frog human beings are far more evolved in our sensory organs and in our repertoires of conjectural knowl g gs, s, we edge. t like our the world we are about to experience. Experiencing an object or an event ental pulse we project and the rebound that we perceive through our xpect consciously or unconsciously. However, we ca t the very act of expecting. ry are as its We constantly apply and evaluate the adequacy of content we are expecting via observing the areas of experience that are defined by content material of an expectation. We are ?anticipation machines? (Dennett, 1991, p. 177) that project the beams of expectation onto our environment in the same way a sonar or radar would. Withou transmitting a pulse of sound and receiving its rebounding echo, we cannot ever detect anything regardless of its presence to us. The pulse of sound in sonar or radar is mental scan of presented to us is only possible by the m sensory organs?the act of expecting. We may e nnot perceive withou In addition, just as we load bullets into a gun, we can load anything into our expecting gun, from pleasant events to boring routines, or from very general ideas to ve specific knowledge. We are able to perceive by our act of expecting. Moreover, we able to expect many things at the same time. The perceptual gun in our mind h outstanding capacity to load and shoot multiple and seamless mental pulses simultaneously. 122 By way of analogy, we are like a turned-on projector loaded with various slides or films (conjectural hypotheses or lay theories about our living). We select and lo contents into our mental projector corresponding to the life tasks we have to deal with at a given moment. Our continuous pro ad jection of what will (should) be to the screen (world we are will n of things (conjecture) always comes first. Once an expecte on (conjec hostile experie lay theo problem Role of experiencing) is then compared with what is on the screen. The lack of fit between what will be and what is redirects our cognitive resources to it. As mentioned, expecting an a priori theory/hypothesis is a necessary condition for any perceptual process and experience. However, a variance?a psychological distance?is created when our cognitive recycling attempt (expecting by applying a hypothesis of what (should) be) yields a failure of fit to what is. Once a person identifies a variance, he or she tries to explain the variance by mending conjectural hypotheses (i.e., changing cognitive knowledge) or by modifying observational contents (i.e., changing reality through action). To summarize, the expectatio d state (e.g., ?The meeting will be pleasant?), which is the content of expectati turing), is found to be discrepant from the observed state (e.g., encountering remarks from others), one detects a problem. This is a problem in that one nces a failure to confirm the adopted hypothesis (e.g., pleasant meeting)?i.e., a ry about a present state. This is how we begin to perceive something as atic. Emotion in Problem Detection A problem must be felt before it can be stated. John Dewey (1938) 123 Studies about information seeking draw heavily on the notion of uncertainty, it as if it istreating a triggering mechanism for the subsequent pursuit of information. negative emotional strain such as anxiety uncerta seek in 2002). the con fifi & Wein conside us to de tem. What causes us to draw our attention from a prese s our , ion he f two distinct systems consisting of a disposi 0) evaluation. That is, it demands an assessment of the effort, the prospects of success and failure of the sequence of actions. For humans, these strategic Uncertainty perceived by a person would lead to . Anxiety can be described as ?the affective (emotional) equivalent of inty? (Gudykunst & Nishida, 2001, p. 59), and it tends to motivate a person to formation to reduce the experienced negative psychological state (Afifi &Weiner, Thus, affective states such as anxiety are commonly used to account for bridging ceptual distance between uncertainty and motivation in information seeking (A er, 2002). However, in the present study, I adopt a more general view, ring affective states other than anxiety. In essence, any affective state can allow tect a problematic situation. Dispositional vs. surveillance sys nt mental task to a new one? What kind of ?interrupt mechanism? redirect attention to a newly emerging problem? Simon (1967) pointed out that, in general theories of human information processing are muted about the interaction of cognit and affect. Simon foresaw some role of emotion in cognitive systems (e.g., goal- terminating systems and interrupt systems), but he himself did not work on it (Marcus, Neuman, & Mackuen, 2000). Marcus et al. (2000) found a few theoretical works on t affect-cognition interaction and developed a model o tion system and a surveillance system in cognitive processing. Marcus et al. (200 explained the dispositional system: Strategic action, behavior designed to achieve a purpose, requires an ongoing success, the current stock of physical and psychic resources, and feedback on the 124 considerations are only occasionally governed by conscious calculation. Mor emotions of the disposition system provide precisely this guidance. When our tell us that we are bursting with confidence, energy, and eagerness. Alternatively exhausted and beaten. Shifts in the direction of increased elation strengthen the e often these executive functions are done subconsciously. Importantly, the feelings are focused on ourselves, changes in mood from gloomy to enthusiastic , when our mood changes in the direction of depression, we conclude that we are motivation to expend effort and strengthen confidence in a successful outcome. d effort and undermine confidence that the outcome will prove successful. toward the action? the disposition system relies on emotional assessment to enthusiastic and we abandon those that cause us despair. The disposition system that are already in their repertoire of habits and learned behaviors. [italics added] he surveillance system, acts to scan the environment for novelty and sudden intrusion of threat. It serves to warn us when we cannot rely and some people are powerful and dangerous. This system uses emotion to of its ongoing analyses. It generates moods of calmness, on the one hand, and anxiety, on the other. Here we focus on its attention-related y on ? So long as the comparison shows no discrepancy etween expectation and reality, the system generates a sense of calm and remains r, it evokes increasing anxiety, it interrupts ongoing activity, and it shifts attention d] (Marcus et al., 2000, pp. 10-11) A key contribution of their model is that emotion plays not only a role in managing routine tasks but also in redirection toward an emerging task. This is a straightforward extension of Simon?s (1967) model of the workings of the human nervous system (i.e., goal-terminating vs. interrupt mechanism). However, to make a Shifts in the direction of increased depression weaken the motivation to expen Accordingly, this emotional calculus is translated into a summary disposition control the execution of habits: we sustain those habits about which we feel provides people with an understanding, an emotional report card, about actions (pp. 9-10) About the surveillance system, they described: The second system, t on past learning to handle what now confronts us and to warn us that some things signal the consequences properties. Identifying two systems in the limbic region of the brain suggests that people rely on their feelings to assess how well they are doing, and they rel their feelings to scan for signs of threat and uncertainty. What is interesting about this second emotional system is that the onset of increased anxiety stops ongoing activity and orients attention to the threatening appearance so that learning can take place b unobtrusive. When the system detects unexpected or threatening stimuli, howeve away from the previous focus and toward the intrusive stimuli. [italics adde 125 more general accounting of the roles of emotion, their ?emotional report card? conce should include more than ?anxiety? or ?calmness.? As I discuss next, any emotional states can play the role of an emotional report card. Trade cognitive complexity for simple affective knowledge. We must adapt handle multiple life problems simultaneously. For example, breathing, seeing, listening, answering the phone, writing, and meeting deadlines can come all a pt and t once. Hence, we develop rrent equires e tasks. e the future of the environment: that is, all future states of the environment. be adapted to the future conditions of the environment; and in this sense general knowledge comes earlier than momentary knowledge, than special knowledge. [italics added] (p. 49) and apply certain ways of behaving, responding, and adapting to such recu situations in an efficient way. Once we get used to maneuvering through a common problem, we find few reasons to maintain our cognitive investment to supervise each solution application. This is ?automaticity??defined as ?capable of operating by itself without any need for conscious guidance, once put in motion? (Bargh, 1996, p. 173). To handle multiple specific tasks simultaneously (which is a mandate for any living organism), we need to devise some way of ?efficient supervision.? This r some simplifying mental mechanism to reduce the complexity of handling multipl Prototype hypotheses can simplify by trading complexity for a certain type of simpler cognitive information, often in the form of affective information (e.g., confidence in dealing with the problem, good or bad, pleasant/comfortable vs. unpleasant/uncomfortable). Prototype hypothesis. Popper (1999) described our epistemic hypothesis testing about future states to life itself: So we reach the conclusion that life must from the start anticipate in some degre Perhaps it is just a question of hours, or perhaps of millions of years. Life must 126 Anticipating with a general knowledge such as affective knowledge allows us to test our hypotheses about current and impending future states. Unless the general knowledge is invalidated, our mental process has no reason to be interrupted. The prototypic hypothesis testing is, thus, managing and supervising multifarious living tasks. We are routinely challenged by a number of mental tasks arising at the same time. We need a meta-cognitive function that guides and manages routine tasks. Meta- cognition supervises how we are doing, rather than what we are doing. Hence, we have evolved to deal with such a juggling problem by trading cognitive complexity for a simple affective state. By lumping various forms of meta-cognitive feedback into a single piece of affective state?i.e., ?emotional report card? (Marcus et al., 2000), we can condense and load them as one bullet into our mental gun of expectation. We can effectively test our hypothesis of ?going well? by summarizing the status of many mental tasks with a single prototypical content such as ?feeling pleasant.? A fishing analogy can illustrate the role of general knowledge and special knowledge in managing multiple cognitive tasks efficiently. Suppose that we throw out many fishing lines (e.g., ten) simultaneously. How can we efficiently handle these multiple fishing lines without sacrificing our need for other mental commitments? Of course, watching them individually makes us busy, indeed too busy to handle any other work (e.g., reading books). A way to control multiple fishing lines is knotting them together. By merging all the lines into one and holding them as a bundle?i.e., lumping?we can reduce the necessary mental resources in supervising all of the individual lines. Until this system fails, we can efficiently meet our need for managing multiple tasks through the cheapest expenditure of cognitive resources. Once the bundle 127 as a whole is in motion, we next look for which specific line causes the movement. We economize our cognitive work until a sudden move arises. Similarly, a human being recognizes a problem and its source, first by lumping manifold life tasks together. On we detect a strike?experiencing a betrayal of our expected content by observa then investigate the causes of the betrayal. Indeed, just a few mental knots (i.e., ma clusters of prototype hypotheses) can handle numerous mental supervisory tasks. betrayal of such a prototype hypothesis is experienced, it is quite conveniently detectable and the person can look for which part caused the problem. By loading affective knowledge i ce tion, we ny Once a n our expectation gun (e.g., the sense of being pleasan sant). ts our ell with observational knowledge, it then tri , an explana situatio t), we can efficiently abstract many physiological, physical, and psychological tasks at any given moment. Any malfunction in any area of these life tasks will refute our expected mental hypothesis of ?affective knowledge? (e.g., sense of being plea This will break into our mental placidness?i.e., perceptual arousal. It then redirec cognitive resources until we found a cognitive solution for the perceptual problem. If we are not able to find a solution (e.g., an explanation of why the ?unpleasantness? happened and/or how to deal with it), then we enter into a situation that is a meta-problem?i.e., a cognitive problem. Emotion is a kind of ?general knowledge? (e.g., pleasantness will be continued) that is supervising other mental tasks (Popper, 1999). Once general knowledge that is loaded in expectation does not fit w ggers a subsequent search for and identification of ?special knowledge? (e.g. tion about how it arises and how to behave in response). Before a problematic n, we can manage many critical living tasks with very simple, general, and 128 abstrac affective knowledge is sufficient to supervise other mental tasks. General knowledge is very primitive, but it provides substantial cues to detect an eruption of a problematic state. However, we subsequently demand more specific knowledge about what causes the discrepancy and how we can narrow the gap. Our prototypic knowledge evolves from primitive conjectural knowledge (e.g., emotion) to comprehensive special knowledge as we continue our problem solving efforts. When a situation is over, we seal the problematic situation by means of the primitive conjectural knowledge (e.g., emotion such as pleasantness of confidence). By reloading the affective state right after a situation (e.g., expecting to be pleasant again), we reduce cognitive complexity that we generated from problem solving. We are trading meaning (confidence that we can handle such a problem) for complexity. In summary, our repertoire of expected knowledge moves to and fro along the general-to-specific continuum in corresponding to opening and closure of problematic situations. We travel a hierarchy of abstractness in conjectural knowledge from ?general? to ?specific? sequentially in a problematic situation. Entering and closing a situation. We not only start a problematic situation with a tradeoff between a meaning and complexity (e.g., unpleasantness), but also end it with a tradeoff between a meaning and complexity (e.g., pleasantness). Once we build a solution successfully, we seal the problematic situation again by trading cognitive complexity for a certain meaning (i.e., a simpler prototype state such as confidence). If we are satisfied with the new solution, we then confidently trade the complexity (i.e., the procedural knowledge of how to deal with such a problem) for a simpler affective state t knowledge. The general conjectural knowledge that often comes in the form of 129 such as restoring pleasantness or feeling competent in dealing with a problem. Ear Cvetkovich (1995) keenly captured our aspiration for achieving simplicity out of cognitive complexity in dealing with life problems: Men, endowed with limited human capacities, are troubled by the complex complex simple, that will give meaning to the complexity?trade meaning for distinction between a key and explanation. A key is meaning and requires no of reason. It is only the result?simplicity?that counts. The force of an evidence. [italics added] (1995, p. 34) We are always struggling to cope with complexity. After our problem-solving efforts, a solution, a procedure suggesting what to do, may result. Yet, it taxes our mental resources to maintain these details. Thus, epiphenomenal mental work is necessary to close or se l and environments into which they are thrown. They seek the key that will make the complexity?and thus allow them to forget it?. And Veyne stresses the crucial argument, no evidence. It works because it reduces complexity, not for any matter explanation, in contrast, depends on the quality of its supporting arguments and al a situation with endowed confidence about our newly acquired problem . otion f., te. Subsequently, we trace back what specifically triggers such a disruptive feeling. solving capability while letting one forget about technical details (complexity) In other words, the person now summarizes the details in problem solving by making it into affective knowledge. Summary. People go through the steps of trading affective knowledge for cognitive complexity from the opening of a problematic situation to its closure. Em (affective knowledge) enables one to detect what goes wrong as an interruption mechanism. With the perceptual mechanism of expectation, affective knowledge makes it possible to conduct prototypical hypothesis testing?i.e., a detection mechanism (c ?attention operator,? Simon, 1994). By an eruption of some emotion, such as uneasiness in a given moment, we recognize that something is falling outside of our expected sta 130 Additionally, emotion plays a role in cognitive closure of a problematic situation. It finalizes and seals the mental chasm. If one successfully learns and understands the details of what to do in a problematic situation (e.g., information), people then assign an affective summary about o e.g., confidence). After making effo ly updated expected state or a renewed o forts at problem solving (e.g., ch otion plays an indexing role tha bel of a given betwee nal conclu ng with that kind of problem (e.g., a eupho e straints, n (i.e., cognitive details) to construct a ne?s competence in problem solving ( rts to come to an end of a situation, the end state should be the new bserved state that reflects our ef anging conditions in and around the problem). Whereas a situation arises through an emotional signal, a problematic situation is resolved, again, with some emotional sealer. It is a compression of information that plays a functional role as a mnemonic device 15 and an interruptive mechanism that diverts our mental concentration to other emergent tasks?i.e., new problems. In other words, em t summarizes and sorts out cognitively complex information. It is a tag or la set of detailed information contained within a solution. Here, a tradeoff occurs n complicated cognitive contents of how to deal with a problem and an emotio sion that summarizes one?s competence in deali ric feeling such as pleasantness, high self-esteem, or confidence regarding the problem). Therefore, the closure of a problematic situation occurs with a tradeoff between cognitive complexity (information) and the relatively simple affective knowledge (emotion). However, in some situations, people cannot solve a problem because of sever constraints, despite their strong desire for problem resolution. Under such con they might skip the process of accruing informatio 15 Mnemonic device refers to ?an active, strategic kind of learning device or method, a rehearsal strategy? which provides an excellent means of retrieving the information (Ashcraft, 1994, p. 199). 131 solution s it, ge in on? J. antecedent situation or he finds himself in a new situation. The situation then ension in the individual and ?motivates? him to find a solution to his problem. t situation into a problematic situation, in that recognition of a problem is the first but turn instead to alternate fillers?i.e., negative emotion seeking. Wherea information helps to close the gap between expected and observed states by fixing appropriate new expectations and/or improving observational states, emotion can create a similar state of closure by filling up the gap with emotional content (e.g., anger). That is, emotion can become a solution of discrepancy that reduces the perceived gap from badness-of-fit between expected and observed states. In such a case, affective knowledge becomes a special form of information that is functionally equivalent to information about a situation. Thus, when one perceives a problem and feels a strong connection to yet also feels strong constraints against doing something about it, one may enga emotion seeking rather than information seeking (i.e., not by cognitive complexity). Definition of Problem and Situation For there is nothing either good or bad but thinking makes it so. Hamlet, William Shakespeare What is a problem? Are problems always negative? And what is a situati Grunig (1968) viewed a problematic situation as ?a function of a change in the situation? (p. 52). He explained environmental factors (e.g., economic resources), individual factors (e.g., psychological attributes), and their interactions as antecedent conditions of a situation. He explicated problem solving and situation as such: Problem solving behavior begins when a change occurs in the individual?s becomes lacking or indeterminant. The lacking situation creates a state of t Recognition of the existence of an indeterminacy converts the indeterminan step in solving it. (J. Grunig, 1968, pp. 32-34) 132 Problem recognition occurs right after an observation refutes the content of an expectation. The situational theory of publics defines problem recognition as ?detec that something should be done about a situation and stop to think about what to do? (J. Grunig, 1997, p. 10). To explain two new dependent variables of c ting ommunicant activen ent er, if re urgent problems (e.g., making te till ist moment, nothing can be perceived as change or difference and thus as problematic. In ess and cognitive entrepreneurship in problem solving, I will explicate in the following section conceptual meanings for problem, meta-problem, situation, problematic situation, and problem recognition. First, I confine problem to those only already perceived. It is of little use to extend the range of problem definition to something unperceivable to a decision-maker. Let?s take an example of poor minority residents who are exposed to high levels of lead in their houses, but who have not yet perceived how serious the situation is. Governm officials may initiate a massive health campaign to do something about it. Howev the most of the affected residents are heavily constrained by mo a daily living), the lead problem would not be a problem to the residents. Despi the danger, they do not perceive themselves to be in a problematic situation. They s feel happy and secure. The problem of lead contamination exists in the health communicator?s mind and not yet in the residents? minds. Putting aside the normative issue of the urgency of doing something about it, a problematic situation will not ex until the residents perceive its consequence themselves (e.g., illness from the lead). Human beings can perceive changes in their environments (a change in the situation) only when one has adopted a frame of reference that gauges degree of departure from it. Without a state to compare with what we psychologically carry at the 133 terms of the situational theory, people perceive something missing or sense a change in a situation when the a priori expectation they hold is not met as anticipated. They perceive a probl e of int in time when the problem solver reduces the perceived discrep w em, a difference or change, when they experience such a betrayal of content material on an expectation. I define problem as a perceptual discrepancy between expected and observed states to the unbearable extent of badness-of-fit. This is a perceptual problem in that it arises when an assumed hypothesis about a current state and perceived reality collides. In Popper?s (1963, 1999) terminology, a problem is the occurrence of which conjectural knowledge is refuted by reality. Problem detection is the moment that a living organism perceives the refuted conjectural knowledge. A perceptual problem triggers subsequent cognitive efforts to mend the discrepant caus conjectural knowledge or to change or deflect perceivable reality. In the past, we considered the term situation as an identical concept with the term problem. However, here I distinguish the terms situation or problematic situation to have exclusive connotation of a temporal period from the point of time at which one detects a meta-problem to that the po ancy by any means. 16 In other words, a situation refers to the temporal distance between the refuted conjectural knowledge and the establishment of an irrefutable ne conjecture?i.e., a period of recovery from badness-of-fit to goodness-of-fit between expected and observed states. Whereas a problem is about a perceptual aspect, a situation or a problematic situation is about a cognitive aspect. Although problems may arise, situations may not occur. I make this distinction because a situation starts only when a meta-problem occurs, not when a perceptual problem occurs. We deal with an almost infinite number of problems throughout our 16 The reduction of perceived discrepancy?i.e., problem solving?can be reached in several ways. 134 lives. At every moment, we encounter more than a manageable number of perceived discrepancies between our expected states and observable states. However, not all problems lead us to desperate situations that make us ?stop to think about what to We learn from past problematic situations and thus anticipate and carry a past solution fo similar problems. Because we have an inclination to develop solutions for problems we commonly enco do.? r unter, we attempt preconscious problem solving by applying an available past so e ., a , is a cogniti a f- r than concept, lution. In case we experience a special moment in which our solution cannot solv the problem, we experience a problem in dealing with a problem. This is a meta-problem. I define a meta-problem as a perceived discrepancy between expected states (e.g ?I think I have a solution for this kind of problem?) and observed state (e.g., ?I have no idea what to do for this problem?), specifically when the anticipated state is one?s belief that one has a readily accessible, available, and applicable new hypothesis pertaining to problem. This problem, the absence of a readymade solution to a problem at hand ve problem. When we fail to do preconscious problem solving and experience meta-problem, we then enter into a problematic situation. A problematic situation thus requires two prerequisites. One is problem detection and the other is meta-problem detection. The closure of a problematic situation occurs when one establishes an applicable solution and one can successfully effectuate it enough to restore goodness-o fit between an expected state and an observed state. I will thus limit my theoretical focus to problematic situations (i.e., only problems coming with a meta-problem) rathe all problems from now on. Analogy for problem and situation. To illustrate, I compare problem and situation with the notion of time and space. A problem is analogous to a geographic 135 space, pectation of a certain s A situation can end, but a problem continuously exists. A perceiver would thus termina e wishful ary ormation forwarder and information processor. Two commu ognition. nt (e.g., a whereas a situation is analogous to a time concept, period. As noted, the STP did not distinguish between problem and situation. Researchers have used the two concepts interchangeably and synonymously. However, I illustrate the differences between the two terms as follows. A problem is a mental space that is opened in a subjective field of experiential interest. This mental space takes up the area between the ex state and the observation of certain state in the perceiver?s mind. The chasm between the two states creates a temporal mental space. In other words, a problem is a mental space that is temporarily opened in a perceiver?s psychological field, which i defined as a domain of experience at a given moment. In contrast, a problematic situation is a subjective time period that is demarcated and exists when a meta-problem occurs. To close a mental chasm (problem), a perceiver makes efforts in narrowing the mental space. te the period of thinking of something as problematic by leaving the mental space (e.g., avoid thinking about the problem on which he or she feels little difference can b made). It may be through fixing the expected state (e.g., setting up a new expectation regarding the problem) or through subjectively deflecting the observed state (e.g., thinking or selective information seeking). In brief, problem recognition is a summ construct encompassing the opening of mental space over a certain time period. The field conception of a problem is useful in accounting for the differential contents of problems between inf nicants may be in the same field but existing with or without problem rec Perhaps the problem content for the information giver on some issue is differe 136 pick-up truck salesman under pressure to meet the minimum quota for monthly sales) from the problem content for the information taker (e.g., a browser who may buy the remote future but not immediately). The contents are different, but the field (subjective field of interest) is the same?that is, car buying. By the fit between the expected and observed states of each perceiver, a problem may exist on one side but not the other side (e.g., the car dealer is active but the customer is not active). Correspondingly, one may become very active in giving information, while another is not active in seeking information (e.g., potential customer is lukewarm bu salesman is desperate in promoting his car). Problems can often present differently even in the same field (e.g., car trade) by the function of fit between the individual perceiver?s a car in t the ding at s e - phase, there is no reason for it to garner more attention than positive information. content of expectation and observation implied by the expected content. Problem and negativity. We assume that negative events are easier to notice than positive events. For example, Slovic (1993) in his asymmetry principle of trust-buil and trust-destroying suggested that negative events carry much greater weight than positive events when events come to our attention. However, this is not necessarily so, least in problem detection. If some series of event history consists of a larger number of failures than successes, the failures (negative events) will be more visible. Likewise, once attention is drawn, negative information gets more weight in judgment than positive case do. However, we need to distinguish the salience of negative information between th attention phase and judgmental phase. Although, in the judgmental phase, negative information naturally carries more weight for its potential consequence, in the attention 137 Because attention depends on context, rather than on severity of consequence, th noticeability of negative and positive information at the attention-phase h e as no reason to differ. kely s at nt. We problem. Failure of our expectation does not mean a failure of our interest. e eived in ne sful and the other player feels defeated. Although the strategic move is perceived as a tive problem In other words, attention is not content-sensitive, but context-sensitive?how conspicuous it is among others. The way I define problem pertains to the attention phase rather than the judgmental phase. Hence, this frees the concept of problem from its content characteristics (e.g., negativity). Problem is a detection of ill fit between expected state and observed state. For problem detection, the content of expectations does not necessarily play a role in perceiving something as problematic. For example, if one receives notice from a doctor that one has a terminal illness, a negative problem is li to be perceived. However, if the poor person is again notified that the prognosis was in error, he must experience a pleasant surprise?that is, a positive problem. The person had detected problems similarly in either case, in that they betrayed his expectation the moment. His doom problem and delight problem were equally discrepa perceive any pleasant problem through an identical detection process as we do for any unpleasant In addition, what makes a problem from the attention phase continuously problematic in the judgmental phase is not the content of observation, but the content of expectation. To clarify, I offer the example of chess players. In a chess game, on player?s strategic and successful move against the competitor?s queen can be perc starkly different ways. Although two chess players share a single identical event, o player feels succes negative problem for the threatened player, it is not a nega 138 for the winning play e as a problem for one side, but not a problem for the other. A problem son, a f l model on rom . The last way is flight from the problem?i.e., leaving the pro t two mega er. Likewise, one team?s scoring of a touchdown in a football gam will likely be perceived is a subjective state if and only if it presumes, intentionally or unintentionally, consciously or unconsciously, some state of being within the situation. For that rea change of situation from normal to problematic happens because of the discrepancy o expectation from what the person observes now. To be precise, nothing can be necessarily considered problematic solely because of observed content. Problem-solving strategies. Generally speaking, the proposed hierarchica of problem detection suggests three strategies for dealing with problems once a pers perceives a psychological distance between the expected and observed states in one?s domain of experience. A problem solver?s strategies in dealing with a problem vary f 1) revision strategy, 2) deflection strategy, to 3) flight strategy. The first strategy is to remedy the contents of the expectation?i.e., revision of expectation content (e.g., lower one?s aspiration). The next strategy is to fix the contents of the observation?i.e., changing or deflecting the observed content (e.g., being selective, such as avoiding discrepancy-causing information) blem field (e.g., avoiding thinking about the problematic situation). The firs approaches directly intervene into either of the sources that cause the discrepancy. Specifically, one may look for revising or reinforcing information through communication behaviors by taking a certain cognitive strategy (e.g., cognitive o strategy). A problem solver?s subjective perception about the problematic situation will influence what kind of problem-solving strategy one would subsequently take. Thus, it is 139 necessary to inquire what kind of perceptual antecedents influence the choice of subsequent problem-solving strategies. They are problem recognition, constraint recogni hesion or collision between the contents of expecta is the to what extent one is competent this kind. d tion, level of involvement, referent criterion, and situational motivation in problem solving. Antecedent Parameters of Communicant Activeness and Cognitive Entrepreneurship in Problem Solving Problem Recognition No problem exists until we recognize it. We do not discover a problem external to us. It is born and lives inside of us. Thus, a problem is not independent of our individuality in thinking. A problem is a joint product of our mind-working and the perceived world in which we reside. The interaction between what we expect and what we observe jointly creates a problem. The size and characteristics of a new-born problem are solely defined by the extent of co tion and observation. I have explicated the differences between perceptual problem (problem) and cognitive problem (meta-problem). A perceptual problem early and necessary part of a problematic situation that mainly involves the attention phase in a problematic situation. In contrast, a cognitive problem is the latter part of a problematic situation. It primarily concerns the judgmental phase of what to do about the perceived psychological discrepancy. The judgmental phase concerns the perceiver?s evaluation of what caused it, how it can be resolved, and in problem solving of Definition. The situational theory of publics adopts a definition of problem recognition as when ?people detect that something should be done about a situation an 140 stop to think about what to do? (J. Grunig, 1997, p. 10). In the previous section, I conceptually illustrated how we detect something as an unbearably poor fit by a hierarchical model of problem detection. Following the introduced concepts, I define problem recognition as one?s perception that there is something missing and there is n immediately applicable solution to it. It is, thus, a meta-problem following a perceptua problem?i.e., a perceptual state one experiences after the failure of preconscious problem solving. It is not only a perceptual problem, but also a cognitive problem in that one recognizes something discrepant and experiences the absence of a handy solution. A person who perceiv o l es a problem but feels incapable of finding an immediate solution (i.e., a mech ived ditional anism to narrow the perceived psychological discrepancy) then enters into a problematic situation. As a result, the person is likely to stop one?s current routines to think about a solution. However, it increases only the probability of ?stop to think about what to do,? not determine it. There are other factors such as the extent of perce connection to the problem and perceived obstacles in doing something. These ad factors jointly influence whether one ?stops? things one is doing ?to think about what to do.? In other words, a person may or may not stop to think about what to do even with high level of problem recognition To conceptualize problem recognition, thus, we need to incorporate two aspects of a problem simultaneously?the perceptual problem and the cognitive problem (attention phase and judgmental phase). Without a precedent perceptual problem, we never feel a modicum of reason to ?stop to think about what to do? (J. Grunig, 1997, p. 10). Our motivation for stopping our routines to think about what to do presumes t a hat we 141 recognize what is currently missing: detect a discrepancy between what we expect a what we observe nd and experience absence of a solution. Level of Involvement Communication and marketing research heavily uses the concept of involvement (Salmon, 1986). The involvement concept demonstrates its utility in segmenting and discriminating between people with varying levels of involvedness. The level of involvement correlates with the differential behaviors regarding products, issues, or problem solving behaviors (J. Grunig, 1989). Specifically, when we know one?s le involvement, communicators can predict how an individual would behave differently regarding problems, products, and ideas. Singling out a cohesive but distinct group of people from a general population almost always rewards communicators with its economy, efficiency, and effectiveness in pu vel of rsuing communication objectives. However, there is f in a d involvement as a m? (1965) and Rothschild and Ray (1974) saw it as a charact much confusion that arises from overuse of the concept among communication researchers. Therefore, I next delimit the conceptual meaning of involvement adopted in the present study. Definition. Lovelock and Weinberg (1984) gave a ?common-sense? definition o involvement as ?degree of importance or concern? that a product or behavior generates in different individuals (p. 73). Previously, the involvement concept carried meaning non-perceptual way. For example, Krugman (1965) define characteristic of a ?mediu eristic of a ?product.? However, J. Grunig (1976) defined it as a ?perception? that people come to have within a given situation. He defined level of involvement as ?the extent to which people connect themselves with a situation? (J. Grunig, 1997, p. 10). 142 People enact their communicative action by their perceived connection?involveme of self to the problematic situation. When their perceived connection is low, they are likely to be passive in communication behavior. When they perceive a close connection, they are likely to be active in communication behavior. Thus, it is better to capture th notion of a perceptual variable of involvement, rather than capturing the notion of triggering variables of perception. Even if the actual connection of some events or problems to us is important, we will not initiate behavior to do something about the problem until we consciously perceive a connection. nt? e er e , ysical gnition. It is only after he recognizes his health problem that he rtant this problem is to him (i.e., a perceived connection). Thus, to underst ant Without awareness of an existing problem, a situation will never start. We nev feel something needs to be done without a preceding perception of something missing. Knowing what is missing is a logical a priori step to estimating how closely we are connected with it. We should see there is a significant leap from one?s actual connection to one?s perceived connection. We do not live in an objective world, but live in a perceived world. We construct our reality subjectively, not taking a single standard version of reality. We thus translate our perception of the world, not the world itself. W cannot do anything at all before we happen upon this perception. Actual connection is thus, different from perceived connection. A person who is not aware that he is terminally ill (i.e., actual connection to the health problem) will not do anything about it until he finds abnormal signs of his ph condition?problem reco evaluates how impo and, explain, and predict subsequent problem-solving efforts (e.g., communic 143 activeness), we need to delimit our conceptual scope to what we perceive as being connected rather than to what we are actually connected. Constraint Recognition Unlike the concept of involvement and problem recognition, few commu and marketing theories use a concept like constraint recognition explicitly (J. Grunig 1989). Constraint recognition is one of two original conceptual variables developed b Grunig (1968) in the earlier version of the situational theory of publics. Constraint recognition has its origin from economics and management science rather than fr nication , y J. om psycho of cons linear p constraints of resources available to a decision maker (J. Grunig, 1968, 1989). Later in social psychology, Bandura (1977) proposed ?personal efficacy? in his social learning theory, which is a very close concept to J. Grunig?s (1968) constraint recognition. Definition. The situational theory of publics defines constraint recognition as when ?people perceive that there are obstacles in a situation that limit their ability to do anything about the situation? (J. Grunig, 1997, p. 10). From studies about large land owners and peasants in Colombia, J. Grunig (1969, 1971) found that ?people have little need to communicate in situations where constraints prevent people from making choices? (J. Grunig, 1997, p. 10). Constraint recognition discourages communication behavior such as information seeking and processing even if communicants have high problem recognition and/or level of perceived involvement. As noted by J. Grunig?s (1969, 1971, 1972) studies of Colombian peasants and landowners, people are less likely logy, unlike many variables in communication theories. A close parallel concept traint recognition from economics and management is a discounting factor in rogramming, a statistical process that can be used to maximize profits within the 144 to communicate about ?problems or issues about which they believe they can do little about behaviors they do not believe they have the personal efficacy to execute? (J. Grunig, 1989, p. 212). Referent Criterion People approach their problems by recalling relevant experiences of success similar to a current problem. If their search for prior experience is fruitless, they will tur to external sources for a so or n lution. Originally, J. Grunig (1968) described a referent criterio iteria y lead to a generalized query of the following kind: ?what criteria can I discover experienced fire fighter asks, ?Are there any underlying principles of fire fighting ] (p. 97) Thus, J. Grunig (1997) formally defined a referent criterion as ?a solution carried from previous situations to a new situation? (p. 11). We can also construe referent criterion as a cognitive ?schema? and ?cross-situational attitude? to those bits of cognitive and attitudinal knowledge that guide problem-solving and decision-making (J. Grunig, 1997). It generally reduces the need for a problem solver to search for additional information. However, the referent criterion as an independent variable has not been a n as a ?gross criterion? or ?general guide? in which other more specific cr will be required to fit (e.g., maximum profits, maximum sales, and survival of the organization) (pp. 27-28). According to J. Grunig (1968), a referent criterion is ?determined by the antecedent condition, especially from the social contacts of the individual and from his past behavior which has partially determined the antecedent conditions? (p. 27). J. Grunig (1968) illustrated the need of a referent in repeated problems by Simon?s (1957) explanation. Simon (1957) explained such a need: When a problem of a particular kind has several times arisen for decision, it ma which can be used whenever a problem of this kind arises?? For example, the which can be applied to the many fire situations with which I deal?? [italics added 145 good predictor for communication behavior as intended (e.g., J. Grunig & Disbrow, 1977). J. Grunig (1997) thus dropped it from the situational theory of publics. He omitted it because he concluded that referent criteria would be ?more of an effect of communication than a cause? (J. Grunig, 1997, p. 11). Although dropping the referent criterion has conceptual and empirical merit find that the notion has some conceptual utility in explaining the new variables in this dissertation. In the earlier chapter on communicant activeness in problem solving, I developed a set of new dimensions of communicant activeness such as information forwarding, information sharing, information forefending, and information permitting. I reason that the referent criterion a solution carries from previous situations would e little variance for information seeking and processing empirically. However, I reason that it may explain and predict the dimension of information transmission and select s, I xplain ion variabl will refine its conceptual meaning in detail. es to some extent. For example, we can run four more regression equations beyond the two conventional regression equations for information seeking and information processing (J. Grunig, 1997, p. 12). Should a referent criterion?a solution toward a given problem? be available to a communicant, we can logically predict that she or he would forward or share it with others, would forefend more and permit less in taking and giving newly available information. If the dropped variable, referent criterion, removes some variance in explaining those dependent variables of communicant behavior, keeping it out of situational theory should be a loss, not only in the empirical sense but also in terms of theoretical validity. In the following section, I 146 Definition. Although it has some perceptual aspect, referent criterion is closer to cognition because it conceptually taps and measures the available knowledge and inferential rules from one?s prior problem-solving experiences. I define referent criterio as any knowledge or subjective judgmental system that exerts specific influence on the way one approaches problem solving. This can include decisional guidelines or decision rules perceived as relevant to a given problem. Problem solvers bring them over from prior problematic situations. However, I discriminate between decisional referents on- duty?that is, applicable and workable referents of which a problem solver is aware?a off-duty ones?those which are applicable and workable but not yet recognized for th value to the given problem-solving situation. Even if we store some useful knowledge decisional referents in our memory, unless they are available and evaluated for their workability within a given problem, they are of no use. By the success of internal searching for retrievable solutions or pieces of knowledge to construct a new solution, one?s degr n nd eir or ee of cognitive entrepreneurship and communicant activeness in problem ecyclable and workable referent criterion, he or she i ke an the te in as wishful solving varies. If a problem holder finds a r s less likely to have an entrepreneurial mindset and less eager to seek for information in dealing with a current problem. In contrast, if one has difficulty in retrieving a workable solution from internal storage, then one is more likely to ta entrepreneurial mindset and communicant activeness to compose a novel solution. In addition to the knowledge aspect of a referent criterion, I now include presence and extent of wishful thinking and/or willful thinking toward an end sta problem solving. Generally speaking, such self-fulfilling decisional referents thinking or willful thinking lower one?s problem solving effectiveness because of their 147 tendency to result in misdiagnosis of problem characteristics and self-fulfilling solution building and evaluation. Once a problem holder retrieves such a self-fulfilling referen (e.g., a goal, a desire, or a preference), this will strongly influence the interpretations and selection of the data encountered during problem solving. The stronger presence of suc self-fulfilling decisional referents t h will result in more information forefending and a less entrepr matic ers e of l erceptual variables and the referent criterion one activates after detecting a eneurial approach in problem solving. In summary, problem solvers do carry referent criteria from previous proble situations. Or they instantly improvise by configuring available knowledge, experience, or judgmental rules derived from similar problems. In the latter case, problem solv often use subjective decisional referents such as wishful or willful thinking about the end result. However, irrespective of the differential contribution to problem solving effectiveness, any decisional referent is functionally identical in that problem solvers deploy it to gain closure on a problematic situation. They become a major premis syllogistic reasoning in a given decision-making or problem-solving scenario. In case one adopts a more self-fulfilling referent (e.g., a terminally ill patient may think my illness must not be that serious), subsequent choice and drawing of the minor premise would be a more self-fulfilling one that corresponds with the subjective major premise components (e.g., sampling of observational reality that is more compatible with self- fulfilling referent: ?Look, most of my physical conditions are not different from norma people?). Situational Motivation in Problem Solving Situational theory offers a means to predict publics? differential responses with its three p 148 problem. One?s perception is subjective to the individual perceivers (individual differences with the same perceptual object and event), situational across time periods (they dissipate and no longer exist after problem resolution), and antecedent to motivation (may or may not do something about the perceived state), cognitive processing (one may or may not think further about the perceived state), and communication behaviors (one may or may not seek, forward, and forefend informatio People act on their perceptions, whereas motivation and cognition (i.e., referent crit are enacted by the perceptions. We can say that the perception of a problematic state, perceived capacity or capability regarding problem, and perceived connectedness are t prime movers at least in our mind that trigger subsequent adaptive behaviors about the perceived states (i.e., problems). Source of redundancy in problem recognition and level of involvement. J. G (1997) has defined problem recognition as a situation wherein ?people detect that something should be done about a situation and stop to think about what to do? (p. 10). He defines level of involvement as the perceived ?extent to which people connect themselves with a situation? (J. Grunig, 1997, p. 10). The measurement item for problem recognition is ?how often do you stop to think about? the issue; whereas for level of involvement the measure is ?to what extent do you see a connection between yourself, personally, and each of these situations? (J. G n). erion) he runig runig, 1997, pp. 45-47). These two constructs are com mmunication plementary but independent in predicting different co behaviors. However, we need to take a finer look at how we come to ?stop to think.? Our perceptual mechanism detects countless ?discrepancies? or cases of ?something is lacking? routinely. Whenever we detect a lack, we instantly judge whether and how it 149 would affect (i.e., connect to) our current or future state of being (e.g., ?Will this ruin m interest??). By instant judgment (i.e., preconscious problem solving), we tend to keep only a handful of ?lacking situa y tions? out of the countless ?something missing situatio missing olecule rtain state as problematic when our current expectation about things turns nitial point at which we experience the existence of some problem. J. Grunig thus explained that ?proble tarily y other situatio of a problematic situation (detect), the perceived connection close enough to keep our mind ns.? The resulting ?stop to think about? state is, thus, an end state from our judgment about relevance?to the extent of connection between ourselves and a state. Even if we detect a problem, unless it affects us significantly, we will soon leave the situation. There is a sequential cognitive process that people lead themselves to a certain communicative behavior. To illustrate, we need to look at the behavioral m that dissects the developmental stages for the human decision-making process. We perceive a ce out discrepant from it. The detect of the behavioral molecule is thus an i m recognition represents the detect segment of the behavioral molecule?people do not stop to think about a situation unless they perceive that something needs to be done to improve the situation? (J. Grunig & Hunt, 1984, p. 149). Even if we momen ?stop to think about? something that has a close connection, we will not continue to pa cognitive taxes in sustaining us to keep us ?stopping to think.? We are under nal demands from more closely connected problems. We are cognitive investors who selectively distribute our limited cognitive resources in terms of some prioritization principle. ?Stop to think about? is thus the outcome of a cognitively active state resulting from a joint function of three perceptual prerequisite conditions: the internal presence 150 alive to think about the problem, and the perceived obstacles prohibiting us from do something about the detected missing state: ing Detecti Both de stop to think? ?percei meritor dditionally, three of the independent variables in the situational theory are perceptual variables such as ?perceived lacking,? ?perceived connection,? and ?perceived obstacles that limit one?s ability to do anything.? To enhance the conceptual coherence among the independent variables, the situational theory needs to eliminate the ?stop to think? notion from its conceptual definition of problem recognition. Again, one?s state of stop to think is not perceptual, but motivational. The following diagram illustrates the conceptual sequence between situational perception and situational motivation. Figure 9. Situational motivation and perceptual antecedent conditions. Stop to think = f (detection, connection, constraint). on that something is missing in a situation alone is not sufficient to stop to think. tection and connection are two perceptual necessary conditions for ?the state of think? about the discrepancy further. Putting it differently, when people ?stop to about something then there would be a sufficient level of ?perceived missing,? ved connection,? and ?perceived lack of obstacles? that make cognitive efforts ious to the perceivers. A 151 Three perceptual variables jointly predict one?s ?heightened cognitive readin and ?augmented epistemic motivation? as cognitive and motivational outcomes during a problematic situation. The situational perceptions and the epistemic motivation are confounded in the current definition and operationalization of problem recognition. It contains a built-in confoundedness that is tapping the level of involvement. We are cognitive economizers who selectively ?stop to think? about only those problems that are closely connected with us. For this reason, ?the extent to which people connect themselves with a situation? is already incorporated into the concept of problem recognition. As a natural consequence, the questions used to measure two constructs, problem recognition and level of involvement, become conceptual Siamese twins; measuring one thing unavoidably taps the other. To summarize: People do not stop to think unless they perceive that something is lacking in situation (problem recognition), unless they perceive that their connection the lacking state is causing some threats or opportunities, and unless they percei ess? the with ve ink r ion. Summary In this section, I reviewed and elaborated key independent variables to account for communicant activeness and cognitive entrepreneurship in problem solving. I first explained the concepts of problem and situation by introducing conjectural knowledge and its testing?i.e., expectation and observation of a living organism. In so doing, I a lack of barriers that prohibit one?s problem-solving efforts. Thus, stop to th about tendency is defined as a situational motivation?i.e., a situational need fo cognitive working to fill out the discrepancy between expected and observed problematic states and to improve the problematic situat 152 discussed the role of emotion (e.g., knotting) in managing multiple cognitive tasks. That is, a piece of general and abstract affective knowledge (e.g., pleasantness) becomes the content a ext, I reviewed and refined four situational antecedent variables from the situational theory ecognition, level of invo that there is ly applicable solution to it. This reworks the pre ink? of a conjectural hypothesis to be tested by observation. I called this prototype hypothesis testing. It gives a supervisory summary of many mental and physical tasks at given moment. Once we experience a refutation in prototypic hypothesis testing, we subsequently pursue more specific knowledge to explain what happens, why it happens, and how something should be handled. Entering and closing a situation is thus signified with conjectural affective knowledge that trades cognitive complexity for simpler affective meaning. N of publics (J. Grunig, 1989, 1997). They are problem r lvement, constraint recognition, and referent criterion. I distinguished problems as emergent in two different phases: perceptual (attention phase) and cognitive (judgmental phase) problems. I refined problem recognition to be one?s perception something missing and there is no immediate vious definition of problem recognition, ?people detect that something should be done about a situation and stop to think about what to do? (italics added, J. Grunig, 1997, p. 10). Although one detects discrepancies between expectation and observation, not all detected discrepancies make one ?stop to think? about what to do. One?s ?stop to th tendency is the outcome of situational perceptions such as detection, connection, and constraints. Thus, I remove the notion of ?stop to think about what to do? from the definition. Conceptual refinement resolves the multicollinearity problem among situational antecedent variables (Kim, Downie, & De Stefano, 2005; J. Grunig, 1997). 153 I also delimited involvement to a perceptual scope?that is, only perceived connection, not an objective or actual connection that is not yet perceived. Constraint recognition refers to any perceived obstacle or barrier that a problem solver perceives in making efforts to gain closure on one?s problematic situation. situa obje r subj hinking on an end state after problem solving) that r as lem tasks. I argued that the motivation toward fluctuates across problems. S tecedents such constraint recognition fect the extent problem solving. I hese t variables with the new dependent variables (i.e., CAPS and endent and Independent Variables he antecedent and consequent variables that have been developed thus far. These integrated variables consist of a new version of the situational lving Next, I reintroduced a dropped variable, referent criterion, as a potentially useful tional antecedent parameter. I modified its conceptual meaning to include any ctive knowledge (i.e., a solution carried from past success in problem solving) and/o ective beliefs (i.e., wishful or willful t becomes a decisional referent. Decisional referents could be any knowledge as fa they are perceived as relevant to problem solving. Finally, I introduced a motivation variable, situational motivation in prob solving. People differ across problems whether they are willing to invest their cognitive capacity and capability to the problem-solving problem solving pecifically, the situational an as problem recognition, , and level of involvement af of situational motivation in n the next section, I will integrate t refined situational anteceden CEPS). Integration of Dep In this section, I integrate t theory of publics. I call this a situational theory of problem solving. The Situational Theory of Problem So 154 Up to this point, I have taken a backward approach to building a situational theory of problem solving. In the first two sections of this conceptualization, I tried to establish two dependent variables. Dependent variables are the most important and often the most problematic variables in any theory. Finding or establishing a dependent variable therefore means problem recognition to the theorist. After constructing two dependent variables or two focal phenomena of interest, I attempted to construct the how and why? i.e., independent variables?of the two phenomena: communicant activeness in problem solving and cognitive entrepreneurship in problem solving. To summarize, I put them into a single theoretical framework. One good way to summarize a theory is to specify a mathematical system of equations with key variables. J. Grunig (1968) noted the utility of such a formal system and the limitation of estimating the mathematical relationships among variables at his original theorizing of situational theory. He (1968) wrote, The system is presented here primarily for its summary and general summary power. Some day it may be possible to determine the exact functional nature of the alue. In this study we are primarily concerned with determining whether the equations are independent variables?as measured for Colombian latifundistas?have a positive, negative, or neutral effect on the aches. Because of this, researchers can propos the mathematical relationships; then the model would have tremendous predictive v valid in their present crude form, and, if so, whether the dependent variables. (pp. 50-51) Since the time the original situational theory was proposed, theorists have made impressive progress in methodological appro e and test mathematical systems of equations to test the viability of their theoretical propositions fairly easily. I here summarize the variables and systems of situational theory of problem solving. 155 S = Situation; OS = Observed State; PP = Percep PR = Problem Recognition; LI = Level of Involvement; ES = Expected State; t = time, one given process; tual Problem; CP = Cognitive Problem; CR = Constraint Recognition; SM = Situational Motivation; CE = Cognitive Entrepreneurship CA = Communicant Activeness ws: 1. PP = f (dSt/dt) = f (ES-OS) 2. CP = f (ES-OS | AS) = PR 4. - AS = Absence of Solution; The system is as follo 3. SM = f (PR, CR, LI | RC) CE = f (SM) 5. CA = f (SM) I illustrate the sequential process from situational perception to subsequent problem solving acts in the following figure. Figure 10: Sequence illustration of the situational theory of problem solving. 156 Situational Theory of Communicant Activeness in Problem Solving and Situational Theory of Cognitive Entrepreneurship in Problem Solving We can break down the structural model of the situational theory of problem solving to highlight each dependent variable. Each derived model can stand on its own as an independent theory. They are as follows: A model of the situational theory of communicant activeness in problem solving (SITCAPS). The first dependent variable I have thus far elaborated is communicant activeness in problem solving. It takes six subdimensions related with information selection, transmission, and acquisition. Four situational antecedent variables explain the likelihood of a problem solver?s communicative activeness regarding a given problem. nition, The following model summarizes the valences of causal paths from problem recog level of involvement, constraint recognition, and referent criterion to communicant activeness. Figure 11: A model of the situational theory of communicant activeness in problem solving (SITCAPS). 157 H11: The higher the problem recognition, the higher the situational motivation in H12: The higher the constraint problem solving. recognition, the lower the situational motivation in problem solving. H13: T problem solving. roblem solving. communicant activeness in problem solving. A model of the situational theory of cognitive entrepreneurship in problem solving (SITCEPS). The second dependent variable I introduced was cognitive entrepreneurship in problem solving. This construct has four subdimensions: cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspension. The following model summarizes the theoretical relationships between four situational independent variables and cognitive entrepreneurship as a dependent variable. Figure 12: A model of the situational theory of cognitive entrepreneurship in problem solving (SITCEPS). he higher the level of involvement, the higher the situational motivation in H14: The higher the referent criterion, the higher the communicant activeness in p H15: The higher the situational motivation in problem solving, the higher the 158 H16: The higher the problem recognition, the higher the situational motivation in H17: The higher the constraint recognition, the lower the situational motivation in problem solving. H18: The higher the level of involvement, the higher the situational motivation in problem solving. H19: The higher the referent criterion, the lower the cognitive entrepreneurship in problem solving. H20: The higher the situational motivation in problem solving, the higher the cognitive entrepreneurship in problem solving. SITCAPS and SITCEPS set the CAPS and CEPS as dependent variables explained by situational antecedent variables. Both models consist of a family of the situational theory of problem solving. However, it is interesting to integrate two models into one by setting CAPS and CEPS as dependent variables simultaneously. However, from the empirical model testing perspective (structural equation modeling), combining the two d es ahead unicative behavior, and CEPS, featuring the cogniti est relative problem solving. models into one will create a model identification problem. Both CAPS an CEPS constructs are accounted for by the same situational antecedent variables. When merging the two models of SITCAPS and SITCEPS, the resulting model becomes just- identified or under-identified (if specifying bidirectional causality between CAPS and CEPS) and thus not testable. Despite testing difficulty, it is interesting to speculate which variable com of the other. CAPS, featuring comm ve approach in problem solving, seem to explain each other to some degree. However, it is hard to define which comes first. A plausible relationship is bidirectional causality between the two constructs. This requires a nonrecursive model to t effects from each other simultaneously. Such a nonrecursive model testing is desirable but often methodologically difficult (e.g., the non-converging issue in model solution). 159 To test simultaneous causal relationships between CAPS and CEPS, I construc the model in Figures 13 and 14. As noted, a nonrecursive model is hard to solve empirically. To ide ted ntify the ch an exogenous variable to identify the model. Surely, it is arguable that a referent criterion affects both CAPS and CEPS. However, as noted, simultaneous causal paths from referent criterion to both CAPS and CEPS as endogenous variables will not be identified. The primary interest in this test is how CAPS and CEPS would affect each other. Thus, for the model identification, I will specify that the referent criterion affects one over the other. Figure 13 and 14 are two possible models (N. B., both models contain the same hypotheses and predictions.). I will analyze both models and will compare the results. Figure 13: A model of a nonrecursive relationship between CAPS and CEPS (Model I). bidirectional causality model, it is necessary to specify one exogenous variable as an exclusive cause for one of the endogenous variables. I specify the referent criterion as su 160 Figure 14: A model of a nonrecursive relationship between CAPS and CEPS (Model II). A problem solver with high cognitive entrepreneurship tends to be high in communicant activeness. Particularly, someone with high cognitive entrepreneurship is more likely to be more enthusiastic in broadening candidate solutions (more inform seeking), more tolerant of competing or incompatible ideas (less information forefend and more willing to disperse knowledge and information to others (m ation ing), ore information forward ess kely to out ing). Thus, I expect a positive effect from cognitive entrepreneurship to communicant activeness. In contrast, I expect a problem solver with high communicant activeness to be l entrepreneurial in problem solving. As a result of heightened communicant activeness in problem solving, a problem solver will be more selective, transmissive, and knowledgeable about the problem. Thus, the problem solver is likely to be more li optimize his or her favored solution (high cognitive retrogression), show less tolerance for incompatible solutions (less cognitive multilateralism), be less enthusiastic ab 161 available solutions and ideas (less cognitive commitment), and be less heedful in evaluating new ideas and candidate solutions (less cognitive suspension). H21: The higher the cognitive entrepreneurship, the higher the communicant activeness H22: The higher the communicant activeness, the lower the cognitive entrepreneurship in in problem solving. problem solving. Summary. I claim that the situational theory of problem solving (STOPS) is a more general theory on the basis of its generality in subsuming variables of STP. The STOPS with the CAPS model takes a further step. This study challenges the traditional sender/receiver dichotomy by treating communication as both a dependent and independent variable. For example, a business communicator prepares advertising campaigns for promoting a new product (information transmission). Here, the message sender not only initiates a communication process to solve a problem (i.e., information forwarding to increase sales) but also should learn about products and target segments of the population (e.g., information seeking to prepare messages). In contrast, a father with an ill child vigorously seeks information about treatment of that disease. Here, he initiates a communication process (information seeking to learn about the disease) to cope with his life problem. Often, such accumulated knowledge for a specific problem leads a communicant to share or transmit information to others (information forwarding with forefending). Communication is a coping mechanism that increases problem- solving potential for problem holders. The situational theory of problem solving accommodates both communicative directions and explains ill-captured aspects of informational giving with its general conceptualization of communicant behaviors. Furthermore, the situational theory of problem solving introduces a new variable, cognitive entrepreneurship in problem solving. With this new construct, we can better 162 understand how people me t times, we have a more entrepr ew on approac ith cant ntally approach their life problems. A eneurial mindset. We take in information to evaluate by its merits before selecting a solution. We value and tolerate incompatible ideas, become enthusiastic about n ideas, and are eager to finalize a solution (i.e., cognitive alpha approach). At other times, we are less entrepreneurial and use information not to evaluate, but to justify a decisi that we have already hastily drawn. We value more ideas that are incompatible with a decision we made. We are less committed to new ideas, and we want to be firm and conclusive as fast as possible (i.e., cognitive omega approach). However, our mental h should be best viewed as rational when we adopt a problem-solving strategy in a more dexterous way. We simply become irrational when we are so entrepreneurial w small problems (e.g., low-involvement problem such as what to have for lunch). Likewise, we become irrational when we are recklessly non-entrepreneurial with big problems (e.g., highly involved problems such as a risky behavior). In this chapter, I first proposed and explicated two new concepts of communi activeness and cognitive entrepreneurship. Then, I refined the situational antecedent variables that affect our communicative and cognitive approaches in problem solving. These are integrated and referred to as a situational theory of problem solving. The new situational theory consists of two models, the situational theory of communicant activeness in problem solving (SITCAPS) and the situational theory of cognitive entrepreneurship in problem solving (SITCEPS). The situational theory of publics (J. Grunig, 1997) now becomes a more general theory that describes unique human communicative and cognitive features in dealing with life problems. I expect the new situational theory to host further normative theorizing in the areas related with human 163 problem solving. In the following chapter, I will discuss how I designed and conducted the tests for the new theory and constructs. In addition, I will discuss the methodologica and ethical consideration in the study. l 164 CHAPTER III: METHODOLOGY In the previous chapter, I first conceptualized two new variables of communicant activeness and cognitive entrepreneurship in problem solving. They become the dependent variables of the situational theory of problem solving. Next, I elaborated the independent variables from the situational theory of publics. Integration of two new dependent variables with the independent variables of the STP introduces a situational theory of problem solving. This theory consists of a situational theory of communicant activen are s; he fically, I neurship in problem solving as sive model of the situational theory of problem ess and a situational theory of cognitive entrepreneurship in problem solving. The new situational theory includes a set of propositions and hypotheses that subject to tests. Empirical tests can never prove any theoretical claims, they only add support for the inferred plausibility of the theory. Therefore, as the first step of testing, I generated measurement items for each conceptual dimension in communicant activenes cognitive entrepreneurship in problem solving; and problem recognition, level of involvement, constraint recognition, referent criterion, and situational motivation in problem solving. Conceptual explications and definitions of each variable in earlier chapters guided the operationalization process (i.e., generating item pools). After t operationalization, the derived measurement scales were submitted to analytic procedures of item analysis and the assessment of reliability and validity studies. Given that those measurement systems are reliable and valid for each key variable, I proceeded to test causal path models derived from the situational theory of problem solving. Speci first tested the causal paths from situational antecedent variables to communicant activeness in problem solving and cognitive entrepre separate models. Then, I tested a nonrecur 165 solving d 3). something unobservable. Typically, survey rates, when p h an explana ed (Weisb number of participants (e.g., explanatory analyses), surveys make large samples feasible. In the present study, the purpose was to test a theory in which many variables need to be wherein two new variables specified as bidirectional each causes each other. In the following sections, I will discuss the choice of research method (survey) to test the models, the data collection approach such as sampling strategy (non-probability sampling), estimation of sample size and power (Monte Carlo simulation), data collection procedure, data analysis procedures (CFA and SEM), and ethical considerations in the study. Survey Research A survey is ?systematic data collection about a sample drawn from a specifie larger population? by using questions (Schwarz, Groves, & Schuman, 1998, p. 14 Researchers use a sample and questions to make inferences about the population and concepts of interest. Thus, researchers sample populations and operationalize a concept instrumentally to infer from something observable to research consists of 1) setting research objectives for information collection, 2) designing the study, 3) preparing a reliable and valid survey instrument, 4) administering the survey, 5) managing data collection (e.g., making efforts to attain high response editing and coding), 6) analyzing survey data, and 7) reporting the results (Fink, 2003). Strength and Weakness of Survey Research Surveys generally are used to describe the characteristics of a large population robability sampling is used. Furthermore, surveys work well for a study wit tory purpose, such as this study, even though nonprobability sampling is us erg, Krosnick, & Bowen, 1996). When the research purpose requires a large 166 analyze ata f d valid le selection and size, researchers have more ability ads 8). ting. fy and ibility and makes it impossible for researchers to change their initial design and research questions survey makes serendipity h is only feasible when there are interactions between the researcher and particip d simultaneously. Thus, survey research design was an appropriate choice for d collection. Surveys are flexible enough to allow researchers to operationalize definitions from actual observations in a backward manner. Such flexibility is hard to attain in experimental design in which an operational definition is a manipulation of a variable o interest. If the measurement items are well established (i.e., they are reliable an measurement scales) with an adequate samp to generalize. However, survey research requires the standardization of questions, which le toward superficial results. Standardized measurement items often measure the least common denominator among the participants? characteristics of interest (Babbie, 199 Survey research is inflexible and blind to natural processes inherent in a research set Qualitative research methods not only allow and encourage a researcher to modi elaborate theory, questions, and design; but they also provide in-depth contextual information (McCracken, 1988; Marshall & Rossman, 1995). In contrast, survey research cannot provide such flex (Babbie, 1998). A impossible, whic ants. Finally, survey research cannot directly measure social action but can only measure the self reports of past action or hypothetical action (e.g., behavioral intention). For that reason, the data collected are often artificial and could misrepresent the participants? characteristics of research interest. This, in turn, makes it difficult to establish the validity of measures (Babbie, 1998). Rationale for Choosing Survey Research 167 I have chosen the survey research method as the most appropriate approach the purposes and specific constraints inherent in this study. I chose the survey me because the purpose of the study was to test causal structures and hypotheses derived from a situational theory of problem solving. If I could follow the necessary procedure (e.g., random assignment) and selection of design frame (e.g., a for thod s Solomon four-group design) a and Bowen sented the advantage of survey research for such a study: be obtained by a survey. You cannot assume that people think in certain ways equivalent to individual data, nor can you use experiments as alternatives to the questions, you can gain much information about what they are thinking?and why play important roles in social science. (p. 20) To measure publics? perceptions, cognitive approaches, and communicative behaviors, therefore, survey research was a most appropriate data collection method. , the experimental method would be the best research design to study causal propositions among the constructs of interest. However, the number of causal variables in the systems of equations was more than five. This unnecessarily complicates the design and procedure (e.g., manipulations for independent variables) beyond the researcher?s control. Alternatively, survey research design allows the study of multiple variables economically. A survey study can yield data of adequate size and quality for causal analysis (i.e., structural equation modeling). Secondly, the present study aimed to explain publics? perceptions of their life problems and their subsequent cognitive and communicative features in dealing with the problems. Thus, asking publics for their own perceptions about their problems was critical to testing the viability of the proposed theory. Weisberg, Krosnick, (1996) cogently pre The explanation of mass behavior often requires mass attitude data that can only without asking them what they think. You cannot regard aggregate data as collection of data in the natural environment. If it is possible to ask people they do things. When public attitudes and mass behavior are of interest, surveys 168 The third consideration is the practicality of the scale-development procedure. Developing a measurement instrument requires a huge number of respondents. Necessary steps su rs must be gned to assist public relations practitioners with applied commu rucial to s chapter, I theoretically specified the two variables to be explained by a set of antecedent variables, including problem recognition, level of involvement, constraint recognition, referent criterion, and situational motivation in problem solving. To test these concepts ch as pilot testing of initial item pools, item analysis for reliability and validity, and cross-validation all demand multiple and large samples. Survey research makes it possible to collect a large amount of data through standardized questionnaires economically and quickly (Babbie, 1998). In addition, practical research situations for public relations practitione considered. The situational theory of publics has been widely adopted for planning communication programs (e.g., public information campaigns). Practitioners segment relevant focal publics for their communication program not only through qualitative study but also through formative survey research. Because the situational theory of problem solving is desi nication practices (e.g., health and risk communication campaigns), it is c establish a set of reliable and valid measurement systems. To meet that purpose, survey research was desirable. Research Design Design of Study This dissertation research required two phases. The first phase was to develop a set of reliable and valid scale items for the two new variables of communicant activenes in problem solving and cognitive entrepreneurship in problem solving. In the previous 169 and propositions (i.e., the specified inter-variable relationships), I first developed corresponding measurement systems for these consequent??explanandum? an antecedent variables??explanans? (Hempel & Oppenheim, 1953). This task was to describe the phenomena in an empirical way for the purpose of making it possible to falsify the concepts and propositions proposed (Popper, 1963). The second phase was to test the viability of the newly developed situationa theory of problem solving. Researchers cannot test a theory directly, but they can test a derived model through empirical observations of the theoretical propositions. From the first phase of study design, I produced a testable measurement system with acceptable levels of reliability and validity. With this measurement system, I proceeded to test caus relationships between antecedent conditions and consequent phenomena. This was an attempt to falsify the models for the purpose of demonstrating the viability of the situational theory of problem solving. In testing logic, if the falsifying attempts fail, we can gain some confidence in the theoretical plausibility of the concepts and propositions. However, the tenability of the theory is only provisional. In the following sections describe the procedure used in the first phase of this dissertational study. Scaling Procedure A typical way to d d l al , I will evelop a scale is to start by refining the purpose of the scale. My pur le lem public relations professionals or risk and health communicators can apply the scales to pose in this study was twofold. The first was to test the concepts and propositions in the situational theory of problem solving. The second purpose was to identify a sca that could be used both for theoretical and practical research related to people?s prob solving. Specifically for practical research, applied communication researchers such as 170 segment a population into more meaningful chunks based on their differential cognitiv and communicative features. e struct structure 17 with ef itself in s rule of three Netemeyer, Bearden, and Sharma (2003) presented a clear and efficient model to illustrate the process of scale development. The first step is to define constructs clearly. The researcher should carefully explicate a construct and its dimensions. Otherwise, subsequent measurement scales have little use in theory and practice. Also, a scale developer should be concerned about construct dimensionality?whether it is unidimensional, multidimensional, or a second-order construct. Netemeyer, Bearden, and Sharma (2003) recommended having an effect indicator (?reflective? items) rather than causal indicator (?formative? items) (Bollen, 1989). In the present study, the two new constructs, CAPS and CEPS, were hypothesized as a second-order con fect indicators. The second step was to construct and choose the best items. Conceptual explication should precede the task of generating item pools so that theoretical meaning can guide the composing of measurement instruments. In doing so, ?nontrivial redundancy? among items was necessary to allow the phenomenon of interest to manifest different ways (DeVellis, 1991, p. 56). Such nontrivial redundancy among item was critical because ?the content that is common to the items will summate across items while their irrelevant idiosyncrasies will cancel out? (DeVellis, 1991, p. 56). A thumb in item construction is clarity. DeVellis recommended having a pool of items Both constructs had enough conceptual generalness that allow further decomposition of the concepts. example, whereas cognitive entrepreneurship in problem solving can generate knowledge to understand 17 In developing CAPS and CEPS, I took a higher-order factor structure approach with a theoretical purpose. For differential problem solving approaches among people, we can further segment it in a conceptually meaningful way. Theorists should explicate conceptual meanings and dimensions to the extent it generates understanding of the phenomenon. Indeed, I intended that researchers can use each subdimension of two constructs as an independent construct if necessary. 171 to four times larger than the final scales. He and other scholars suggested that good scale items should be unambiguous, relatively short (e.g., avoid wordiness), easy to t multiple negatives (e.g., innuendos), not double barreled, without ambigu , , rmat g., ong agreement). Finally, the scale developer should ask both ex , I ), comprehend, withou ous pronoun references, and without misplaced modifiers (Netemeyer, Bearden & Sharma, 2003; Specter, 1992; DeVellis, 1991). In addition, the researcher should choose items that prevent some common respondent biases such as acquiescence affirmation, or agreement. Wording items both positively and negatively within the same scale can be a good preventive strategy (DeVellis, 1991). Following item construction, the researcher should choose a response format. In the present study, I used a 9-point Likert format spanning from ?not at all? to ?extremely.? I used a unidirectional fo (e.g., from absence of agreement to extreme agreement) rather than bidirectional (e. from strong disagreement to str perts and laypeople whether the items match theory?i.e., to make item judgments. This makes it possible to test content and face validity (Netemeyer, Bearden, & Sharma, 2003). To get a laypeople perspective in item selection, I recruited three graduate students who were not familiar with the situational theory. To get an expert perspective asked Dr. James E. Grunig, the developer of the STP, to review the fit between theory and the proposed items. The third step was to design and conduct empirical tests to develop and refine the scale. By drawing a relevant sample, the scale developer can do item-trimming (e.g., rewriting items), test psychometric properties, do exploratory factor analyses (e.g., checking dimensionality), check internal consistency estimates (e.g., Cronbach?s alpha 172 make initial estimates of validity, and sort out items for cross-validation (Netemeyer, Bearden, & Sharma, 2003). pers can des alyses consist confirm equivalence). In addition, researc , y related with each other (i.e., the nonrecursive and construct validity?whether the scale has theoret e The last step was to finalize the scale. By using several samples, scale develo ign and conduct various types of reliability and validity analyses. These an of exploratory factor analysis (e.g., removing low or cross loading items) and atory factor analysis (e.g., checking factor structure hers should test conventional reliability statistics and validity testing procedures. In the present study, I first did a series of principal component analyses. I conducted an internal consistency test using Cronbach?s alpha. I also examined construct reliability using coefficient H, and the variance extracted confirmatory factor analysis for CAPS and CEPS (Hancock & Mueller, 2001) to check as construct validity. Regarding validity, I checked face validity?whether the measurement items appear to represent well the construct to be measured; concurrent validity?whether two or more related measures were statisticall model between CAPS and CEPS); ically relationships with other relevant variables (e.g., SITCAPS and SITCEPS). Above all, construct validity was the most important testing procedure in scal development. I tested construct validity of two new dependent variables by connecting their measures to other theoretical variables in the situational theory. Earlier, I theoretically specified a priori causal relationships among the key constructs in the situational theory of problem solving. The structural interrelationships between the situational antecedent variables and two new variables were tested against the data. 173 Should there be close fit between the specified model and the observed data, I could in a corresponding degree of construct validity for the scale developed here. Sampling Strategy fer ility s ry. In e sample is not necessary if statistical generalization of the finding is not the goal. They emphasized that it is the theory that is applied beyond the research setting and thus ?the research sample need only allow a test of the theory?any sample within the theory?s domain (e.g., any relevant sample), not just a representative one, can provide such a test? Rationale for non-probability sampling strategy. For the present study, I have chosen a convenience sampling (i.e., a nonprobability sampling) strategy for practical and theoretical reasons. A nonprobability sample does not allow those members in a population to have an equal probability of being chosen. The choice of a nonprobab sampling strategy makes it impossible to estimate unbiasedness and sampling variability for a chosen sample. Thus, it does not allow the researcher to draw any inference about a population (Schwarz, Groves, & Schuman, 1998). In addition, data analysis with nonprobability sampling can result in statistical problems such as biased parameter estimates and standard errors. Nevertheless, a nonprobability sampling strategy is best for the research problem in this study. The purpose of this study was to develop new variables and new theo practice, ?most early tests of nascent theories? adopt non-probability student samples, and such a strategy is, not ideal, but useful for initial theorizing and hypothesis testing with ?multivariate relationships? (Caplan, 2005, p. 732). Besides, I delimited the scop to ?theoretical generalizability? rather than statistical generalizability. Calder, Philips, and Tybout (1981) pointed out that to make theoretical generalizations, a representative 174 (Calder, Philips, & Tybout, 1981, p. 200). In addition, Shapiro (2002) addressed the issue: [R]ejecting a study that seeks to expand theory and that detects a potentially important effect on the basis of a nonrandom sample usually reflects a important effect, no matter what kind of sample is used, it is clearly true for some may be to conduct a theoretically driven boundary search to determine to whom s ata which are nalysis misunderstanding of the nature of generalizability. If a study detected an group of people, in some setting, at some time, for some message. The next step the effect applies and to whom it does not. (p. 499) Sample size demands for relational analysis are often huge. Meeting the required number of participants using probability sampling would not be practical. Thus, in this dissertation, I combined the two most common nonprobability sampling methods of convenience sampling and multiplicity sampling (a.k.a., snowball sampling or network sampling) to test theoretical propositions. Multiplicity sampling consists of using previously identified members of a group to find other members of the population. This sampling method was useful to trace patterns of influence among connected participant (Weisberg, Krosnick, & Bowen, 1996). Sample Size Estimation I have conducted a simulation study to estimate the minimum sample size for d collection. In most empirical research methods, the researcher can use Kraemer and Thiemann?s (1987) classical guideline to estimate sample size and to determine power. However, they did not provide a procedure for estimating sample size for confirmatory factor analysis and causal path modeling, which are used in my dissertation. Hancock (2006) advised that ?a priori? power analysis is more desirable than ?post hoc methods,? often merely a ?self-pitying? (p. 103). He reminded us that a post hoc a 175 is ?merely establishing what is already apparent? (p. 103) 18 . In this vein, I chose an a te sample size with onte Carlo 2) illustrated how plus. Notably, the and two APS and STICEPS) were of primary interest. To see the sample ombined ition, constraint recognition, level nt of missing data, reliability of the g the variables (L. Muth?n & B. Muth?n, ually by eights. Normality of data in SEM was a key assumption. I sting will priori Monte Carlo simulation approach to determine appropria power .80 and a .05 level of significance. The necessary sample size for this study was estimated using Mplus M Simulation option (2004 Version 3.01). L. Muth?n and B. Muth?n (200 to estimate optimal sample size and to determine power by using M Mplus approach focuses on individual parameters within the model, not the entire model. 19 In my study, two confirmatory factor models (i.e., CAPS and CEPS) causal path model (SITC size and a rough check for power for individual path estimates, I used a model c with all the variables in one model (i.e., problem recogn of involvement, referent criterion, CAPS, and CEPS). As a first step, I specified a priori conditions for key factors such as the size of the model, distribution of the variables, amou variables, and strength of the relationships amon 2002). Then, each residual variance of endogenous variables is computed individ considering a priori path w presumed that data transformation (i.e., power transformation) before model te satisfy the normality assumption. No missing data assumption was made. Although such 18 Hancock (2006) provided a general guideline for a priori power analysis: ?when ? = .02 [? = RMSEA] ve appear to be achieve power of ? = .80, samples sizes of n = 500 or above appear to be sufficient for testing overall data-model fit? (p. 103); 000 or above models assuming the models were specified correctly. Thus, I followed the Mplus simulation approach that hinges on the power for ?parameters? within a model. and models have df ? 60, to achieve power of ? = .80, samples sizes of n = 300 or abo sufficient for testing overall data-model fit; when ? = .02 and models have df ? 30, to ?when ? = .02 and models have df = 10, to achieve power of ? = .80, samples sizes of n = 1 appear to be sufficient for testing overall data-model fit? (p. 104). 19 Hancock (2006) clarified two different foci in power analysis: 1) parameter(s) within a model and 2) data-model fit as a whole. In the present study, I mainly focused on the structural parameters within the 176 an assumption was too idealistic, I considered the possibility of using imputation options strategy. Reliability of factor indicators, ior studies. tudies (e.g., past studies of the STP) provided rough approximations to generate , able estimation of sample size with preset power (.80) and significance o, I assumed the overall model had ould not guide any decision ntire model. Despite some inherent inaccuracy in a prior analysis, it g y analysis: ?By choosing the values in models carefully and conservatively, their studies hoping for the best and uational olving, f involvement, ferent criterion, and situational motivation in problem solving. ment scale items for each variable. The instruments I used are as follows: (e.g., EM algorithm) as a possible compensatory factor correlations, and causal paths were all rough approximations from pr Specifically two pilot tests for the STOPS in addition to prior STP s regression weights in data in this simulation. Because the purpose of this Monte Carlo study was modest?i.e. to check a reason level, such gross estimation could be justified. Finally, I mainly focused on determining power to detect individual path coefficients. In doing s been correctly specified. Thus, the present simulation study c regarding power for e is desirable in this study to use SEM analysis. Hancock (2006) recommended conductin preparator researchers have the ability to ensure that they enter prepared for the worst? (p. 104). Instruments. The key variables within the models derived from the sit theory of problem solving were as following: communicant activeness in problem s cognitive entrepreneurship in problem solving, problem recognition, level o constraint recognition, re Explication of each construct in the conceptualization chapter guided the process of drafting measure Instruments for Communicant Activeness in Problem Solving (CAPS) Information Forefending 177 Others r ts about this problem are worthless. I understand this problem. his problem. e to deal with this problem. en to opposite views and they are related to the problem. es from. d interests cepted conflicting information about this problem. ve conversations on this problem. knowledge and perspective about this problem. spending some time to persuade others about this problem. knowledge and thoughts about this problem. I actively seek out opportunities to participate in public opinion polls about this problem. espect my perspective about this problem because it is simple and clear. Some publicized statemen I have invested enough time and energy so that I know where to go when I need updated information regarding t I have studied this problem enough to judge the value of information. I feel like resisting some persuasive efforts around this problem. I express my opinions confidently about what should be don I am picky in choosing information sources when I think about this problem. Information Permitting To make better decisions regarding this problem, I list information as long as For this problem, I welcome any information regardless of where it com I am careful in accepting information about this problem because of the veste of those who provided the information. (R) I listen to even opposite views on this problem. At times, I find that I have ac Information Forwarding Sometimes I find I am engaging in aggressi It is one of my top priorities to share my If it is possible, I take time to explain this problem to others. It is worthy I look for chances to share my 178 I love to start a conversation on this problem with others. I volunteer to inform others about the problem. rship role in initiating conversation about the problem. in passing on information related to this problem in riends and others come to learn more about this problem. egarding this problem. rn what kind of solutions I regularly check to lem on the Internet. ts containing relevant knowledge about the problem. roblem oblem. hat I will be quite active in passing on information related to this problem in y friends and others come to learn more about this problem. In the past, I researched about this problem seriously. I feel happy when I provide new information about this problem to others. I often play a leade Information Sharing I am sure that I will be quite active the near future. I am a person to whom my f In the past, I researched about this problem seriously. At times, I am asked to give advice r Information Seeking From time to time, I contact people about this problem to lea there are. I regularly visit Websites relevant to the problem. see if there is any new information about this prob I would request bookle I visit an online or regular bookstore to find useful information about the p My friends think that I take too much time for learning about this pr Information Processing I am sure t the near future. I am a person to whom m 179 Instruments for Cognitive Entrepreneurship in Problem Solving (CEPS) Cognitive Retrogression I know how I should behave for this problem. e spent too much time on this problem to change my position now. y decision on this problem. position even if someone challenges it with contradictory evidence. pension think before I finalize my position on this problem. all the evidence is in. ng Saddam Hussein while he orst in the world, was enough justification to It is too late to change the position I now have on this problem. I hav It is too late to shake the conclusion I have drawn for this problem. I have made efforts to justify m I will keep my current I have found counter evidence that rejects the positions different from mine. I feel it is costly to change my mind on this problem. Cognitive Sus I want to take more time before making up my mind for this problem. I hesitate to make up my mind about what should be done for this problem. I need more time to For this problem, I will try to suspend any judgment until Pro-War in Iraq Statements The war in Iraq can be justified because the cost of controlli was in power was higher than that of war. Saddam was connected with terrorists. Saddam's human rights record, among the w go to war. Con-War in Iraq Statements 180 With th s a ing, Affirm rt the current situation deviate from what you think it should be? Constra e economic and domestic security problems the United States was facing, it wa bad time to go to war in Iraq. A pre-emptive attack by the U.S. gives credibility to those who describe the U. S. as an aggressive nation. The war in Iraq has increased anti-American sentiment. Pro-Affirmative Action Statements Affirmative action levels the playing field because minority students, generally speak start out at a disadvantage. Diversity is desirable yet won't always occur if left to chance. Con- Affirmative Action Statements ative action demeans true minority achievement because success is labeled as a result of affirmative action rather than hard work or ability. Because of affirmative action, a wealthy minority student who doesn?t put in much effo could be chosen over a poor white student who works harder. Instruments for Situational Antecedent Variables Problem Recognition To what extent do you think there is something missing in this problem? How much does the current situation differ from your expectation? How strong do you feel that something needs to be done to improve the situation for this problem? How much does int Recognition 181 Please think of whether you, personally, could do anything that would make a difference in the way these problems are handled. If you wanted to do something, would your efforts make a difference? (R) To what extent do you believe this problem is a problem that you can do something about (R) To what extent do you bel ? ieve that you could affect the way this problem is eventually Level o t t this problem? Referen hink about each of these three problems? To wha tter. solved if you wanted to? (R) f Involvement In your mind, how much of a connection do you see between yourself and this problem? To what extent do you believe this problem could involve you or someone close to you a some point? How much do you believe this problem affects or could affect you personally? How strong would you say your opinions or thoughts are abou t Criterion I am confident about my knowledge about this problem. I strongly support a certain way of resolving this problem. I have a preference for how the problem should be settled. I am pretty sure I know how to solve this problem. Past experience has provided me with guidelines for solving this problem. Situational Motivation in Problem Solving How often do you stop to t t extent would you say you are curious about this problem? Please, indicate how much you would like to understand each of these problems be 182 Administering the survey instrument. The researcher may administer a survey instrument via a self-administered questionnaire, an interview, a structured record r and structured observation (Fink, 2003). I chose a self-administered questionnaire. A self-administered questionnaire is cheap, produces data fast, is free of interviewer bias, and provides confidentiality and privacy, which encourage candid resp eview, onses (Fink, 2003; Babbie m: when they are not fam on to n should be used. In the present study, I omitted the no opinion option. However, participants were allowed to skip questions. Pilot test. To prepare a final version of the survey instrument, I conducted four waves of pilot tests in August 2003, December 2003, May 2004, and December 2004. , 1998). Participants completed question items through a Web-based survey for i.e., computerized self-administered questionnaire (Babbie, 1998). Participants read questions and made choices on a screen by using a computer keyboard and mouse. Participants who did not have access to a personal computer or access to the Internet were offered a printed questionnaire, but no one requested one. No opinion options. Experts (Weisberg, Krosnick, & Bowen, 1996) on survey research recommend including no-opinion options to prevent respondents from giving meaningless answers. Often, participants feel pressured to respond even iliar with the issue. Thus, a no-opinion option makes participants more comfortable and accurate. However, the experts also advise some caveats for the use of no-opinion option. Specifically, less-motivated respondents use the no-opinion opti avoid thinking about the question. Thus, survey experts recommend omitting no-opinio options concerning well-known issues (e.g., legalization of abortion). If the issue of inquiry is obscure or unfamiliar to participants, a no-opinion option 183 These p g sampling strategy). Although I used the results of these to construct ments, I did not report ertation. Data C . n owball tial were in the purpose of study is to understand better about information traffick re. The researcher explained how to find the survey website (URL address and log-in retests tested preliminary item pools for the variables in the situational theory of problem solving (e.g., correlations between the testing items and social desirability scale items), determined response format (e.g., fractional scale vs. Likert-type scale), and tested the data collection method (a snowballin the final version of the survey instru them in this diss ollection Procedure Data were collected during April and May 2005 at the University of Maryland. Students registered for an introductory communication course were invited to participate To encourage participation in the study, students were all provided with extra credits i discussion with their instructors. The participation was voluntary, and they were given alternative assignments if they decided not to participate. Because I chose a sn sampling technique, I devised a clustering method as follows. First, I identified an initial contact person who was interested and volunteered for participation in exchange for extrapoints in the communication classes. Once the ini contact person agreed to participate, participants were instructed in how to recruit other candidate participants from their personal relationships. Importantly, participants were advised repeatedly not to force participation in ?snowballing.? Instead, participants encouraged to expla ing patterns among people and that the recruiting student would get some extra- credit from participating. Next, participants were shown the website that contained a survey questionnai 184 method) 20 and illustrated how to complete the questionnaire directly. However, participants were asked to explain that the participation would be voluntary to the candida f ld enter ? tial e te respondents and, most importantly, confidential. Once the acquaintances agreed to participate, the initial contact person provided his or her own login code to the others to form a cluster under the initial contact person?s login code. To start the survey questionnaire, participants had to provide the first letter o their last name with the last four digits of their social security number. For example, a participant whose name is John Smith (Social Security Number: 123-45-6789) entered ?S6789? to login in the first page of Website. Similarly, participants who were recruited indirectly through the initial contact persons entered the initial contact person?s login code (e.g., ?S6789?) with their own code (e.g., Mary Adams with the 987-65-4321 wou A4321?). This login method ensured that the participant?s identity was confiden because no other identity information was asked in the questionnaire. 21 Once they successfully logged in, participants read the informed consent form. The consent form instructed participants that there was no foreseeable personal risk with their participation and that they had the right to withdraw participation at any time without any penalty and/or decline to answer certain questions. After participants read and agreed to the informed consent form by clicking an ?agree? button, they started th questionnaire. Every participant read an identical set of questions. firm, CreateSurvey ( 20 The website containing the questionnaire has been constructed by a professional online survey research http://www.createsurvey.com). 21 For a confirmatory purpose to assign extra credits, the initial contact persons in the communication classes provided a confirmatory r acquaintances after they learn the eceipt that contained their names and the names of participated ir acquaintances have completed the questionnaires. The receipts was destroyed immediately once their participations are confirmed. Thus, there was no way to identify personal information with responses. 185 One possible risk that lowers the data quality was a multiple response by a single participant. To prevent such multiple responses, I used a ?duplicate tracking function,? technical service provided by the CreatSurvey. The CreatSurvey limited each respondent to one entry by using a special cookie device. Data Reduction and Data Analysis Model Testing a ing the ; s using ension. After a set of best measur the ed The first research goal was to develop and validate reliable and valid measurement scales for two new constructs, communicant activeness in problem solv (CAPS) and cognitive entrepreneurship in problem solving (CEPS). I here specified theoretical models and subdimensions under CAPS (six dimensions) and CEPS (four dimensions). I adopted a two-step structural equation modeling approach (Kline, 1998 Byrne, 1994). In the first step, the measurement phase, I analyzed completely confirmatory models that allow covarying among all the latent variables and stand-alone variables in the model. Then, I checked for correlated residuals and cross-loading LM tests and removed low loading items for a given dim ement items was identified for each dimension, I tested its goodness-of-fit to overall measurement structure. This required second-order factor analyses that subsum subdimensions under the higher order constructs (e.g., communicant activeness in problem solving). In the second step, the structural phase, I compared the final confirmatory models with the proposed structural models. I then respecified the initial structural models with applications of the LM test and Wald test, if necessary. It turned out that most of the 186 initial structural models were reasonable data-model fits based on multiple fit indices Hence, most models were tested as th . ey were originally specified. e proposed structural equation models, I adopted commonly adopted c t is to use ? 2 /df to reduce the sample ten s ative of an acceptable model. , To evaluate th model fit indices. They are ? 2 and its degree of freedom, Comparative Fit Index (CFI), Root-Mean-Square-Error-of-Approximation (RMSEA), Standardize Root Mean Square Residual (SRMR), Non-Normed Fit Index (NNFI), and Akaike Information Criterion (AIC). Often ? 2 values are sensitive to large sample size. Thus, the ? 2 statisti would be significant even if the differences between observed data and model-implied covariances were small (Kline, 1998). A common treatmen size sensitivity. Although there is no rule of thumb to use, researchers of consider the minimally acceptable value of ? 2 /df ratio to be less than 3 (Kline, 1998). CFI and NNFI are incremental measures of fit, and acceptable levels of fit are values close to 1.00. SRMR is a standardized summary of the average covariance residuals. When the average discrepancy between the observed and imposed covariance is perfect, the SRMR value becomes 0. As the discrepancy grows, SRMR values increase as well. SRMR values less than .10 are commonly accepted as indic RMSEA (Steiger & Lind, 1980) tends to correct for ?tendency of the chi-square statistic to reject a model? with a large sample size (Netemeyer, Bearden, & Sharma, 2003, p. 152). Its values of .08 or less are indicative of acceptable fit (Browne & Cudeck 1993). AIC is a parsimony index, and values close to 0 or lower are considered as indicative of acceptable model (Akaike, 1987). AIC is useful when a researcher needs to compare nonhierarchical models, thereby direct comparisons, using the differences of ? 2 and its degree of freedom. 187 Recently, Hu and Bentler (1999) recommended the use of multiple indices, as CFI, RMSEA, and SRMR. With the joint-criteria approach (Hu & Bentler, 1999), model is considered viable when it has CFI ? .96 such a and SRMR ? .10 or RMSEA ? .06 and o interpret the individual parameter estimates to test hypotheses and explore research questions. m truct reliabil y of re ta. When a SRMR ? .10. Once the models have met this proposed model evaluation approach, I proceeded t In using a sample, I made sub-samples from the data set by using the rando selection function in SPSS 11.5. Thus, I divided the first subsample (n = 457) as the developmental samples of CAPS and CEPS and the other subsamples as validation samples (e.g., n = 917). Finally, I ran the models by using the total sample size (n = 1,380), once I reached the best models. For the CAPS model, I conducted a series of nested model tests among one factor, six factor oblique, and a second-order factor model to see its dimensionality. For the CEPS model, I directly proceeded to the second-order structure testing and found acceptable data-model fits. For the causal models of situational theory, I used the total sample (n = 1,380). In addition to the tests of factor analysis structures, I analyzed conventional internal consistency and reliability measures such as Cronbach?s alpha and cons ity. In addition, I computed coefficient H (Hancock & Mueller, 2001) and variance extracted for the latent variables used in the structural model analysis. At the next phase of analysis, I tested three models from the situational theor problem solving. These specified models with a priori hypothesized causal paths we subject to first overall model testing by checking their goodness-of-fit to the da 188 model achieved a reasonable model-data fit, I interpreted its paths to evaluate the hypotheses and research questions. Model s a les. Thus, the second-order factor model was identified and could be tested ( mitment, could b derived from the situational theory of problem solving (STOPS). I erived three structural models from the situational theory of problem solving. Using CAPS, first, I constructed the model of situational theory of communicant activeness in problem solving (SITCAPS). Using CEPS, I constructed the model of situational theory of cognitive entrepreneurship in problem solving (SITCEPS). Both models had identical model specification in their structural relationships (see Figure 11 and 12 in the Chapter II: Conceptualization). Finally, I specified a nonrecursive model to test bidirectional Identification Communicant activeness in problem solving (CAPS). A measurement model with four or more indicators for the construct to be measured is always identified. CAPS ha second-order factor structure with six subdimensions under the construct of communicant activeness. Each subdimension (e.g., information seeking) had at least three or more measurement variab see Figure 5 in the Chapter II: Conceptualization). Cognitive entrepreneurship in problem solving (CEPS). CEPS has four subdimensions: cognitive retrogression, cognitive multilateralism, cognitive com and cognitive suspension. Notably, cognitive multilateralism and cognitive commitment had been measured by a single measurement item because they were computed by conceptually derived formulas. However, the overall model was identified because the remaining two dimensions had four or more measurement items. Thus, the CEPS model e tested (see Figure 8 in the Chapter II: Conceptualization). Three models d 189 causal influences between CAP t least one exogenous variable that affects one of ffecting the other. I created ere rm (Bogdan & Biklen, 002) advised informing respondents, protecting respondents, and explaining benefits, if any, to , J. Grunig and L. Grunig (2000) on, and preserving privacy as the three major e esearch protocols and ins S and CEPS. This requires a the endogenous variables while not a two nonrecursive models with the referent criterion as exogenous variable (see Figure 13 and 14 in the Chapter II: Conceptualization). Both nonrecursive models w identified. Ethical Considerations In studies using human participants, two most dominant ethical issues are obtaining informed consent and protecting participants from potential ha 1998; Babbie, 1998). I chose a survey for the current study. Survey research, like other scientific inquiries, has ethical responsibilities in treating human beings as objects of investigation. Fowler (2 respondents. Similarly listed participant consent, prevention of decepti thical issues in public relations research. In research with human subjects, there almost always is a ?tension? between a researcher?s ?scientific need? and participant?s ?right to decline? of participation (Schwarz, Groves, & Schuman, 1998). To manage such tension, I paid attention for those major ethical issues throughout the research. After I finalized the research method and plan, I submitted the r truments for review by the University of Maryland Institutional Review Board (IRB). Upon approval of the research, I contacted the faculty and instructors in the Department of Communication to get approval and make alternative assignments if participants refused to participate in the study. Thus, those participants or non- 190 participants in the respondent pool had equal opportunity to get extra-credit (i.e., 20 points). As participants agreed to participate, they were first presented with the informed consen as n contact information if they wanted to ask a question or had concern ith tiality ing any te. mmarize what I found from the data analysis and hypothes t form, in which they found the title and purpose of the study, procedures such time, assurance of confidentiality, possible risks and benefits, most importantly, their freedom to withdraw from study or decline for some questions. Respondents were give the researcher?s name and s to share. All of the responses were kept confidential as promised, although respondents entered a self-made individual code (e.g., last four digits of social security number w first letter of their last name) for assigning responses to their clusters (i.e., who ?snowballed? the participant). However, it was practically impossible for the researcher to connect the personal information with the individual responses. Thus, confiden was secured. Notably, it was acknowledged clearly that participation would not br possible harm and personal benefits to the respondents before they decided to participa In the next chapter, I will su is testing. 191 CHAPTER IV: RESULTS Descriptions of Survey Participants One of the main purposes of this study is to develop and test measurement item for the new concepts of communicant activeness and cognitive entrepreneurship in problem solving. Because of the number of conceptual factors in the study, the le the survey questionnaire was very long (i.e., 100 question items regarding thr s ngth of ee scriptive measures regarding particip he ipate in social on in and 97, n ubdimensions. For the war in Iraq issue, the percentages of missing values varied from 3.8%-5.8%; for losing weight, the percentages individual and social issues). Thus, I removed basic de ants (e.g., gender, income, age) to reduce participants? fatigue. A total of 1,380 University of Maryland undergraduate students participated in t survey. I invited students from an undergraduate communication course to partic exchange for extracredit. Preliminary Analysis Missing Value Analysis (MVA) and Normality The survey instruments in this study consisted of 100 items regarding three and individual issues: war in Iraq, losing weight, and elimination of affirmative acti American higher education. There were 49 items to measure communicant activeness its subdimensions, 20 items to measure the cognitive entrepreneurship and its subdimensions, and 22 items for the situational antecedent variables. I conducted missing value analysis (MVA) and found no significant difference between this data pattern and a random data pattern: i.e., Little's MCAR test: ? 2 (115 = 1,380) = 35.204, p = 1.000. In addition, I conducted missing value analysis (MVA) for the individual constructs and their s 192 193 varied from etric variables did not seriously deviate from the In all of the structural equation model analyses, I used the maximum likelihood (ML) estimation method to reduce the undesirable effects from non-normal distribution of Finch (1995 substantially nonnormal, ML or GLS should be used as an estim th e nonnormal variables in the data et, I conducted s al SEM a alyses in whic the same models to two different data sets (i.e., transformed and non- transformed data). The com ilar fits and s ame estimat . Thus, I proce ith a non-tran ed data set with a ?ML? estimation method in subsequent analyses. 22 Reliability Tests and Exploratory Fact onstructs In the present st developed two in prob m solving and cognitive entrepreneur ip in problem solving. Each construct has sub-constructs that they are correlated but possess unique conceptual meanings toward th nstructs. As a first step, I examined these new constructs and 3.8%-5.9%; and for elimination of affirmative action, the percentages varied from 3.9%-5.9%. Little?s MCAR tests showed that all of the variables had completely random missing patterns. Before analyzing the data, I examined the univariate outliers and kurtosis and skewness in the univariate distribution of each variable. The majority of variables looked symmetric, whereas those non-symm symmetric distribution. data. West, , and Curran ) suggested that given variables that are not ation method. To secure e use of som h I fit s ever n parisons resulted in sim imilar par ter es eded w sform or Analysis of the Key C udy, I new constructs of communicant activeness le sh eir higher co 22 An alternative estimation method (i.e., ML, Robust) was run and gave comparable results. In general, ML, Robust estimation resulted in a similar estimation of parameters but more favorable model fit information (e.g., CFI and RMSEA) than ML estimation. Although I could report the ML, Robust estimation outcomes to highlight the better model fits, I chose ML outputs to be more conservative in interpretation. 194 subdim s of their effectiveness in measuring the intended constructs. Speci , I asses and internal consistency of the data using the Cronbach?s alpha test. I also conducted an exploratory factor analysis (i.e., principal comp t analysis) to assess its loadings, dim Eigenvalues using the SPSS 11.5 program. The following section presents the items and the results of exploratory factor analyses the item t I po . Communicant Activeness in Problem Solving (CAPS) In Table 3, I summarize the result of the reliability tests and principal component problem solving. I used three problems, or CAPS in three different problems showed acceptable levels of pha coefficients: .9 r w in Iraq, .94 for losing weight, and .95 for eliminating mative action. Although I treated CAPS as a ensional construct here, I nsions: information forefending, information itting, information forwarding, infor aring, inform dim y, I conducted nsions for the inating mative action in term .33% of variance correspondingly. In eliminating affirmative action, four dimensions cou for 54.28% of variance. ensions in term fically sed the reliability onen ensionalities, and and reliability tests for s tha pro sed analysis (PCA) of communicant activeness in war in Iraq, losing weight, and elim Cronbach?s alpha tests f al affir proposed CAPS to have six subdim perm inform principal component analysis. PCA showed th war in Iraq and losing weight issues and f affir in Iraq and losing weight, five dim 55 ac inating affirmative action, in this study. The 3 fo ing. To see whethe s of the Kaiser ru ar unidim h at there were five subdim e ensions mation s r CAPS has m our dimensions for the issue of elim le (i.e., Eigenvalues higher than 1.00). In war explained about 56.14% of variance and ation seeking, and ensioation process ulti nalit e nted 195 ultidimensional structure under CAPS (i.e., subfactors were emerged from sis) s rther broke down CAPS into its six dimensions and examined its reliability and structure. I summarize the results and items in following tables. Table 2 reports the results from one facto alysis of all the variables as indicators of C able 4-9 report the results of the separate factor analyses in six subdim sion . Table 3 Reliability for Communicant Activeness in Problem Solving (n = 1380) Construct War in Iraq Losing Weight Eliminating Affirmative Action I fo the on und e f a m actor analy , thu , I fu r an AP s of S. T CAen PS Alpha .94 .95 Cronbach?s .93 10.77 11.62 12.95 3.38 3.04 3.15 1.63 1.41 1.69 1.18 1.16 1.21 ige lues E nva 1.01 1.03 % of Variance Explained 54.28 Cumulative 56.14 55.33 Table 4 Items, Factor Loadings, and Reliability for Information Forefending (n = 1380) War in LosiIraq ng Weight Eliminating Affirmative Action Construct It Factor ng a ac Load l em Loadi Alph F tor in A g pha Factor Loading Alpha Others respect my perspective about this problem .57 .63 .61 because it is simple and clear. Some publicized statements about this problem are .54 .50 worthless. I have invested enough time and energy so that I understand this problem. .81 .81 .82 I know where to go when I need updated information regarding this problem. .68 .73 .6 7 I have studied this problem enough to judge the value of inform .80 ation. .81 .8 1 I feel like resisting some persuasive efforts around this problem. .53 .59 I express my opinions confidently about what should be done to deal with this problem. .75 .74 .7 6 Forefending .49 .80 .50 .79 .46 .81 Information I am picky in choosing information sources when I think about this problem. Eigenvalues 3.45 3.01 3.54 % of Variance Explained 43.08 50.21 44.23 196 Table 5 Items, Factor Loadings, and Reliability for Information Permitting (n = 1380) War in Iraq Losing Weight Eliminating Affirmative Action Construct Item Factor Loading Alpha Factor Loading Alpha Factor Loading Alpha To make better decisions regarding this problem, opposite views and information as long as they are rela to the problem. I listen to ted .74 .66 .70 For this problem, I welcome any information regardles where it comes from. s of .42 .57 .46 I am careful in accepting information about this problem -.60 -.59 -.60 because of the vested interests of those who provided the information. (R) I listen to even opposite views on this problem. .79 .75 .74 In n formatio Permitting At times, I find that I have ted conflicting information about this problem. . accep .66 65 .71 .65 .68 .64 Eigenvalues 2.14 2.19 2.07 % of Variance Explained 42.88 43.71 41.31 197 Table 6 Items, Factor Loadings, and Reliability for Information Forwarding (n = 1380) War in Iraq Losing Weight Eliminating Affirmative Action C Factor Alpha Factor L onstruct Item Loading oading Alpha Factor Alpha Loading Sometimes I find I am engaging in aggressive conversations on this problem. .65 .66 .69 It a .77 .75 .78 is one of my top priorities to share my knowledge and perspective bout this problem. If to others. it is possible, I take time to explain this problem .78 .76 .77 It pr is worthy spending some time to persuade others about this oblem. .56 I pr .81 .76 .74 look for chances to share my knowledge and thoughts about this oblem. I a actively seek out opportunities to participate in public opinion polls bout this problem. .64 .72 I love to start a conversation on this problem with others. .80 .74 .76 I volunteer to inform others about the problem. .76 .77 .78 I feel happy when I provide new information about this problem to others. .67 .64 Inf Forwarding I often play a leadership role in initiating conversation about the problem. .78 .89 .73 .88 .78 .90 Eigenvalues 4.53 4.55 4.95 % of Variance Explained 56.64 50.52 54.97 ormation 198 199 Table 7 Items, Factor Loadings, and Reliability for Information Sharing (n = 1380) War in Iraq Losing Weight Eliminating Affirmative Action Construct Item Factor Loading Alpha Factor Loading Alpha Factor Loading Alpha I am sure that I will be quite active in passing on information related to this problem in the near future. .76 .74 .77 I am a person to whom my friends and others come to learn more about this problem. .82 .83 .81 In the past, I researched about this problem seriously. .76 .78 .78 At times, I am asked to give advice regarding this problem. .79 .80 1 nInformatio Sharing .78 .81 .83 .8 Eigenvalues 2.45 2.48 2.54 % of Variance Explained 61.12 62.01 63.55 200 Table 8 Items, Factor Loadings, and Reliability for Information Seeking (n = 1380) War r L g gh liminating in I aq osin Wei t E Affirmative Action Constr Item A a a o g p to i Alpha uct Factor Loading lph F ctor L adin Al ha Fac r Load ng From tim to ti e, I onta wh kin of s utio the e m c ct p o i b t r at d ol ns re . 9 peo le ab ut th s pro lem o lea n are. .69 66 .6 I re ularl visi web tes rg y t si el t he b . . 2 evan to t pro lem .80 80 .8 I re ularl che k to e if this problem o the I tern g y c se th s n n ation about n n et . 3 ere i any ew i form . .77 83 .8 I w uld r ques boo ets c the problem. o e t kl o i el t w e ut . 3 ntain ng r evan kno ledg abo .76 74 .7 I vi it an nlin or re ular s o e g b st o ul r about the problem . 5 ook ore t find usef info mation .74 79 .8 Informa Seekin f d nk t e m ch tim o r a . 8 9 .87 tion g My rien s thi tha I tak this problem. too u e f r lea ning bout .81 73 . 5 .7 Eigenvalues 2.83 3 9 .46 3.6 % of Varian Explain 2 5 4 56.5 ce ed 7.65 61. 5 201 Table 9 Items, Factor Loadings, and Reliability for Information Processing (n = 1380) War in I n i Eliminating Affirmative Action raq Losi g We ght C tr Item Factor Loading h r L n A a a pha ons uct Alp a Facto oadi g lpha F Lo ctor ding Al I am sur that will e qu te ac in rmation related o thi prob e I b i ti n in fo t s l ut 7 .64 ve i pass g on em in the near f ure. . 8 .75 I am a person to whom m frie learn m e ab ut th pro em. y n nd er m or o is bl 8 .84 ds a oth s co e to . 6 .85 Information roce ing In he p t, I r sear ed a out t P ss t as e ch b h o seriously. .73 0 .72 8 is pr blem .7 .78 .70 .5 Ei a .8 1.63 genv lues 1 8 1.89 % o Varia Explai . 1 4.50 62 80 62.9 5f nce ned Table 4 summarizes the items, loadings, Cronbach?s alpha, and variance explained in three problems for information forefending. In information forefending, I used eight items and found only one dimension: i.e., Eigenvalues = 3.45, 3.01, and 3.54 in war in Iraq, losing weight, and eliminating affirmative action. Factor loa dings varied 43.08%, 50.21%, and 44.23% and the e rnal ied % .64, correspondingly. Yet, I found the reliabil from .6 ude from .46 to .81. The amounts of variance explained were Cronbach?s alphas were .80, .79, and .81, correspondingly. Therefore, I conclud that information forefending reached an acceptable level of reliability and inte consistency. Table 5 summarizes the items, loadings, Cronbach?s alpha, and variance explained in three problems for information permitting. In information permitting, I used five items and again found only one dimension: i.e., Eigenvalues = 2.14, 2.19, and 2.07 for war in Iraq, losing weight, and eliminating affirmative action. Factor loadings var from .42 to .79. The amounts of variance explained were 42.88%, 43.71%, and 41.31 and the Cronbach?s alphas were .65, .65, and ity coefficients were less than desirable (i.e., .70 to .80). The fewer number of items would be a possible cause for marginal reliabilities. Table 6 summarizes the items, loadings, Cronbach?s alpha, and variance explained in three problems for information forwarding. In information forwarding, I used 10 items. I also found a single dimension: i.e., Eigenvalues = 4.53, 4.55, and 4.95 for war in Iraq, losing weight, and eliminating affirmative action. Factor loadings vary 4 to .81. The amounts of variance explained were 56.64%, 50.52%, and 54.97% and the Cronbach?s alphas were .89, .88, and .90, correspondingly. Therefore, I concl 202 203 nd the Cronbach?s alphas were .79, .81, that information forefending reached an ems, load explained in three problems for informat ation seeking, I used six it ension: i.e., Eigenvalues = 2.83, 3.46, and 3.69 for war in Iraq, losing weight, and eliminating affirmative action. Factor loadings varied from .66 to .85. The amount of variance explained was 56.52%, 57.65%, and 61.45% and the Cronbach?s alphas nd . herefore, I conclude that information forefe n acc ility and internal consistency. Table 9 summ ms, explain d in three problem ation processing. In information processing, I used three item ension: in Iraq, losing weight, and eliminating affirmative action. Factor loadings varied from .64 to .86. The variances explained were 62.80%, 62.91%, and 54.50% and the Cronbach?s alphas were .70, .70, and .58 correspondingly. Although the reliability of eliminating affirmative action resulted in a marginal alpha coefficient, the other alpha coefficients reached an acceptable level of reliability. that information forwarding measures reached an acceptable level of reliability and internal consistency. In information sharing, I used four items and found a single dimension: i.e., Eigenvalues = 2.45, 2.48, and 2.54 in war in Iraq, losing weight, and eliminating affirmative action. Factor loadings varied from .74 to .83. The amounts of variance explained are 61.12%, 62.01%, and 63.55% a and .81, correspondingly. Therefore, I conclude acceptable level of reliability and internal consistency. Table 8 summarizes the it ings, Cronbach?s alpha, and variance ion seeking. In inform ems and found a single dim were .81, .85, a 87, correspondingly. T nding reached a eptable level of reliab arizes the ite loadings, Cronbach?s alpha, and variance e s for inform s and found a single dim i.e., Eigenvalues = 1.88, 1.89, and 1.63 in war 204 In summary, I found that the new construct communicant activeness in problem solving achieved a reasonable internal consistency and reliability. I treated CAPS not only as a unidim t, but also as a set of multi ts. In both und acceptable agreements between data and proposition. Cogn trepreneurship in Problem Solving (CEPS) The second new construct I proposed is cognitive entrepreneurship in problem solving (CEPS). CEPS consists of four conceptual subdimen s: cognitive retrog ion, cognitive suspension, cognitive m, an g e commitment. s for cognitive retrogression and cognitive susp le items to measure n u and cognitive commitment: two putational formulas in the conceptualization. For CEPS, I conducted an analysis of d internal co m liability and principal component analysis. arize the result of ponent is (PCA) of cognitive entrepreneurship easurement, roblems: war in Iraq and inating affir itive teralism and cognitive comm ent strategies using , whi i i s r CEPS in two different problems. Here a PS as if it were a unidimensional construct by entering a our ensions. Analysis showed alpha coefficients of .66 for war in Iraq, which is ensional construc dimensional construc cases, I fo itive En sion ress multilateralis d co nitiv ension, whereas I introduced multiple item I used sing com reliability an retrogression and suspension. In the followi re analys I used two p multila respondents? evaluations about a conf weight problem I tre subdim cog cy an n-con itive m alysis elim itment require licting so trovers ltilateralism only for the subd ng tables, I sum re in problem cial issue. Thus, I excluded the losing al i nsisten s a no imensions of cognitive arize the results of the In Table 10, I summ liability tests and principal com dif sue. solving. In CEPS m mative action. Cogn ferent measurem ch I first conducted Cronbach?s alpha tests fo ted CE ll items for f marginal, and .74 for eliminating affirmative action. To see whether CEPS has multi io ty, I conducted prin l component analysis. PCA showed that there are fo di ns in both issues: Eigen lu 71, 2.53, 1.37, and 1.01 for war in Iraq and 4.72, 2.43, 1.49, and 1.28 for eliminating affirmative action. For war in Iraq, four dimensions explained about 63.45% of variance, and for eliminating affirmative action, 58.39% of variance. Table Reliability for Cognitive Entrepreneurship in Problem Solving (n = 1380) Construct i q liminating ffirmative Action dimens ur sub nali cipa mensio va es 2. 10 War n Ira E A Alpha .6 .74 Cronbach?s 6 2.55 2 2 % 2 (27.78%) .71 ( ) 4.7 2 2 % 3 32.53 ( 1.10 ) 2.4 (14. %) 1 1 % 9 (8.77%) .37 ( 1.29 ) 1.4 1 (8 8 (7.52%) Eigenvalues .01 .41%) 1.2 % of Variance Explained 8.39 Cumulative 45 63. 5 205 206 Table 11 Items, Factor Loadings, and Reliability for Cognitive Retrogression (n = 1380) War in Iraq m ng c Eli inati Affirmative A tion Construct Item c ad Alpha Factor L n Fa Lo tor ing oadi g Alpha I now ow I houl behk h s d a r this p e 6ve fo robl m. . 3 .58 It is too late to change th pr blem e position I now have on this o . .80 I have spent too much tim p ition now. e on this problem h os to c ange my .68 .77 It s too ate t shak the pr blem i l o e c usion I e n t o . oncl hav draw for his .66 .74 I ve ade forts o juha m ef t st y deci o s l .58 ify m sion n thi prob em. I will k p m curr nt po it with contradictor ee y e sition even if m y evidence. so eone challenges .71 .71 I ve f und ount evi in ha o c er dence that rej t os s different from m e. 62 ects he p ition . Cognitive e re n fe is tl c e my mi 2 . R trog ssio I el it cos y to hang nd on this problem. .7 .74 82 E va .53 igen lues 2 3.18 % of a Explained 2.08 9 4 Vari nce 52.9 207 Table 12 Items, Factor Loadings, and Reliability for Cognitive Suspension (n = 1380) m e on .70 .72 .75 .69 War in Iraq Eliminating Affir ativ Acti Construct Item L o g l ct Loading Factor adin A pha Fa or Alpha I t k re time o ak ng up my p e .77 .73 wan to ta e mo robl m. bef re m i mind for this I ta e up my nd about what should be done for this problem .68 .61 .81 .79 2.20 2.07 hesi te to mak . mi I to thin fo fi ze p o t p e need more time robl m. k be re I nali my ositi n on his Cognitive Suspension or th s pro lem, vide ce is n. F i b I will try to spen n g t a e e n i su d a y jud men until ll th Eigenvalues % of Variance Explained 54.90 51.73 As I did for CAPS, I conducted subsequent analyses for cognitive retrogression and cognitive suspension in which I adopted a multiple-item measurement strategy. .80. sed four items a , ore reneurship in problem Table 11 summarizes the items, loadings, Cronbach?s alpha, and variance explained in two problems for cognitive retrogression. In cognitive retrogression, I used eight items and found single dimensionality in both issues: i.e., Eigenvalues 2.53 for war in Iraq and 3.18 for eliminating affirmative action. Factor loadings varied from .58 to The amount of variance explained was 42.08 and 52.99% and the Cronbach?s alphas were .72 and .82, correspondingly. Table 12 summarizes the items, loadings, Cronbach?s alpha, and variance explained in two problems for cognitive suspension. In cognitive suspension, I u nd again found only one dimension in both issues: i.e., Eigenvalues 2.20 for war in Iraq and 2.07 for eliminating affirmative action. Factor loadings vary from .61 to .81. The amount of variance explained was 54.90 and 51.73% and the Cronbach?s alphas were .72 and .69. Antecedent Variables of the CAPS and CEPS The situational theory of publics (J. Grunig, 1997, 2003) specifies situational antecedent conditions explaining when and why a communicant actively seeks or passively processes information. They are problem recognition, constraint recognition level of involvement, and referent criterion. In addition, I introduced a variable called situational motivation in problem solving that conceptualizes and measures one?s problem-solving motivation in a problematic situation. The primary purpose of the present study is to generalize the situational theory of publics by introducing two m general constructs of communicant activeness and cognitive entrep 208 209 solving. Thu uld lics (J. Grunig, 1997). I found only one dimension in all three problems: i.e., Eigenvalues = 2.17, 2.11, and 2.11 for war in Iraq, losing weight, and eliminating affirmative action, respectively. Factor loadings varied from .82 to .87. The amounts of variance explained are 72.29%, 70.40%, and 70.33% and the s, it is necessary to investigate how the existing antecedent conditions wo conceptually link with the two new variables. In doing so, I should note that two antecedent variables of problem recognition and referent criterion have been conceptually refined and thus their measurement items have been modified. Hence, to proceed, I checked the new and modified items? reliabilities and internal consistencies. In the following Tables 13-17, I will summarize the analysis and findings for the situational antecedent variables. Problem recognition. Table 13 summarizes the items, loadings, Cronbach?s alpha, and variance explained in three problems for problem recognition. I used four items that had been modified from the past measurement items in the situational theory of publics (J. Grunig, 1997). I found only one dimension in all three problems: i.e., Eigenvalues = 2.26, 2.29, and 2.26 for war in Iraq, losing weight, and eliminating affirmative action. Factor loadings varied from .63 to .82. The amounts of variance explained are 56.50%, 57.21%, and 56.45% and the Cronbach?s alphas were .74, .75, and .74, correspondingly. Therefore, I conclude that problem recognition reached an acceptable level of reliability and internal consistency. Constraint recognition. Table 14 summarizes the items, loadings, Cronbach?s alpha, and variance explained in three problems for constraint recognition. I used three items as in the situational theory of pub 210 Cronbach?s alphas were .81, .79, and .79, correspondingly. Therefore, I conclude that constraint recognition reached an ity and internal consistency. Level of involvement. Table 15 summarizes the items, loadings, Cronbach?s alpha, and v c plained in three pro s for level of involvement. I used four items as in the si o of lics (J. Grunig, 1997). I found only one dimension in all three problem : i envalues = 2.07, 2.26, and 2.32 for war in Iraq, losing weight, and eliminating affirmative action, respectively. Factor loadings varied from .61 to .82. The amounts of variance explained are 51.71%, 56.54%, and 58.04% and the Cronbach?s alpha r an , e di . efore, I conclude that level of ent reached an acceptable level of reliability and internal consistency. Referent criterion. Table 16 summarizes the items, loadings, Cronbach?s alpha, s for referent criterion. I used five items that had odified from s in the situational theory of publics. For e d). In the other two problems, I found a ension: i.e., E (53.44% of variance explained) and 2.47 ight and eliminating affirmative action .64 to .77 (in the first dimension) and the spondingly. The two dimensions in the ension was maller than the first d addition, in the other two problems, I found ensions. Therefore, with some , I de that referent criterion cceptab ncy. acceptable level of reliabil arian tuati s e ex blem nal .e., theo Eig ry pub s we e .69, .74, d .76 corr spon ngly Ther involvem and variance explained in three problem been m war in Iraq, I found two dim explained) and 1.03 (20.64% of variance explaine single dim (49.35% of variance explained) for losing we correspondingly. Factor loadings vary from Cronbach?s alphas were .76, .78, and .81, corre war in Iraq issue could be an issue speci relatively s single dim reached an a the past measurem nsions: i. ime ent item e., Eigenvalues 2.57 (51.39% of variance fic result. Yet, the second dim nsion. In igenvalues = 2.67 caution conclu le level of reliability and internal consiste 211 Table 17 summarizes the items, loadi ha, and ia e ems for situational motivation in problem solving. I used three items. I found a single dimension in all three problem of variance explained), 1.97 (65.65% of varia xpla and 1.62 (53.99% of variance explained) for war in Iraq, losing weigh d el affirma a , .71 to .85 and the Cronbach?s alphas were .67, .74, and .57, correspondingly. ysis and reliability analysis helped identif l n nt and reliable measurement items. s were reliab lly consistent in most of the problems with a few es ensionality of the construct. T ean the construct has a temeyer, Bearden, & Sharma, tiveness in problem solving to have six ensions as a second-order construct. ensional structure, especially a second- confirmatory factor is that tested the d a priori conceptual Situational motivation in problem solving. ngs, Cronbach?s alp var nce xplained in three probl s: i.e., Eigenvalues 1.81 (60.34% nce e t, an ine imi d), nating tive ction respectively. Factor loadings varied from In summ y the dim ary, the exploratory factor anal ensionality and the interna ly co siste Item exceptions. However, reliability analys unidim m 2003). Indeed, I conceptualized communicant ac subdim communicant activenes order structure. In the following section, I analys models. le and interna is such as Cronbach?s alpha assum hus, although alpha level would be high, it does not ensionality (Ne Thus, it is necessary to find whether im report the results from chieved un s has such a multi-d imensionality of the new construct using idim Table 13 Items, Factor Loadings, and Reliability for Problem Recognition (n = 1380) War in Iraq Losing Weight Eliminating Affirmative Action Construct Item Factor Alpha Factor Loading Loading Alpha Factor Alpha Loading To what extent do you think there is something missing in this problem? .79 .76 .72 How much does the current situation differ from your expectation? .63 .73 .75 How strong do you feel that something needs to be done to improve the situation for this problem? .79 .82 .80 .78 .74 .70 .75 .74 .74 Problem Recognition How much does the current situation deviate from what you think it should be? Eigenvalues 2.26 2.29 2.26 % of Variance Explained 56.50 57.21 56.45 212 Table 14 Items, Factor Loadings, and Reliability for Constraint Recognition (n = 1380) War in Iraq eight Affirmative Losing W Eliminating Action Construct Item Factor oad Alpha Factor ad Alpha Factor ad Alpha L ing Lo ing Lo ing Please think of whether you, that would make a difference in the way these problems are .84 .84 .85 personally, could do anything handled. If you wanted to do some ma thing, would your efforts ke a difference? (R) To what extent do you believe this problem is a problem that you can do something about? (R) .87 .85 .84 Constra ecogni uld a the w you wanted (R) .83 .79 int R tion To what extent do you believe that you co this prob ffect ay lem is eventually solved if to? .84 .81 .82 .79 Eigenvalues 2.17 2.11 2.11 % of n 70.33 ce Varia Explained 72.29 70.40 213 214 q t Table 15 Items, Factor Loadings, and Reliability for Level of Involvement (n = 1380) War in Ira Losing Weigh Eliminating Affirmative Action C a r g onstruct Item Factor Loading Alph Facto Alpha Loadin Factor Loading Alpha In n your mind, how much of a connection do you see betwee yourself and this problem? .61 .74 .73 To what extent do you believe this problem could involve y someone close to you at some point? ou or 2 .77 .7 .78 H ct y 2 ow much do you believe this problem affects or could affe ou personally? .78 .8 .79 Inv or thoughts are about this problem? 69 4 6 Level of olvement How strong would you say your opinions .71 . .72 .7 .75 .7 Eigenvalues 2.07 2.26 2.32 % of Variance Explained 51.71 56.54 58.04 215 Table 16 Items, Factor Loadings, and Reliability for Referent Criterion (n = 1380) inating ff ti ction War in Iraq Losing Weight Elim A irma ve A Constr Factor Loading ctor d ph o in A a a a lpha uct Item Fa Loa ing Al a Fact r Load g lph F ctor Lo ding A I a confidenm t t problem. 7 1 4 . abou my knowledge about this . 5 -. 3 .6 69 I s ongl sup thi prob em. tr y po c n s g s l 7 4 7 . rt a ertai way of re olvin . 6 -. 1 .7 74 I h ve a refer be settled. a p e fo w p e should 7 4 7 . nce r ho the robl m . 5 -. 5 .7 77 I a pre y surm tt e n o s t problem. 6 .5 3 .73 , I k ow h w to olve his . 8 2 .7 Refer Criter st e e p d e id es s g le 6 .6 4 .56 ent ion Pa exp rienc gu elin for has rovi ed m with olvin this prob m. . 4 1 .7 Eigenvalues 2 7.5 1.0 6 7 2.3 .7 2.6 .78 47 .81 % o Varia Explained . 20. 44 9f nce 51 39 64 53. 4 .35 Table 17 Items, Factor Loadings, and Reliability for Situational Motivation in Problem Solving (n = 1380) War in Iraq i e Act Los ng W ight Eliminating Affirmative ion Construct Item Fa p o i a Factor oading lpha ctor Loading Al ha Fact r Load ng Alph L A How often do you stop to think about each of thes re problems? .75 e th e .80 .71 To what extent would you say you are curious ab hi problem? .83 out t s .85 .82 Situational Motivation in Problem Solving Please, indicate how much you would like to und n each of these problems better. .7 7 8 .68 .57 ersta d 5 .6 .7 .74 Eigenvalues 1.81 1.62 1.97 % of Variance Explained 53.99 60.34 65.65 216 217 quation Modeling Analysis In the prev s section, I reporte e characteristics o e sample and results of exploratory analysis of the variables in eed to the confirma i hase tw ne cons t C PS d C sing the EQS .1 a B er, 5 a z the x the CAP and CEPS. Then, I analyzed a series of causal models in which I specified con tual relationships with other conceptual variables (e.g., level of involvement) to . In t e e n, I w irst report lts from confirmatory factor analysis of CAPS. Problem onfirmatory c A ysis of C ne ale p rrent study, I have a large sample size (n = 1,380). Thus, using a sub-sampling strategy, s nction. e , ss-v a f a sca is st al ed b g samples different populations. Although the sub-sampling strategy has a limitation in ity of the measurement system and the factor model. The e s ple n 4 about 33% of 1,380 survey participants. Following the two-step process of SEM, I co atory is that allowed covarying all th odels using the candidate easurement ecking for poor or cross-loading items and 6.1 produced a set of best item Structural E iou d th f th owcluded in the study. I n proc tory analys pr s p entl for 200 the ), I o naly w ed truc data s of to e A amine factor structures of an EPS. U 6 ogr m ( S cep test the nomological validity of CAPS and CEPS the resu C cu I drew a few sm fu from conducting a cross-validation anal structure and conceptual valid first developm analys m residual covariances with LM tests in EQS he n xt s ctio ill f Communicant Activeness in Solving (CAPS) Fa sirable to te tor nal st a APS eIt is de w m asurement sc with different sam les. In the aller developmental amples by using SPSS 11.5 random selection How ver cro alid tion or new le be ev uat y usin ysis, it still allows a better opportunity to evaluate the ntal am had = 67, e factors in the m first conducted the nfirm items. Subsequent analysis ch s across three 218 problems in the current study: war in Iraq, losing weight, and eliminating affirmative action. he in l ph , factor oblique, and second-order structure model. These three models were nested hierarchically under a six-factor oblique m l. Specifically, the one-factor model is the onious of the three models compared. Thus, I first compared the one-factor o th f r iqu odel. And, e s c o u odel turned out be a statistically better model, I compared the six-factor oblique model against the which is mo onious than the six-factor oblique o I summarized the results from model testing in Tables 18-20. Table 18 reports the o goo ces an T n, the structura ase I imposed three factor structures: one factor, six- ode most parsim m to second-order structure model, m m del wi the six- acto obl e m if th ix-fa tor bliq e m re parsim del. del dness-of-fit indi d model comparison results for the war in Iraq issue. 219 ? 2 f Table 18 Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (War in Iraq) Model d ? 2 / df RMSEA SRMR CFI NNFI AIC ? df ? ? 2 p Independence 4 0 4574.10 21 21.781 4154.104 One-Facto ) r (Md A 705.166 . e 2 . r 9 . e 6 0 169 4.173 .083 (.076, .089) . 155 .877 .847 367.166 Six-Factor Obliqu (Md B) 396.32 154 2.574 058 (.051, .065) . 062 .944 .924 88.322 A vs. B 308.844 15 p < .001 Developmental Sample (n = 467) Second-Orde C) (Md 399.52 159 2.513 057 (.050, .064) 063 .945 .927 81.529 B vs. C 3.207 5 p > .250 Independenc Model 4529.27 21 21.568 4109.276 One-Facto ) r 6 . e 3 . r 6 . e 3 0 (Md A 785.47 169 4.648 .089 (.083, .096) . 162 .857 .823 447.476 Six-Factor Obliqu (Md B) 395.63 154 2.569 059 (.051, .066) . 056 .944 .924 87.633 A vs. B 389.843 15 p < .001 Validation Sample (n = 458) Second-Orde C) (Md 408.36 159 2.568 059 (.052, .066) 057 .942 .924 90.366 B vs. C 3 12.73 5 p =.025 Independenc Model 4725.84 21 22.504 4305.843 One-Facto ) r 5 . e 3 . r 2 . p e 3 0 (Md A 755.17 169 4.468 .087 (.081, .093) 168 .870 .839 417.175 Six-Factor Obliqu (Md B) 401.76 154 2.609 059 (.052, .066) . 055 .945 .925 93.763 A vs. B 353.412 15 p < .001 Validation Sample (n = 460) Second-Orde C) (Md 408.32 159 2.568 058 (.051, .065) 056 .945 .927 90.322 B vs. C 6.559 5 > .250 Independenc 13239.66 21 63.046 12819.663 r 9 . Total (n = 1380) Second-Orde 0823.50 15 5.179 .055 (.051, .059) 055 .949 .933 0505.50 220 Table 19 CAPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (Losing Weight) Model 2 ? 2 I ? ? 2 p ? df / df RMSEA SRMR CF NNFI AIC df ? p c .2 1 1 3 Inde enden e 4529 53 210 2 .568 4 09.25 On Fact ( d A) x-Fa or Ob ( d B) l 7) Seco d-Or ( d C) Inde enden odel e- or M .480 17 3. 4 .0 0, ) .1 68 29 Si ct li M .352 15 2. 1 .0 2, ) .0 45 22 v 0 1 p < .001 Developmenta Sample (n = 46 n de M .799 16 2. 3 .0 1, ) .0 46 16 4. 5 p > .250 p c M .8 0 9 4 642 3 71 76 (.07 .083 55 .891 .8 6.480 que 338 8 14 49 (.04 .057 48 .958 .9 .352 A s. B 3 4.128 5 r 342 3 10 49 (.04 .056 49 .958 .9 .799 B vs. C 447 e 4367 14 210 2 .799 3 47.81 On Fact ( d A) x-Fa or Ob ( d B) 8) Seco d-Or ( d C) Inde enden odel e- or M .706 17 3. 9 .0 9, ) .1 69 27 Si ct li M .861 15 2. 1 .0 4, ) .0 39 31 7 1 p < .001 Validation Sample (n = 45 n de M .788 16 2. 4 .0 5, ) .0 37 39 17 5 p ?.005 p c M .8 0 8 5 622 3 59 75 (.06 .082 44 .892 .8 6.706 que 347 8 01 51 (.04 .058 46 .954 .9 .861 A vs. B 2 4.845 5 r 365 3 24 52 (.04 .059 48 .951 .9 .788 B vs. C .927 e 4316 55 210 2 .556 3 96.83 On Fact ( d A) x-Fa or Ob ( d B) 0 ) Seco d-Or ( d C) Inde enden e- or M .352 17 3. 3 .0 3, ) .1 88 20 Si ct li M .205 15 1. 2 .0 9, ) .0 50 -2 3 1 p < .001 Validation Sample (n = 46 n de M .431 16 1. 1 .0 8, ) .0 52 -9 3. 5 p > .500 p c 2. 0 1 9 552 3 19 69 (.06 .076 48 .908 .8 6.352 que 313 8 98 46 (.03 .054 44 .962 .9 .795 A vs. B 2 9.147 5 r 316 3 94 45 (.03 .053 44 .963 .9 .569 B vs. C 226 e 1261 329 210 6 .059 12 92.32 Total (n = 1380 n de .8 3. 8 .0 2, ) .0 52 30 ) Seco d-Or r 628 38 163 85 46 (.04 .049 42 .962 .9 2.838 221 Table 20 CAPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (Eliminating Affirmative Action) M ? 2 f f RMSEA SRMR CFI F ? p odel d ? 2 / d NN I AIC ? df ? 2 Ind ende ce 515ep n 1. 0 2 4 73 735 21 24.53 731. 5 O -Fact r Md A) 80 ix-F or O ique Md B) 38 pmen l n = 7) Second-Order Md C) 38 Ind ende ce odel 471 ne o ( 2. 8 (.0 09 9 .8 .8 S act bl ( 3. 5 (.0 06 6 .9 .9 78 vs. B 4 02 p < .001 Develo ta Sample ( 46 ( 6. 4 (.0 06 6 .9 .942 10 C 2 p > .250 ep n M 4. 0 8 4 07 480 175 4.586 .08 82, . 4) .17 73 278 160 2.395 .05 48, . 2) .05 55 610 163 2.372 .05 47, . 1) .05 55 072 21 22.44 48 452.480 41 63.2 A 19.2 15 60.6 B vs. 3.33 3 294. 2 O -Fact r Md A) 69 ix-F or O ique Md B) 38 ation n = 8) Second-Order Md C) 38 Ind ende ce odel 494 ne o ( 8. 1 (.0 08 8 .8 .8 S act bl ( 8. 6 (.0 06 8 .9 .9 16 vs. B 3 31 p < .001 Valid Sample ( 45 ( 9. 5 (.0 06 8 .9 .935 10 C p > .750 ep n M 3. 0 8 4 07 447 175 3.991 .08 75, . 7) .15 84 616 160 2.429 .05 49, . 3) .04 49 610 163 2.390 .05 48, . 2) .04 50 076 21 23.53 61 348.447 33 68.6 A 09.8 15 63.6 B vs. .994 3 523. 6 O -Fact r Md A) 73 ix-F or O ique Md B) 41 ation n = 0) Second-Order Md C) 41 Ind ende ce 1414 ne o ( 5. 4 (.0 09 1 .8 .8 S act bl ( 1. 9 (.0 06 9 .9 .9 13 vs. B 3 28 p < .001 Valid Sample ( 46 ( 8. 8 (.0 06 0 .9 .930 46 C 3 p ?.05 ep n 5 0 0 1 5.572 941 175 4.205 .08 77, . 0) .18 81 613 160 2.573 .05 52, . 5) .04 47 846 163 2.570 .05 52, . 5) .05 46 .572 21 67.36 58 385.941 30 91.6 A 24.3 15 92.8 B vs. 7.23 3 372 = 1380) Sec nd-O er 77 Total (n o rd 6. 37 2 (.0 05 .9 3 163 4.763 .05 49, . 6) .046 .956 43 450.337 In the developmental sample (n = 467), the one-factor model was a significantly better model than the independence (null) model: ? 2 df (169) = 705.166 vs. ? 2 df (210) = 4574.104 and ? 2 /df ratios of 4.173 vs. 21.781. Next, I compared the one-factor model with th e-e six-factor oblique model. The six-factor oblique model was better than the on factor model: ? 2 df (169) = 705.166 vs. ? 2 df (154) = 396.322 and ? 2 /df ratios of 4.173 vs. 2.574. The nested model test showed that the six-factor oblique factor model is a better representation of the observed covariances structure: ?? 2 ? df (15) = 308.844, p < .001. According to Hu and Bentler?s (1999) joint criterion: i.e., CFI ? .96 and SRMR ? .10 or RMSEA ? .06 and SRMR ? .10, the six-factor oblique model is close to an acceptable fi in terms of the joint criterion: CFI = .944, SRMR = .062, and RMSEA = .058. Howev the CAPS model is designed to be a second-order structure that is more parsimonious. Hence, I conducted the nested model testing between the six-factor oblique and second- order factor models. Comparison showed the second-order factor model to be better than t er, the six- e 01. factor oblique model: ? 2 df (154) = 396.322 vs. ? 2 df (159) = 399.529 and ? 2 /df ratios of 2.574 vs. 2.513. In addition, the nested model testing showed that the second-order model is more parsimonious than the six-factor oblique model: ?? 2 ? df (5) = 3.207, p > .250. The second-order factor model is also close to an acceptable fit in terms of the joint criterion: CFI = .945, SRMR = .063, and RMSEA = .057. Thus, I conclude that th proposed second-order model is a more viable model structure for the data. 23 Table 19 summarizes the model comparisons for the war in Iraq problem. In the losing weight problem with the development sample (n = 467), the six-factor oblique model was significantly better than the one-factor model ?? 2 ? df (15) = 304.128, p < .0 23 I added error covariances and disturbance covariances in modeling when they were conceptually reasonable. Thus, in some models, the degrees of freedom were slightly different even in the same factor models (e.g., the second-order factor models) across three problems. 222 The six-factor oblique model ? 2 /df ratio was 2.141. It was close to an acceptable fit terms of the joint criterion: CFI = .958, SRMR = .048, and RMSEA = .049. And, the nested model in testing between the six-factor and second-order factor model showed the gain to be statistically not worse: ?? 2 ? df (5) = 4.447, p > .250. The model f re for the el to be statistically not worse: ?? 2 ? df (3) = 3.332, rder odel. % of the rawn. I er factor wo odels ere better than the single-factor model. In the first validation sample (n = 458), however, second-order model a it indices were CFI = .958, SRMR = .049, and RMSEA = .049. Thus, I conclude that the proposed second-order model is statistically a more viable model structu data. Table 20 reports the model comparisons results for eliminating affirmative action as done in previous two problems. In the development sample (n = 467), again the six- factor oblique model is significantly better than the one-factor model ?? 2 ? df (15) = 419.202, p < .001. The six-factor oblique model ? 2 /df ratio was 2.395. It was close to an acceptable fit in terms of the joint criterion: CFI = .955, SRMR = .056, and RMSEA = .055. And the nested model testing between the six-factor and second-order factor model again showed the second-order mod p > .250. The model fit indices were CFI = .955, SRMR = .056, and RMSEA = .054. Thus, from the developmental samples, I conclude that the proposed second-o model structure better represents the observed data than the six-factor oblique m Next, I proceeded to the validation samples, which randomly selected 33 total sample (n = 1,380). Two validation samples of n = 458 and n = 460 were d conducted the same model tests of one factor, six-factor oblique, and second-ord models. The results of model comparisons are summarized in Tables 16-18. In t validation samples for the war in Iraq, both six-factor and second-order factor m w 223 the second-order structure was statistically worse than the six-factor structure: ?? 2 ? df (5) = 12.733, p = . 025. In the second validation sample (n=460), the second-order factor : ?? 2 ? df (5) = 6.559, p > .250. Similarly, in the losing weight problem, both the six-factor and second-order factor models were better than the single-factor model. In the first validation sample (n = 458), again, the second-order structure was statistically worse than the six-factor structure: ?? 2 ? df (5) = 17.927, p ? .005. However, in the second validation sample (n = 460), the second-order factor model was not worse: ?? 2 ? df (5) = 3.226, p > .500. Finally, both the six-factor and second-order factor models were better than the single-factor model in the eliminating affirmative action issue. In the first validation sample (n = 458), the second-order factor model was statistically not worse: ?? 2 ? df (3) = .994, p > .750. In the next validation sample (n = 460), the six-factor structure was a better model representation than the second-order structure: ?? 2 ? df (3) = 7.233, p ? .05. In the validation samples, in three out of six tests, the six-factor oblique structure odel better represented the data than the second-order structure, whereas the other three tests showed that the second-order factor models better represented the covariance structures. Despite the three tests favorable to the six-factor oblique structure, six out of nine nested model tests indicated the second-order structure as a better representation. The inconsistency (i.e., three tests out of nine) seems to be due to sampling fluctuation. In addition, the three tests favoring six-factor oblique models were close to non-significant and had almost identical model fit values in terms of joint-criterion. This suggests that the second-order structure is a viable representation of the data. In summary, I found six subdimensions in the CAPS observed datasets from the analysis. This provides evidence model was statistically not worse m 224 225 ation permitting, ation forwarding, information sharing, information seeking, and information processing. odels H1: The higher the communicant activeness in problem solving, the higher the information forefending. H2: The higher the communicant activeness in problem solving, the higher the information permitting. H3: The higher the communicant activeness in problem solving, the higher the information forwarding. H4: The higher the communicant activeness in problem solving, the higher the information sharing. H5: The higher the communicant activeness in problem solving, the higher the information seeking. H6: The higher the communicant activeness in problem solving, the higher the information processing. that there are six sub-constructs, information forefending, inform inform Hypothesis Testing In the development samples, I found the second-order factor structure to be the best structure for the new CAPS construct. To test the hypotheses regarding the subdimensions of CAPS, I used the total sample (n = 1,380) across the three issues in the study. Tables 16-18 summarize the overall model fits in three problems. For war in Iraq, ? 2 df (159) = 823.500, CFI = .949, SRMR = .055, and RMSEA = .055. For losing weight, ? 2 df (163) = 628.838, CFI = .962, SRMR = .042, and RMSEA = .046. For eliminating affirmative action, ? 2 df (163) = 776.337, CFI = .956, SRMR = .046, and RMSEA = .052. Thus, for all three models the second-order structure is a reasonable representation for CAPS. Next, I proceed to interpreting the parameter estimates in the second-order m to test the hypotheses I proposed earlier. To conceptualize the new construct of communicant activeness in problem solving, I proposed six hypotheses (see Figure 5 in the conceptualization). They are: Figures 15-17 summarize the three second-order models the standardized structural path ates and standardized loadings in three problems. Fig ar in Iraq). and estim ure 15: CAPS (W Fig : CAPS (Losing Weight). ure 16 226 Figure 17: CAPS (Eliminating Affirmative Action). ssess t, ?a to le 21 summarizes the variance explained and the coefficient H of CAPS and its subdimensions. For those items in the CFA models, I examined the extracted variance to a construct validity and the coefficient H for construct reliability. The coefficient H is a better alternative indicator of construct reliability than other previous measures based on the composite score with the equal weight assumption (Hancock & Mueller, 2001). Specifically, H is not affected by the loading sign, additional indicators can never detrac and it is never smaller than the reliability of the best indicator, which is logical in that factor inferred from multiple indicator variables should never be worse (i.e., less reliable) than the best single indicator alone? (Hancock & Mueller, 2001, p. 195). For the extracted variance, 50% or above is considered good, whereas the coefficient H is considered good with .70 to .80 or above. Most subdimensions were higher or close 50% of variance explained and close to or above .70 in the coefficient H. More importantly, communicant activeness in problem solving, which is the focal construct, exceeded 50% of the extracted variance and was above .95 on coefficient H. Tab 227 228 Table 21 Variance Explained and Coefficient H in CAPS L g ig minatin fi iv tion War in Iraq osin We ht Eli g Af rmat e Ac Variance l H an E i Varianc xplaineExp ained Vari ce xpla ned H e d H E Information f ng 3. 4 3 8 7 . Fore endi 5 93 .8 1 51. 9 .81 53.9 820 Infor atiom n m g 1. 8 3 3 8 . Per ittin 3 37 .5 8 31. 9 .58 29.9 590 Information w g 1. 1 5 1 4 . For ardin 5 46 .8 9 46. 7 .78 50.3 803 Inform io 2. 7 1 4 7 . at n Sharing 5 64 .7 6 56. 6 .79 53.7 782 Inform o 6. 9 5 3 6 . ati n Seeking 4 56 .7 4 54. 0 .89 57.6 852 Information c g 8. 5 4 8 6 . Pro essin 4 33 .7 0 40. 6 .67 27.0 558 C mmu ican P ble o n t i ss ro m 8. 5 5 7 7 . Act vene in Solving 5 73 .9 1 66. 2 .96 70.2 981 As shown in the figures, in all three problems (n = 1380), the structural paths from the second-order construct (i.e., communicant activeness) to first-order constructs were al d g cant (i.e., tiveness in problem s that a problem solver with heightened motivation in problem solving ers ?s eeking and processing increases as one becomes more active in problem solving. In summary, from the data analysis and hypothesis testing, I found six l significant at the p < .001 level. The loadings in the observed variables were all significant at the p < .001 level. Most of the parameter estimates in the measurement items varied between .313 and .826. The core premise in the CAPS model is that as a communicant becomes active in problem solving, he or she will become more selective, transmissive, and acquisitive regarding information about the problem. The analysis of CAPS in three individual an social problems provides supportive evidence for the main postulate and the six a priori hypotheses. As a communicant becomes active in problem solving, one?s information selectivity tends to increase (i.e., H1 and H2). In general, one?s information forefendin tendency increased more than his or her information permitting tendency as communi activeness heightens. Interestingly, in all three problems, information transmission H3 and H4) was the most salient dimension in the communicant ac solving. This indicate will evolve to be more active in sharing and forwarding information with oth about the problem. The next salient dimension is information acquisition (i.e., H5 and H6). These two variables are the default dependent variables highlighting a public active and passive communication behaviors in the situational theory of publics (J. Grunig, 1997). Consistent with what past studies of the STP found, I found that information s 229 230 conceptual subdim sed two formulas that are conceptually derived for cognitive multilateralism and cognitive commitment. Specifically, I measured the latter two subdimensions by using participants? agreement with factual and evaluative statements of conflicting position about the problems. Thus, these two dimensions were measured by single items. I avoided nested model testing between one-factor, multiple-factor oblique, and second- order factor models because of modeling and programming difficulties, although it would have been possible to conduct these tests. Instead, I constructed and tested second-order factor structure directly whether it achieves a reasonable model fit across issues. Hence, if the model tests showed the second-order factor models to be viable, I considered these tests to be supporting evidence for the proposed second-order factor structure for CEPS. As done in the confirmatory factor analysis of CAPS, I used randomly drawn subsamples: the developmental sample (n = 467, 33% of total n = 1,380) and the validating sample (n = 917, 66% of total n = 1,380). Table 20 and 21 report the CFA ensions of CAPS as proposed. Each subdimension has a positive relationship with its higher dimension of communicant activeness. In the following section, I will report another new construct I developed in this study. Cognitive Entrepreneurship in Problem Solving Confirmatory Factor Analysis of CEPS The second new construct is cognitive entrepreneurship in problem solving (CEPS). CEPS captures our varying cognitive approaches across different problem situations. It possesses four conceptual dimensions: cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspension. In the present study, I applied a multiple-item approach for cognitive retrogression and cognitive suspension. In contrast, I u 231 results in two subsamples for two issues: i.e., war in Iraq and eliminating affirmative acti First, in the developm ple (n = odels were statistically viable m rding 9) joint-criterion. For war in Iraq, the second-order model resulted in ? 2 df (29) = 58.935, CFI = .977, SRMR = .055, and RMSEA = .047. For eliminatin irmative action, the second-order model resulted ? e is a viable representation of the observed data. In the li n sample (n 7), I found si good model fit indices. For war in Iraq, the co order model resulted in ? df 154.220, C 55, SRMR = .067, and MSEA = .069. For eliminating affirmative action, the second-order model resulted in df (28) = 7123.734, CFI = .964, SRMR = .056, and RMSEA = .061. With the CFA sults, I consider the second-order factor representation a viable model structure for proble olving. Henc proceeded to interpret the eter estimates in the second-o mod the owing. For the hypothesis testing, I used the the model fits in the Tabl ? 2 df inating mative a found ? 2 df (28) = 175.969, CFI = .963, SRMR = .056, and RMSEA e to interp eter estimates within odels. on. ental sam 467), the second-order factor m odels, acco to Hu and Bentler?s (199 g aff in a second-order factor structur va se R ? 2 df (28) = 70.080, CFI = .969, SRMR = .059, and RMSEA = .057. I found, thus, that datio nd- = 91 milarly ) = 2 (29 FI = .9 2 re cognitive entrepreneurship in param Hypothesis Testing second-order CFAs. I reported (29) = 187.542, CFI = .961, SRMR = affir = .062. Thus, these results allowed m the m m s e, I rder el in foll total sample (n = 1,380) to conduct the e 19 and 20. For war in Iraq .061, and RMSEA = .063. For elim ction, I ret the specific param For those items in the CFA models, again I computed the Coefficient H for con b d racted vari a onstruct validity. Table 24 mmarizes the variance explained and the coefficient H of CAPS and its subdimensions. C nly e ogression and cognitive suspension were measured by ul items. Most subdimensions were lower than 50% of variance explained. Thus, rrent measurement items seemed t construct rather poorly. But, coefficient H c 70 o r oth C itive rship in problem solving, ain, had low variance explained. The coefficient Hs were .612 and .627, which were w an .70. I c d that the construct th resent study. H e, I interpret the findings more carefully related with the CEPS n t. struct relia ility an the ext ance to ssess c su In m cu is ag lo in co EPS, o cognitiv retr tiple o tap the lose to . r highe for b issues. ogn entrepreneu er th e p onclude enc CEPS has marginal validity and reliability struc 232 233 Table 22 CEPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (War in Iraq) Mo df f RMS R FI NNFI del ? 2 ? 2 / d EA S MR C AIC Indepen 45 3 13 dence 73.1013 1 0.5 1283.101 Developmental Sample (n = 467) Second- 29 .047 (.030, .064) 0 77 .965 .Order 5 8.935 2.032 . 55 .9 935 Independe 45 6 98 nce 12.39428 2.4 2722.394 Validation Sample (n = 917) Second-Order 4 9 .069 (.058, .079) .067 .955 .20015 2 5.317 .930 96.200 Independe 1 45 9 43 01 6 nce .446410 1.1 4 1.44 Total (n = 1380) Second-Order 7 9 .063 (.054, .072) .061 .961 .939 29.542 .54218 2 6.467 1 234 Table 23 CEPS Model Comparisons: Chi-square Differences and Goodness-of-Fit indices (Eliminating Affirmative Action) M l ? 2 df ? 2 / df RMSEA SRMR CFI NNFI AIC ode Independence 14 1.762 6 45 31.595 1331.766 pm n = 67) Second-Order 7 .080 D o ental a ( 4 0 28 2.503 .057 (.040, .073) .059 .969 .951 14.080 evel S mple Independence 27 4.740 6 45 60.105 2614.74 6 atio 17) Second-Order 12 .734 Val S mple id n a (n = 9 3 28 4.419 .061 (.059, .072) .056 .964 .942 67.734 Independence 40 0.587 2 45 90.457 3980.582 o n = 13 5 28 6.285 .062 (.053, .071) .056 .963 .941 119.969 T tal ( 80) Second-Order 17 .969 Table 24 Variance Explained and Coefficient H in CEPS War in Iraq Eliminating Affirmative Action Variance Explained H Variance Explained H Cognitive Retrogression 30.93 .667 39.38 .699 Cognitive Suspension 34.88 .853 36.27 .785 C .627 ognitive Entrepreneurship in Problem Solving 19.66 .612 19.96 Figures 18-19 summarize the two second order models and the standardized structural path estimates and standardized loadings in two problems. I proposed four hypotheses for cognitive entrepreneurship in problem solving. They are: H7: The higher the cognitive entrepreneurship in problem solving, the lower the cognitive retrogression H8: The higher the cognitive entrepreneurship in problem solving, the higher the cognitive multilateralism. H9: The higher the cognitive entrepreneurship in problem solving, the higher the cognitive commitment. H10: The higher the cognitive entrepreneurship in problem solving, the higher the cognitive suspension. 235 Figure 18: CEPS (War in Iraq). Figure 19: CEPS (Eliminating Affirmative Action). 236 As shown in the figures, in both problems (n = 1,380), the structural paths from the second-order latent variable (i.e., cognitive entrepreneurship) to the first-order latent variables (e.g., cognitive retrogression) were significant at the p < .001 level, ex path to cognitive commitment for the war in Iraq issue (i.e., p < .01). In addition, the loadings in the observed variables were all significant at the p < .001 level. Most of the parameter estimates in the measurement items varied between .400 and .700 (the lowest .285, the highest .909). The core premise in the CEPS model is that as one becomes active in problem solving, one will become less backward reasoning (cognitive retrogression), be more possessing of cognitive multilateralism, and show more commitment and more suspending in finalizing a solution. The analysis of CEPS in two social problems cept the provide m ns. ive action issue (i.e., -.258, p < .001). This would suggest that there is a stronger backward reasoning tend omega approach) for the war in e, as expe s supportive evidence for the main postulate and the four a priori hypotheses. In H7, I conceptually predicted that the one?s cognitive retrogression in proble solving will increase as the cognitive entrepreneurship in problem solving heighte Cognitive retrogression can be highlighted by one?s backward reasoning?i.e., ?a conclusion directs certain evidence.? This is an optimization process for a priori conclusion. For H7, I found supporting evidence in both issues. Notably, the magnitude of the path coefficient was higher for the war in Iraq issue (i.e., -.754, p < .001) than the eliminating affirmat ency (i.e., cognitive Iraq issue. However, in both issues, the directions of the path were consistently negativ cted. 237 In H8 and H9, I adopted a single-indicator approach in measurement. I applie derived formulas for cognitive multilateralism and cognitive commitment. In H8, I predicted that as cognitive entrepreneurship increases, one will more possess cognitiv breadth?a greater number of decision alternatives and a greater extent of tolerance in dealing with competing information during the problem-solving process. I found sup for this prediction in both issues: .184 in war in Iraq, and .334 in eliminating aff d the e port irmative action. e a the i.e., - I ognitive suspension refers to a problem solver?s nitive resources in evaluating and reevaluating a selected ative action issue. ore nd In H9, I expected that as cognitive entrepreneurship heightens one will mak greater cognitive commitment?a greater degree of enthusiasm and extent of patronizing of the proposed solutions for a given problem solving. I found support for the H9 in eliminating affirmative action issue (.166, p < .001), but not in the war in Iraq issue ( .113, p < .01). Interestingly, the signs were opposite in the war in Iraq issue, which suggests, for the war in Iraq issue, that as one becomes more entrepreneurial in problem solving, one would be less enthusiastic and patronizing to the proposed solutions. Thus, found partial support for the cognitive commitment. In H10, I proposed that as one?s cognitive entrepreneurship increases, one?s cognitive suspension increases as well. C heightened willingness to invest cog solution before finalizing it. In the two issues, I found supporting evidence for H10: .414 (p < .001) in the war in Iraq issue and .770 (p < .001) in the eliminating affirm Overall results in the CEPS CFA model tests indicated that as the problem solver?s entrepreneurial mindset in problem solving heightens, one tends to adopt m backward reasoning, have cognitive breadth, and show tolerance for competing ideas a 238 239 s?seems to differ across issues. Considering the single-item approach with formulas in the cognitive multilateralism and cognitive commitment, such a finding requires further analysis. A new study using a multiple-indicators approach, as in the cognitive retrogression and suspension, will allow a clearer understanding for the two dimensions measured with a single item. In summary, I proposed two new variables of CAPS and CEPS to develop a more general version of the situational theory of publics (J. Grunig, 1997, 2003). Although the previous confirmatory factor analyses provided some confidence in the new constructs, it is hardly conclusive in terms of their construct validity. However, construct validity is hard to test or establish by the newly developed concept alone. One alternative way to demonstrate validity of a construct is to examine its conceptual relationships with other established constructs?a ?nomological network? (Cronbach & Meehl, 1955). Specifically, researchers can evaluate the validity of a new construct when they test ?distinct antecedent causes? and find ?consequential effects and/or modifying conditions, as well as quantitative differences in the degree to which a construct is related to antecedents or consequences? (Netemeyer, Bearden, & Sharma, 2003, p. 82; Iacobucci, Ostrom, & Grayson, 1995; Nunnally & Bernstein, 1994). Thus, it is necessary to test how and to what extent the two new constructs are explained by the antecedent variables in the situational theory. In practice, structural equation modeling provides an ideal methodological framework to test nomological validity (Bollen, 1989, Hoyle, 1995). opinions and to suspend judgment before finalizing a solution. Yet, cognitive commitment?the degree of enthusiasm and the extent of patronizing the proposed solution Earlier in the zation, I posited conceptual relationships between the new constructs and existing antecedent variables in the situational theory of publics. si ruc q ion li p , I tested the presumed conceptual lationships between new and existing variables. Specifically, I examined how the newly tr c s co a our antecedent variables from the tuational theory of publics. In the following section, I will report a series of models of m gical ne rks betwe e antecedent variables of tuational theory u lysis tuational Theory Communicant Activeness in Problem Solving (SITCAPS) Model T i int t is st i generalize the situational theory of publics TP) by introducing CAPS. In the earlier exploratory and confirmatory factor analysis, I un v S liable for the three dividua social pr ms. To further explore the validity and utility of the new n th model with four antecedent variables in the STP oblem recognition, constraint recognition, level of involvement, and referent criterion. a io in uce tion a e, situational motivation in problem lv (s i 11 h ma odel, Chapter II: Conceptualization). Table 25 reports the SITCAPS model testing results. According to Hu and entler?s (1 t crit , all three ls reached an acceptable level of model fit. r the ndic (154) = 927.538, CFI = .930, SRMR .077, and EA = .06 espite a r ly low CFI value, the combination of eets the recommended joint criterion (i.e., either CFI ? .96 and conceptuali U re in si no si Si (S fo in co pr In so B Fo = SRMR and RMSEA m ng a st tural e uat mode ng a proach oduced oncept are nceptu lly linked with the f olo two en the two new constructs and th . of Str ctural Ana he pr mary eres of th udy s to d that the new ariable, CAP , is internally consistent and re l and I con oble caucept, structed a sal pa ddit ing n, I ee F trod gure d a m for t otiva e sum al v ry m riabl 999) join aq, RMS erion fit i 0. D mode re ? war in Ir model es we 2 df elative 240 SRMR ? .10 or SRMR ? .10 and RMSEA ? .06). For losing weight, the model fit indices ere ? 2 df (259) = 855.524, CFI = .960, SRMR = .036, and RMSEA = .041. Finally, for liminating affirmative action, the model fit indices were ? 2 df (154) = 752.310, CFI = .954, SRMR = .072, and RMSEA = .053. Thus, I will proceed to interpreting the model parameter estimates to test hypotheses. Hypothesis Testing Earlier I posited five hypotheses between communicant activeness in problem solving and the four situational antecedent variables and one motivational variable. They are: H11: The higher the problem recognition, the higher the situational motivation in problem solving. H12: The higher the constraint recognition, the lower the situational motivation in problem solving. H13: The higher the involvement recognition, the higher the situational motivation in problem solving. H14: The higher the referent criterion, the higher the communicant activeness in problem solving. H15: The higher the situational motivation in problem solving, the higher the communicant activeness in problem solving. Figures 20-22 summarize the SITCAPS models and their parameter estimates in three social and individual problems. w e 241 242 Table 25 SITCAPS Model with Situational Motivation in Problem Solving: Chi-square Differences and Goodness-of-Fit indices Model ? 2 df ? 2 / df RMSEA SRMR CFI NNFI AIC Independence 11278.805 231 48.826 10816.805 War in Iraq SITCAPS 927.538 154 6.023 .060 (.057, .064) .077 .930 .895 619.538 Independence 13360.036 231 57.836 12898.036 Losing Weight SITCAPS 855.524 259 3.303 .041 (.038 .044) .036 .960 .950 337.524 Independence 15274.422 325 46.998 14624.422 Eliminating Affirmative Action SITCAPS 752.310 154 4.885 .053 (.049, .057) .072 .954 .932 444.310 Figure 20: SITCAPS (War in Iraq). Problem Recognition Constraint Recognition Referent Criterion Level of Involvement Communicant Activeness in Problem Solving (CAPS) . 1 1 5 * - . 140 * .5 59* ** Situational Motivation in Problem Solving War in Iraq (n = 1380) ? 2 df (154) = 927.538, p < .001 CFI = .930 SRMR = .077 RMSEA = .060 . 8 3 1 * * * . 4 9 3 * * * .798***-.527*** .337*** -.378*** Information Forefending Information Processing Information Seeking Information Sharing Information Forwarding Information Permitting .8 3 4 ** *.95 1 ** * .8 9 1 * ** . 7 6 1 * * * .5 6 5 * ** . 8 0 9 * * * * p < .05, ** p < .01, *** p < .001, 243 Figure 21: SITCAPS (Losing Weight). 244 Figure 22: SITCAPS (Eliminating Affirmative Action). . 7 8 4 * * * m em are some obstacles in a situation that limit one?s ability to do anything about the In H11, I expected a positive relationship between problem recognition (i.e., ?detect something is missing and should be done?) and situational motivation in proble solving (i.e., the likelihood of ?stop to think about? the problem). I found positive path coefficients in all three problems: .115 (p < .05) for war in Iraq, .181 (p < .001) for losing weight, and .262 (p < .001) for eliminating affirmative action. Thus, as the probl recognition increases, the more one becomes motivated in problem solving. In H12, I expected a negative relationship between constraint recognition (i.e., ?perceive that there 245 situation?) and situational motivation in problem solving. I found all negative p coefficients across all three problems: -.140 (p < .05) for war in Iraq, -.657 (p < .001) for losing weight, and -.158 (p < .001) for eli ath minating affirmative action. Thus, I conclude that as decreases. I t ve to a t, such as wishful thinking or willful thinking about the or example, a terminally ill patient would have a strong willful thinkin ay bjective referent?for a PC virus infection from a past experience. ill one?s constraint recognition increases, one?s situational motivation in problem solving In H13, I expected a positive relationship between level of involvement (i.e., perceive some ?connection? between a situation and oneself) and situational motivation. found strong positive relationships: .831 (p < .001) for war in Iraq, .420 (p < .001) for losing weight, and .630 (p < .001) for eliminating affirmative action. I thus conclude that as one?s perceived involvement increases, his or her situational motivation toward the problem will increase. H15 asks how one?s problem solving would differ in the presence of a referen criterion. Referent criterion was redefined in this study as any knowledge or subjecti judgmental system that exerts specific influence on the way one approaches problem solving. This can be any decisional guideline or decision rules perceived as relevant given problem: i.e., either an objective referent, such as one carried from prior problem solving, or a subjective referen problem outcomes. F g on his problem (?I will be fine??i.e., subjective referent) and subsequently he thinks and look for information that reinforce his strong ?subjective? belief. Or, one m carry a solution?i.e., o The availability and applicability of such cognitive knowledge (internal referent) w increase communicant activeness regardless whether it is subjective or objective. 246 247 ansmission of one?s knowledge on the problem will increase as one possesses a stronger referent criterion about the problematic situation. From the analysis, I found support on this prediction: .493 (p < .001) for war in Iraq, .149 (p < .01) for losing weight, and .492 (p < .001) for eliminating affirmative action. Finally, I expected that the situational motivation in problem solving will increase communicant activeness in problem solving (H15). In all three problems, I found support for this prediction: .559 (p < .001) for war in Iraq, .887 (p < .01) for losing weight, and .542 (p < .001) for eliminating affirmative action. Thus, I conclude that as situational problematic perception increases, one will experience a heightened situational motivation toward problem solving; and it subsequently increases one?s active communication in information selection, transmission, and acquisition. The presence of a strong referent criterion, regardless of its subjectivity (e.g., a willful thinking toward the outcome), is likely to increase subsequent communicant activeness about the problem. It is notable that the coefficients were rather fluctuating across different issues. I reason the fluctuation as originated from the issue sensitivity in this particular sample (i.e., students). For this homogeneous student group, respondents were similarly sensitive in some issues than others (e.g., war in Iraq). In addition, it is notable that the current data were drawn from a non-random sample, not a possible population. With a more heterogeneous samples (e.g., random samples drawn from a national population), the path coefficients could have been more similar across three problems. Finally, I took out some The CAPS model includes variables in information selection (e.g., information forefending) and information transmission (e.g., information forwarding) as new key dimensions. Thus, I conceptually predicted that a problem solver?s selectivity and tr corr et xo s v recognition and constraint recognition in Figure 20) because specifying those correlations ade a m del convergence impossible. Situational Theory of Cognitive Entrepreneurship in Problem Solving (SITCEPS) o Another new construct I developed is cognitive entrepreneurship in problem lv (CEPS). CEPS alone is an independent concept that features different cognitive proaches across dif nt types of problems. However, CEPS can also be explained by tu ante les. he conceptual relationships with the antecedent riables (see Figure or the summary model, Chapter II: Conceptualization). The m d del b es one of the situational theories of problem solving. Table 26 reports the results of the SITCEPS model tests. For war in Iraq, the od ce ? = 5, CFI = .963, SRMR = .041, and RMSEA .052. Fo iminating affir tion, the model fit indices were ? 2 df (94) = 397.388, FI RM 04 R = .048. Model fit indices suggest that both odels are good enough to interpret. Thus, I proceed to interpreting the model parameter ti s st hy ese yp s sting I posited five hypotheses between cognitive entrepreneurship in problem solving d the four tional antece riables and the situational motivation variable. e s are: 16: The hi the problem recognition, the higher the situational motivation in problem solving. tional motivation in problem solving. elations b ween e genou ariables (e.g., a correlation between problem m M so ap si va co m = C m es H an Th H H17: The higher the constraint recognition, o del ing fere va 12 f ational cedent riab I posited t bine mo ecom el fit indi s were 2 df (67) 319.31 r el 4, S mative ac and = .96 R = . 1, MSEA mate othe to te is Te poth s. situa ese gher dent va se hypoth the lower the situa 248 H18: The higher the involvement recognition, the higher the situational motivation in problem solving. H19: The higher the referent criterion, the lower the cognitive entrepreneurship in H20: The higher the situational motivation in problem solving, the higher the cognitive entrepreneurship in problem solving. Figures 23-24 summarize the structural parameter estimates. problem solving. 249 250 Table 26 SITCEPS Model with Situational Motivation in Problem Solving: Chi-square Differences and Goodness-of-Fit indices d 2 RMSEA SRMR CFI N Mo el ? 2 df ? / df NFI AIC n n e 9 4 8.I depe denc 6 85.4 7 120 5 212 6745.447 W in I aq ar r SITCEPS 3 1 . 4 5 1 . 19.3 5 67 4 766 .052 (.0 7, .0 8) .04 .963 934 185.315 n n e 6 0 7. I depe denc 8 68.1 8 136 6 736 8396.108 Eli ffirm minating A ative Action SITCEPS 3 8 . 4 5 1 . 97.3 8 94 4 228 .048 (.0 3, .0 3) .04 .964 949 209.388 Figure 23: SITCEPS (War in Iraq). Figure 24: SITCEPS (Eliminating Affirmative Action). 251 The hypotheses from H16 to H18 are conceptually identical predictions to those in H11-H13. In general, I found similar support as found in the SITCAPS analysis. In H16, I expected a positive relationship between problem recognition and situational motivation in problem solving. I found partial support: .464 (p < .001) for war in Iraq -.080 (n. s.) for eliminating affirmative action. In H17, I expected a negative relationship between constraint recognition and situational motivation in problem solving. I found all negative path coe and fficients across all three problems: -.136 (p < .01) for war in Iraq and - e action. In H18, I expected a positive relation his ial in problem solving ., s suggests that .535 (p < .001) for eliminating affirmativ ship between level of involvement and situational motivation. I found positive relationships: .581 (p < .001) for war in Iraq and .655 (p < .001) for eliminating affirmative action. H19 inquires about the presence of a referent criterion during a problematic situation. From the analysis, I found support for this prediction: -.489 (p < .001) for war in Iraq and -.589 (p < .001) for eliminating affirmative action. T suggests that the presence of referent criteria tend to decrease one?s entrepreneurial mindset?i.e., with applicable referents one becomes less entrepreneur . With the deployable referent for a given problem, one will be more likely to jump into a conclusion (i.e., a solution carried from prior situations) and to turn to information that optimizes the chosen solution?i.e., backward reasoning. Finally, in H20, I expected that the situational motivation in problem solving will increase cognitive entrepreneurship in problem solving. However, from the analysis, I found no support for this prediction: -.487 (p < .001) for war in Iraq and -.422 (p < .001) for eliminating affirmative action. Notably, the signs of the paths were both opposite (i.e negative) from situational motivation to cognitive entrepreneurship. Thi 252 253 the more one is situationally m test to examine the significance of the otivation and cognitive retrogression. The Sobel test allows testing of whether the indirect an, West, & Sheets, 2002; Sobel, 1982). In both issues, the mediation effects between situatio gnitive retrogression were significant: .454 (p < .001) for war in 001) for eliminating affirmative action. For cognitive multilateralism, the indirect paths were not significant for both issues: -.007 (n. s.) for war in Iraq and .014 (n. s.) for eliminating affirmative action. For cognitive commitment, I found a significant mediation effect: as the situational motivation heightens, cognitive commitment increases: .111 (p < .001) for war in Iraq and .075 (p < .01) for eliminating affirmative action. Finally, for cognitive suspension, I found that as situational motivation increases, cognitive suspension decreases: -.096 (p < .01) for war in Iraq and -.078 (p < .01) for eliminating affirmative action. The tests show that the CEPS construct in the SITCEPS model is primarily characterized by the cognitive retrogression dimension (i.e., the more entrepreneurial, the less retrogression). The standardized path coefficients were -.933 for war in Iraq and - .972 for eliminating affirmative action. Cognitive retrogression refers to a cognitive otivated for problem solving, the less one becomes entrepreneurial in problem solving. Careful examination of the subdimensions in CEPS shows that the more one is situationally motivated, the more he or she engages in backward reasoning (i.e., cognitive retrogression). Cognitive retrogression was the strongest dimension among four subdimensions of CEPS. I conducted Sobel?s effect between situational m effect of an independent variable on a dependent variable via the mediator is significantly different from zero (MacKinnon, Lockwood, Hoffm nal motivation and co Iraq and .410 (p < . tendency for backward reasoning (i.e., ?a conclusion dictates evidence that secures and warrants the chosen conclusion.?). In other words, as situational motivation in problem solving grows, one tends to move backward in thinking about the problem. Thus, the result indicates that situational motivation does not create cognitive entrepreneurship in problem ore one is motivated toward problem solving, the more one is likely to become non-entrepreneurial. From the finding, I conclude that situ al motivation tends to pressure problem solvers to hastily turn to a conclusion and optimize the hastily drawn conclusion. umm simi t for H16-H18 for situational perceptual riables as in the SIT PS analysis. ddition, for H19, I found the more one s s feren terio e le e becomes entrepreneurial in problem solving. r H20, I failed to find support. The finding suggests the opposite relationship between tu ti a itive entrepreneurship in problem solving: i.e., as one comes m vated in prob solving, one tends to become non-entrepreneurial. It g ro ed ion, problem solvers tend to mobilize available gnitive resources backwardly?to optimize the selected solution to be more conclusive d vi g. la s etw AP d CE The previous two structural models were designed to test the nomological validity the two ne onstructs. tion, they tested the more generalized version of the tua of pro . Each model was derived from the situational eory of pro solving and stands by itself as a conceptual model with a unique CAPS is designed to understand and explain different solving. On the contrary, the m ation In s ary, I fo CA und lar suppor In ava po Fo si be su co an Re of si th theoretical purpose. SIT sesse a re t cri n, th ss on ational mo vation nd cogn oti hat f lem htengests t m the heig motivat con tion ncin hip b een C S an PS w c eory blem In addi m solvtional th ble ing 254 communicant behavior during problematic situations. SITCEPS is intended to describe and understand differential cognitive approaches in problem solving. Because of model identification issues, I constructed two nonrecursive models that examine the simultaneous causal influences between CEPS and CAPS (see Figure 13 and 14 rmative action, the model fit indices g H21: The higher the cognitive entrepreneurship, the higher the communicant activeness in problem solving. H22: The higher the communicant activeness, the lower the cognitive entrepreneurship in problem solving. for the summary model, Chapter II: Conceptualization). As discussed, nonrecursive models are often difficult to solve mathematically. I tried both of the model specifications but found that only the Figure 16 model (i.e., referent criterion?CEPS) converged successfully in both issues of war in Iraq and eliminating affirmative action. The Figure 17 model (i.e., referent criterion?CAPS) failed to converge in both issues. Table 25 reports the result of the converged nonrecursive model tests. For war in Iraq, the model fit indices were ? 2 df (217) = 794.508, CFI = .961, SRMR = .045, and RMSEA = .044. For eliminating affi were ? 2 df (153) = 932.838, CFI = .931, SRMR = .066, and RMSEA = .061. According to Hu and Bentler?s joint criterion, these fit indices suggest that both models are good enough to interpret. Thus, I interpret the model parameter estimates to test hypotheses. Hypothesis Testing I posited two hypotheses between cognitive entrepreneurship in problem solvin and communicant activeness in problem solving. These hypotheses are: 255 256 Table 27 Nonrecursive Model: Chi-square Differences and Goodness-of-Fit indices Model ? 2 df ? 2 / df RMSEA SRMR CFI NNFI AIC Independence 14918.052 276 54.051 14366.052 War in Iraq Nonrecursive 794.508 217 3.661 .044 (.041, .047) .045 .961 .950 360.508 Independence 11563.657 190 60.861 11183.657 Eliminating Affirmative Action Nonrecursive 932.838 153 6.097 .061 (.057, .065) .066 .931 .915 626.838 CAPS, featuring communicative behavior, and CEPS, featuring the cognitive approach in roblem solving, seem to explain each other to some degree. However, it is hard to define one as the antecedent condition to the other. Indeed, it is most plausible to conclude that CAPS and CEPS affect each other simultaneously (i.e., bidirectional causality). Figures 25-26 summarize the parameter estimates for the converged models. For H21, I expected a positive causal influence from cognitive entrepreneurship in problem solving to communicant activeness in problem solving. However, I found no support for this prediction in both issues: -.959 (n. s.) for war in Iraq and -.892 (n. s.) for eliminating affirmative action. Notably, the signs were opposite, unlike what I expected before model testing; and the standard errors of the parameter estimates were relatively large (i.e., 1.150 of unstandardized parameter estimate with 2.401 of S. E. in war in Iraq and .459 of unstandardized parameter estimate with .633 of S. E.). For H22, I expected a negative causal influence from communicant activeness in problem solving to cognitive entrepreneurship in problem solving. I found support for this expectation in both issues: -.919 (p < .001) for war in Iraq and -.622 (p < .001) for eliminating affirmative action. This suggests that the more active communicative behaviors in information forefending, forwarding, and seeking, the less one will become entrepreneurial in problem solving. The accumulated knowledge and experience from roblem-solving efforts in dealing with information seems to reduce entrepreneurial mindset regarding the problem. p p 257 Figure 25: Nonrecursive model between CEPS and CAPS (War in Iraq). Referent Cognitive Multilateralism Cognitive Retrogression Cognitive Commitment Cognitive Suspension Communicant Activeness in Problem Solving Information Forefending Information Processing Information Seeking Inf Shar ormation ing Information Forwarding Information Permitting . 9 7 5 * * * . 8 7 4 * * * . 7 0 0 * * * . 7 0 2 * * * . 6 * * * 6 0 . 7 * * * 8 3 - . 9 4 * 6 * * -. 0 0 7 ( n .s .) -. 2 0 9 * * * . 1 8 9 * * * Cognitive Approach Communicative Behavior Cognitive Entrepreneurship in Problem Solving -.959 (n.s.) -.919*** War in Iraq (n = 1380) ? 2 df (217) = 794.508, p < .001 SRMR = .045 RMSEA = .044 CFI = .961 * p < .05, ** p < .01, *** p < .001, Criterion - . 0 9 8 ( n . s . ) 258 Figure Action). 26: Nonrecursive model between CEPS and CAPS (Eliminating Affirmative Cognitive Multilateralism Cognitive Retrogression Cognitive Commitment Cogn Suspension itive Communicant Activeness Solving in Problem Information Forefending Information Processing Information Seeking Information Sharing Information Forwarding Information Permitting . 9 8 * 8 * * . 7 9 * 0 * * . 7 4 8 * * * . 6 6 9 * * * . 5 6 3 * * * . 6 8 0 * * * - . 8 3 5 * * * -. 0 2 2 ( n .s .) -. * 1 5 7 * * . 1 3 4 * * * Cognitive Approach Communicative Behavior Cognitive Entrepreneurship Solving in Problem -.892 (n.s.) Eliminating Affirmative Action (n =1380) ? 2 (153) = 932.838 , p < .001 df CFI = .931 SRMR = .066 RMSEA = .061 * p < .05, ** p < .01, *** p < Referent Criterion - . 4 ( n . s . ) 1 7 -.622*** .001, I pay special attention to the H21. I expected that the higher cognitive solver could also be very active in communication behavior. In conceptualizing cognitive entrepreneurship in problem solving, I explicitly said that a non-entrepreneurial mindset entrepreneurship in problem solving would lead to higher communicant activeness. However, in retrospect, I failed to consider that the low cognitive entrepreneurial problem 259 would also be very active in problem solving but in a different way from the entrepreneurial. In other words, high cognitive non-entrepreneurship does not mean passive lly itive en, f as, repreneurial. Therefo g affirmative action. ness in communication behavior. As we found routinely, many non- entrepreneurial problem solvers tend to vigorously seek and forward information that is consistent with a hastily drawn conclusion. Thus, it was not conceptually and empirica valid to specify a positive causal relationship between high cognitive entrepreneurship and communicant activeness in problem solving. The nonsignificant path from cogn entrepreneurship to communicant activeness seems to reflect such a notion. In contrast, the reverse causal flow seems to be logical conceptually. As a problem solver experienced heightened communicant activeness, she or he is likely to develop a good deal of knowledge and preference on how to solve the problem. Th such preference (i.e., high information forefending) would increase cognitive retrogression (i.e., optimizing a preferred conclusion backwardly) and less tolerating o incompatible information, less committed to all the available candidate solution and ide and less suspending. As a result, the problem solver becomes more non-ent re, the negative path from CAPS to CEPS seems to reflect such a negative conceptual relationship. A careful examination of the two nonrecursive model reveals the domination of cognitive retrogression and the information forefending in both issues. Information forefending is the strongest dimension in CAPS and cognitive retrogression is strongest dimension in CEPS. The relationship between cognitive retrogression and information forefending is positive: .885 (p < .001) for war in Iraq and .734 (p < .001) for eliminatin 260 In summary, from the H21 and H22 tests, I found that communicant activenes problem solving tends to reduce the problem solver?s cognitiv s in e entrepreneurship treated with caution. dditional Analyses In the test of a nonrecursive relationship between CAPS and CEPS, I found that formation selectivity and cognitive retrogression dominated the other dimensions. In ddition, I found a nonsignificant path from CEPS to CAPS. As I discussed, the H21 rediction was a conceptual mistake in that both high and low entrepreneurial problem solvers could be high in communicant activeness. In other words, there is no conceptual reason that less entrepreneurial problem solvers should be low in communicant ctiveness. To test this reasoning, I conducted additional analysis to inquire how different they are in terms of communicant activeness. In conceptualizing CEPS, I proposed a model of cognitive alpha and omega groups (CAOS) in terms of their reasoning direction: i.e., ?evidence?conclusion? for forward reasoning and ?conclusion?evidence? for backward reasoning. I named those forward reasoning problem solvers as the cognitive alpha group, and those backward reasoning problem solvers as the cognitive omega group (see Figure 9, Chapter II: Conceptualization). Thus, I computed the average scores of cognitive retrogression (i.e., backward reasoning) for all the respondents and regrouped them high, medium, and low scorers in cognitive retrogression. For analysis, I selected the high and low groups and named them eventually, while the opposite cannot be assumed. However, this interpretation should be A in a p a 261 262 as cognitive omega (i.e., high in cognitive retrogression) and cognitive alpha (i.e., low in cognitive retrogression). I expected that similar model structures of CAPS in both the cognitive alpha group and cognitive omega groups. However, I predicted that they will differ in terms of information selectivity. Specifically, I expected that the cognitive alpha group would have a low path coefficient in information forefending and a high one in information permitting. In contrast, I predicted that the cognitive omega group would show the opposite pattern?i.e., high path coefficient in information forefending and low path coefficient in information permitting. If supported, this pattern of structural paths in CAPS would provide some evidence for the explanation of nonsignificant path in H21 (i.e., the low entrepreneurial problem solvers can also be active in communication behavior, just as those high entrepreneurial problem solvers are active in communication behavior.). The higher or lower entrepreneurial problem solvers do not differ in their information transmission and acquisition but only in their information selectivity (i.e., non-entrepreneurial problem solvers tend to more forefend information.). Table 28 reports the CAOS and CAPS comparison models. For war in Iraq, the CAPS model fit of cognitive alpha group were ? 2 df (139) = 314.960, CFI = .946, SRMR = .061, and RMSEA = .053 and for the cognitive omega group were ? 2 df (145) = 265.130, CFI = .969, SRMR = .047, and RMSEA = .043. For eliminating affirmative action, the CAPS model fit of cognitive alpha group were ? 2 df (97) = 230.285, CFI = .936, SRMR = .054, and RMSEA = .055 and for the cognitive omega group were ? 2 df (96) = 175.514, CFI = .961, SRMR = .039, and RMSEA = .043. Thus, I interpret the model parameter es su Fi timates to ex e w r th ttern re si r as ecte igur -28 mmariz e structura ths of both gr in two issue gure 27: CAPS comparisons between cognitive alpha and omega groups (War in Iraq). amin hethe e pa s we mila I exp d. F es 27 e th l pa oups s. Communicant Activen in Prob Solvi ess lem ng Information Forfending Information Permitting Information ardingForw Information Seeking Information Sharing Information cessingPro . 9 4 4 * * * .4 3 3 ** * . 9 2 9 * * *. 4 8 8 * * * . 5 3 8 * * * . 8 7 7 * * * Cogni lpha Gro War in Iraq (n = 456) ? 2 df (139) 4.960, p < .001 SRMR = .061 RMSEA = .053 tive A up = 31 CFI = .946 * p < .05, ** p < .01, *** p < .001, Communicant ess lem Solving Cognitive O Gr War in Iraq 451 ? 2 df (145 p CFI = .969 SRMR = .047 RMSEA = mega (n = .130, oup ) ) = 265 < .001 .043 Activen in Prob Information Forfending Information Permitting Information ardingForw Information Seeking Information Sharing Information ingProcess . 9 0 8 * * * .4 8 5 * ** . 8 6 7 * * * .6 5 6 * ** . 4 9 6 * * * . 2 * * * 8 7 * p ** p < .001, < .05, < .01, *** p 263 Figure 28: CAPS comparisons between cognitive alpha and omega groups (Eliminating Affirmative Action). . 9 6 * * 3 * . 8 7 7 * * * . 5 2 4 * * * . 9 * * * 7 9 . 8 6 3 * * * . 9 0 5 * * * . 2 4 6 * * * . 9 8 0 * * * 264 265 Table 28 Additional Analysis of CAPS (Cognitive Alpha versus Cognitive Omega Groups): Chi-square Differences and Goodness-of-Fit Indices Model ? 2 df ? 2 / df RMSEA SRM NN IC R CFI FI A end nce Indep e 3471.731 190 18.272 1.731 309 gnit e ha G up = 45 ) gnit e meg rou = 45 ) Co iv Alp ro (n 6 314.960 139 2.266 .053 (.045, .060) .06 .9 .960 Indep e 4104.376 190 21.602 4.376 War in Iraq Co iv O a G p (n 1 CAPS 265.130 145 1.828 .043 (.035, .051) .04 .960 .130 Independe 2232.467 136 16.415 0.467 CAPS end nce nce 1 .946 7 .969 27 36 372 -24 196 gnit e ha G up ) e ) Co iv Alp ro (n = 451 CAPS 230.285 97 2.374 .055 (.057, .065) .054 .911 .285 Independe 2155.632 136 15.850 3.632 Eliminating Affirmative Action Cognitiv Omega Group (n = 458 CAPS 175.514 96 1.828 .043 (.032, .052) .039 .944 .486 .936 .961 36 188 -16 nce A careful examination indicates that information selectivity was different between the cognitive alpha group and cognitive omega group in both issues. Specifically, the cognitive alpha groups in the two issues were lower in information forefending (i.e., .488 for war in Iraq and .449 for eliminating affirmative action) and relatively higher in information permitting (i.e., .538 in war in Iraq and .524 in eliminating affirmative action). The pattern reversed in the cognitive omega group. Cognitive omega groups were higher in information forefending (i.e., .656 for war in Iraq and .776 for eliminating affirmative action) and relatively lower in information permitting (i.e., .496 for war in raq and .246 for eliminating affirmative action). I tabulated these findings in the Tables 29-32. Notably, the information selectivity shows contrasting reflections between the cognitive alpha and omega groups. For the information transmission and acquisition dimensions, I expected a similar pattern between two groups. I visualized these dimensions in Tables 29-32. Even with eyeball examination, two groups possess similar structural path coefficients in information transmission and information acquisition. Thus, I conclude that the additional analysis of CAOS and CAPS seems to support why H21 failed to get support. In other words, the low entrepreneurial problem solvers can also be active in communication behavior, just as those high entrepreneurial problem solvers are active in communication behavior. Specifically, the low and high entrepreneurial problem solvers seem to differ in information selectivity. However, those higher or lower entrepreneurial problem solvers do not differ in their information transmission and acquisition (i.e., those non-entrepreneurial problem solvers tend to more forefend information, while they are similarly seeking and forwarding information as those entrepreneurial problem solvers do.). I 266 Figure 29: Standardized structural path coefficients between cognitive alpha and cognitive omega group (War in Iraq). CAOS and Information Selectivity (War in Iraq) Cognitive Alpha Cognitive Alpha Cognitive Omega Cognitive Omega 0.1 0. 0.4 0.8 S t an d a C en t 0.5 0.6 0.7 0.9 1 r d i z ed o ef f i ci 0 0.2 3 Information Forefending Information Permitting Cognitive Alpha Cognitive Omega mative Action). Figure 30: Standardized structural path coefficients between cognitive alpha and cognitive omega group (Eliminating Affir CAOS and Information Selectivity (Eliminating Affirmative Action) 0.9 1 Cognitive Omeg Cognitive Alpha Cognitive Alpha a Cognitive Omega 0.1 3 0 0.2 0. 0.5 0. 0.8 Information Forefending Information Permitting St a n ed C ent 0.4 0.6 7 dar d i z o ef f i ci Cognitive Alpha Cognitive Omega 267 Figure 31: Standardized structural path coefficients between cognitive alpha and cognitive omega group (War in Iraq). CAOS and Information Transmission and Acquisition 0.5 0.6 0.7 0.8 0.9 1 d i z ed C ffi c i e n t (War in Iraq) 0 0.1 0.2 0.3 0.4 St anda r o e Cognitive Alpha Cognitive Information Forwarding Information Sharing Information Seeking Information Processing Omega Figure 32: Standardized structural path coefficients between cognitive alpha and cognitive omega group (Eliminating Affirmative Action). CAOS and Information Transmission and Acquisition (Eliminating Affirmative Action) 0 0.2 0.3 0.4 0.5 0.6 0.7 0.8 9 Information t andar d ed C oef f ent 0.1 0. 1 Forwarding Information Sharing Information Seeking Information Processing S i z i ci Cognitive Alpha Cognitive Omega In the next chapter, I will discuss the findings and their implications to theory and practice. Following that, I will discuss the limitations of current study and the suggestions for future research. 268 CHAPTER V: CONCLUSIONS AND IMPLICATIONS All life is problem solving. Karl Popper The primary purpose of this study was to develop two new concepts, commu ip in the and ecame a em omes acquisitive about information pertain o ior to ) rather than to spend cognitive efforts after fi r the nicant activeness in problem solving (CAPS) and cognitive entrepreneursh problem solving (CEPS). I then added CAPS and CEPS as the dependent variables to situational theory of publics (J. Grunig, 1968, 1989, 1997). By this way, I elaborated refined the existing situational theory of publics (STP) further. As a result, STP b more general theory of human problem solving, and I called the resulting theory the situational theory of problem solving (STOPS). In CAPS, I started with the guiding premise: The more one commits to probl resolution, the more one becomes selective in dealing with information, the more one becomes transmissive, and the more one bec ing to the problem. In CEPS, I postulated: A problem solver with heightened cognitive entrepreneurship tends to 1) generate a large number of mental syllogistic models before he or she finally selects one for problem solution; 2) commit more t proposed solution proposals, as if they are a solution, during evaluation; 3) be more heedful in finalizing a conclusion; and 4) be more likely to invest cognitive labor pr finalizing a conclusion (i.e., an evaluation purpose nalizing a conclusion (i.e., justification purpose). Parallel to the conceptual explication, I developed measurement systems fo new concepts and tested their validity and utility with the situational antecedent variables such as problem recognition, constraint recognition, level of involvement, referent 269 criterion, and situational motivation in problem solving. In doing so, I refined the concepts of problem recognition and referent criterion to address some issues raised against the situational theory of publics. After the introduction of the key variables in the new situational theory of problem solving, I constructed a series of confirmatory factor analytic models and full causal structural equation models. ns. valid Overall, I posited 22 hypotheses among the key variables and their subdimensio I collected data using the survey method and analyzed them using structural equation modeling with the EQS 6.1 program. In general, I found a good amount of support for most hypotheses. Those few not supported revealed conceptual mistakes I made in deriving predictions (e.g., situational motivation?CEPS in the SITCEPS model). Also, some serendipitous findings provide a valuable chance for conceptual refinement. This final chapter consists of the detailed summary of what I found from the model and hypothesis testing, discussions regarding supported and non-supported hypotheses, the implications of CAPS, CEPS, and STOPS to theory and practice, and finally a summary of limitations and ideas for future research. In the following section, I will first recapitulate the results of this study with discussions of major findings. Summary and Discussion Reliability and Validity of the Measurement Instruments Of two new concepts, CAPS and CEPS, it is necessary to have reliable and measurement systems to be useful. I conducted a series of exploratory tests such as Cronbach?s alpha and Principal Component Analysis. In general, analysis of three 270 problems (war in Iraq, losing weight, and eliminating affirmative action) showed that CAPS and CEPS measurement systems are reliable and internally consistent. For nomological validity test purposes, I introduced five situational antecedent variables: problem recognition, constraint recognition, level of involvement, referent criterio erformed an acceptable job bility and validity. Tests of the New Constructs and Hypotheses After the reliability and validity tests, I proceeded to confirmatory factor analysis and causal model analysis to examine construct validity of the two new constructs. I summarize the findings of the SEM analysis in following. CAPS Using the sub-sampling strategy, I divided total sample of n = 1,380 into developmental (n = 467) and validation samples (n = 458, n = 460). In all the nested model tests, the six factor oblique models were always better than one-factor model structure for CAPS. This suggests that CAPS has multidimensionality, as it was conceptualized. Of six out of nine nested model tests in three issues, the second-order factor structure turned out to be a better model than the six factor oblique model. In general, CAPS seems to be better represented as a second-order construct. Hence, I conducted three second-order confirmatory factor analyses using the total sample (n = 1,380) to test hypotheses regarding CAPS. n, and situational motivation in problem solving. I examined the internal consistency and reliability of the situational antecedent variables and found, overall, the measurement items for the situational antecedent variables p in relia 271 The evaluation of the model fit to data showed that in all three models the second order structure is a reasonable conceptual representation for the CAPS data. Thus, examined the parameter estimates in the second-order models to test H1-H6. All hypotheses found support as expected at (p < .001). The loadings in the observed variables were all significant at the p < .001 level. Most parameter estimates in - I the measur mes active ). n a desirable characteristic in problem solving (e.g., in the issues o tion s isition in ement items vary between .313 and .826. The core premise in the CAPS model was that as a communicant becomes active in problem solving, one will become more selective, transmissive, and acquisitive regarding information about the problem. Specifically, as a communicant beco in problem solving, one?s information selectivity tends to increase (i.e., H1 and H2 Notably, the subject?s information forefending tendency increased more than the information permitting tendency as communicant activeness heightens: for informatio forefending .691, .613, .748 vs. for information permitting .530, .612, .538 in three problems. Thus, as problem solvers become more active, they tend to lose information permissiveness, which is often f social conflicts). In all three problems, information transmission (i.e., H3 and H4) was the most salient dimension in communicant activeness in problem solving: .921, .903, .984 for information forwarding; .946, .973, .968 for information sharing. This indicates that as problem solvers become more active, their dominant characteristic is to give informa about a problem to others. As discussed earlier, the previous situational theory of public and other communication models often focused exclusively on information acqu conceptualizing communicator activeness. Such findings tell us how the omission of 272 information giving in previous research was a loss in understanding communicator?s activeness in dealing with his or her life problems. To name just a few, with the inclusion of information selectivity and information transmission variables, now we can explain better how certain ideas are dispersed among people (diffusion of innovations), how and why ce t es of s more ing I divided the total sample of n = 1,380 into developmental (n = 467) and validation samples (n = 917). For CEPS, I proceeded directly to the second-order confirmatory factor analysis in the development and validation samples. I considered the second-order confirmatory factor analyses as a good model structure if the model tests resulted in the acceptable fit indices according to Hu and Bentler?s (1999) joint criterion. rtain social problems are enduring and become chronic (conflict resolution), and how activist and active publics can be conceptually distinguished (public relations). The last dimension of information acquisition consists of information seeking and information processing (i.e., H5 and H6). These two variables are the defaul dependent variables highlighting publics? active and passive communication behaviors in the situational theory of publics (J. Grunig, 1997). Consistent with what past studi the STP found, information seeking and processing were increased as one become active in problem solving: .893, .899, .794 for information seeking; .474, .819, .912 for information processing. Problem solvers with heightened motivation in problem solv tend to make efforts in gaining information about the problem. In summary, I found support for the hypotheses regarding six conceptual subdimensions of CAPS. Each subdimension has a positive relationship with its higher dimension of communicant activeness in problem solving. CEPS 273 Results n of 001 1). In CEPS model, I postulated that as one becomes active in problem solving, one wil i.e., a er. For H7, I found evidence in both issues. Notably, the magnitude of the path coefficient was higher for the war in Iraq issue (i.e., -.754, p < .001) than the eliminating affirmative action issue (i.e., -.258, p < .001). This would suggest that there is a stronger backward showed that the second-order structure was a good conceptual representatio the data. I tested the two second-order models and the standardized structural path estimates and standardized loadings in two problems. In both problems (n = 1,380), the structural paths from the second-order latent variable (i.e., cognitive entrepreneurship) to first-order latent variables (e.g., cognitive retrogression) were significant at the p < . level, except the path to cognitive commitment in the war in Iraq issue (i.e., p < .0 addition, the loadings in the observed variables were all significant at the p < .001 level. Most parameter estimates in the measurement items varied between .400 and .700 (the lowest .285, the highest .909). In the l use less backward reasoning (cognitive retrogression), will possess more cognitive multilateralism, will exhibit more commitment, and will exhibit more suspending in finalizing a solution. The analysis of CEPS in two social problems provides supportive evidence for the main postulate and the four hypotheses. In H7, I conceptually predicted that one?s cognitive retrogression in problem solving will decrease as cognitive entrepreneurship in problem solving increases: negative relationship. The most salient feature in cognitive retrogression is one?s backward reasoning?i.e., ?a conclusion directs certain evidence.? This is an optimization process for an a priori conclusion one drew earli supporting 274 reasoni r the and H9, I adopted a single indicator approach in measurement. In H8, I predict e and the extent of patroni e war ar in s seemed to feel tired of the issue, because it had been prolong n osition, it will result in a low score in cognitive commitment. In contrast, the eliminating affirmative action issue was relatively a less ?entrenched? ng tendency (i.e., cognitive omega approach) for the war in Iraq issue than fo eliminating affirmative issue. In H8 ed that as cognitive entrepreneurship increases one will possess more cognitiv breadth and tolerance?the number of decision alternatives and the extent of tolerance in dealing with competing rival information during the problem-solving process. I found support for this prediction in both issues: .184 in war in Iraq and .334 in eliminating affirmative action. H9 predicted that as cognitive entrepreneurship heightens, one will make more cognitive commitment?the degree of enthusiasm zing the proposed solutions for a given problem solving. I found partial support for H9 in the eliminating affirmative action issue (i.e., .166, p < .001), but not in th in Iraq issue (i.e., -.113, p < .01). The signs were opposite in the war in Iraq issue, which suggests that for the war in Iraq issue, as one becomes more entrepreneurial in problem solving, one would be less enthusiastic and patronizing to the proposed solutions. About the reversal of sign in cognitive commitment, I speculate that for the w Iraq issue, survey respondent ed about two years at the time the data were collected (spring 2005). Thus, eve respondents who were high in entrepreneurship in problem solving would stick exclusively to a certain perspective. Hence, the cognitive commitment was less because survey respondents became entrenched in certain positions even if their cognitive entrepreneurship increased. Considering the computation formula, if one adheres exclusively to a single p 275 problem o ds, e ) of usiasm and the extent of patronizing the propose ral uct , so the participants with high entrepreneurship in problem solving seemed t think and value different perspectives more than in the war in Iraq issue. In other wor people result in more ?hedging? (i.e., commit to different ideas at the same time), and th resulting commitment scores become higher as one?s cognitive entrepreneurship heightens. H10 predicted as one?s cognitive entrepreneurship increases, one?s cognitive suspension increases as well. Cognitive suspension refers to a problem solver?s heightened willingness to invest cognitive resources in evaluating and reevaluating a selected solution before finalizing it. In both issues, H10 was supported: .414 (p < .001 for the war in Iraq issue and .770 (p < .001) for the eliminating affirmative action issue. Overall results in the CEPS CFA model tests indicated that as the problem solver?s entrepreneurial mindset in problem solving heightens, one tends to adopt more a backward reasoning strategy, have cognitive breadth, exhibit more tolerance of competing ideas and opinions, and suspend judgment before finalizing a solution. Yet, cognitive commitment?the degree of enth d solutions?seems to differ across issues. This requires further study with different types of issues. Considering the single-item approach with formulas for cognitive multilateralism and cognitive commitment, such a finding begs further analysis. A new study using a multiple-indicator approach as in the cognitive retrogression and suspension will allow a clearer understanding for the two dimensions measured with a single item. In sum, I proposed two new variables, CAPS and CEPS, to make a more gene version of the situational theory of publics (J. Grunig, 1997, 2003). To examine constr 276 validity (i.e., ?nomological network,? Cronbach & Meehl, 1955), I created several causal nets using situational antecedent variables. Earlier in the conceptualization, I posited conceptual relationships between the new constructs and existing antecedent variables in the situational theory of publics. Using structural equation modeling, I tested the presumed conceptual relationships between new and existing variables. In the following section, I summarize a series of models of nomological networks between the two new constru t m h cognition onstraint recognition (i.e., ?percei ll 7 cts and the antecedent variables of situational theory. SITCAPS From the SITCAPS model testing, I found that the three SITCAPS showed acceptable fit to the data. I thus, examined the model parameter estimates to tes hypotheses. In H11, I expected a positive relationship between problem recognition (i.e., ?detect something is missing and should be done?) and situational motivation in proble solving (i.e., the likelihood of ?stop to think about? the problem). I found positive pat coefficients in all three problems: .115 (p < .05) for war in Iraq, .181 (p < .001) for losing weight, and .262 (p < .001) for eliminating affirmative action. As problem re increases, one becomes more motivated in problem solving. In H12, I expected a negative relationship between c ve that there are some obstacles in a situation that limit one?s ability to do anything about the situation?) and situational motivation in problem solving. I found a negative path coefficients across all three problems: -.140 (p < .05) for war in Iraq, -.65 (p < .001) for losing weight, and -.158 (p < .001) for eliminating affirmative action. 277 Therefo I the that erceived as relevant to a given problem: i.e., either an obje rease hether it is subjective or objective. The CAPS udes variables in information selection (e.g., information forefending) and informa lysis, I found good support for this prediction: .493 (p < .001) for war in Iraq, .149 (p < .01) for losing weight, and .492 (p < .001) for eliminating affirmative action. re, as one?s constraint recognition increases, one?s situational motivation in problem solving decreases. In H13, I predicted a positive relationship between level of involvement (i.e., perceive some ?connection? between a situation and oneself) and situation motivation. found strong positive relationships: .831 (p < .001) for war in Iraq, .420 (p < .001) for losing weight, and .630 (p < .001) for eliminating affirmative action. I thus conclude that as one?s perceived involvement increases, his or her situational motivation toward problem will increase. H14 investigated the role of a referent criterion in a problematic situation. I redefined referent criterion as ?any knowledge or subjective judgmental system exerts specific influence on the way one approaches problem solving.? This can be any decisional guideline or decision rules p ctive referent, such as one carried from prior problem solving, or a subjective referent, such as wishful thinking or willful thinking about the problem outcomes. The availability and applicability of such cognitive knowledge (internal referent) will inc communicant activeness regardless of w model incl tion transmission (e.g., information forwarding) as new key dimensions. Thus, I conceptually predicted that the problem solver?s selectivity and transmission of one?s knowledge on the problem will increase as one possesses a stronger referent criterion about the problematic situation. From the ana 278 Finally, by H15, I predicted that the situational motivation in problem solvin increase communicant activeness in problem solving. In all three problems, I found support for this prediction: .559 (p < .001) for war in Iraq, .887 g will (p < .01) for losing weight, tic n, transmission, and acquisition. The presence of strong referent criteria s e previous s?i.e., information seeking and information processing (J. Grunig, 1997). Thus, the refe ss, ined. is directio ble, and .542 (p < .001) for eliminating affirmative action. As situational problema perception increases one will experience a heightened situational motivation toward problem solving, which subsequently increases one?s active communication in information selectio regardless of their subjectivity (e.g., a willful thinking toward the outcome) i likely to increase subsequent communicant activeness about the problem. In th situational theory, the referent criterion had explained little variance in communication behavior rent criterion had eventually been dropped from the independent variables. J. Grunig (1997) discussed the referent criterion as an effect or outcome of communication behavior such as information seeking. I support his reasoning as logical and strategic in theory building. Nonethele the referent criterion still could be useful as an antecedent variable in a way I redef For many problem-solving contexts, we observe that our preconception, subjective beliefs, or carried-over knowledge from past situations affects our subsequent communication behaviors. Specifically, our subsequent information seeking nally tuned by what a referent criterion prescribes and proscribes. We tend to see what the referent criterion implies. Hence, our information seeking and forwarding or sharing are selective. If so, a new situational theory should reintroduce the varia referent criterion. In other words, I theorize that a referent criterion is not only an effect 279 and outcome of communication behaviors, but also an antecedent condition contextualizing subsequent communication behaviors. At times, a referent criterion remains intact (i.e., reinforced); at other times it is revised. g. The ing t item top to think about the issue?? were considered as measures for situational motiva es In addition, communication behavior has become more general. Information seeking would be reduced in the presence of a referent criterion. However, a referent criterion seems to trigger and drive one to be more selective and to be willing to share and forward information to others. I found supportive evidence for this reasonin higher the referent criterion, the higher the communicant activeness in problem solving. Finally, I revised problem recognition to be more of ?detect? something miss rather than ?stopping to think.? I followed Kim, Downie, and De Stefano?s (2005) conceptual explication to distinguish situational motivation (i.e., ?stop to think about? tendency) from the joint function of ?detect,? ?perceived connection,? and ?perceived obstacle? in doing something about the problem. Hence, I used a new measuremen for problem recognition. The previous problem recognition measures, such as ?how often do you s tion in problem solving. Analyses with these refined measures were consistent with the previous situational theory?s prediction (e.g., the more one exhibits problem recognition, the more one will do information seeking.). Besides, the reported multicollinearity issue (e.g., standardized beta coefficients greater than 1.00) between the independent variables was not found. In summary, the causal networks I posited between newly refined situational antecedent variables and CAPS have gained a good amount of support. The finding giv 280 us some confidence in the construct validity for the new construct of CAPS and the situational theory of problem solving. SITCEPS From the SITCEPS model tests, I found acceptable model fits to the data, t proceeded to interpreting the parameter estimates to test hypotheses. For hypotheses from H16 to H18, which are conceptually identical predictions in H11-H13, I fou hus I nd similar support s in ionship between level of involvement and situatio and eliminating affirmative action. This suggests that the presence of a referen as found in the SITCAPS analysis. In H16, I expected a positive relationship between problem recognition and situational motivation in problem solving. But, I found partial support: .464 (p < .001) for war in Iraq and -.080 (n. s.) for eliminating affirmative action. In H17, I expected a negative relationship between constraint recognition and situational motivation in problem solving. I found significant negative path coefficient both problems: -.136 (p < .01) for war in Iraq and -.535 (p < .001) for eliminating affirmative action. In H18, I expected a positive relat n motivation. Again, I found significant positive relationships in both problems: .581 (p < .001) for war in Iraq and .655 (p < .001) for eliminating affirmative action. H19 inquired about the presence of a referent criterion in a problematic situation. From the analysis, I found support for this prediction: -.489 (p < .001) for war in Iraq -.589 (p < .001) for t criterion tends to decrease one?s entrepreneurial mindset. With applicable decisional referents, one becomes more non-entrepreneurial in problem solving. I 281 predicted so because, given a deployable referent for a problem, one will be more likely to jump into a conclusion (i.e., a solution carried from prior situations) and turn to information that optimizes the chosen solution?i.e., backward reasoning. Lastly, H20 predicted that situational motivation in problem solving will increa cognitive entrepreneurship in problem solving. However, I found no support for this prediction: -.487 (p < .001) for war in Iraq and -.422 (p < .001) for eliminating affirmative action. Notably, the signs of the paths were both opposite (i.e., negative) fro situational motivation to cognitive entrepreneurship. This suggests that the more situationally motivated for problem solving, the more one becomes non-entrepreneuri in problem solving. To understand better this su se m one is al rprising finding, I conducted Sobel?s test to see the effect b r r s ffirmative action. etween situational motivation and cognitive retrogression via CEPS. In both issues, the indirect effects between situational motivation and cognitive retrogression via CEPS were significant: .454 (p < .001) for war in Iraq and .410 (p < .001) fo eliminating affirmative action. For cognitive multilateralism, in both issues the indirect paths via CEPS were not significant: -.007 (n. s.) for war in Iraq and .014 (n. s.) fo eliminating affirmative action. For cognitive commitment, I found a significant effect. A the situational motivation heightens, cognitive commitment increases: .111 (p < .001) for war in Iraq and .075 (p < .01) for eliminating affirmative action. Finally, for cognitive suspension, I found that as situational motivation increases, cognitive suspension decreases: -.096 (p < .01) for war in Iraq and -.078 (p < .01) for eliminating a 282 Notably, in the SITCEPS model, I found that cognitive retrogression is the mos salient conceptual subdimension to conceptualize cognitive entrepreneurship in prob solving (i.e., -.933 for war in Iraq and -. 972 for eliminating affirmative action). Cognitive retrogression refers to a cognitive tendency of doing backward reasoning (i ?a conclusion dictates evidence that secures and warrants the chosen conclusion.?). other words, as situational motivation in problem solving grows, one tends to move backward in thinking about the problem. From the finding, I conclude that situa motivation tends to pressure problem solvers to hastily turn to a conclusion and op the hastily drawn conclusion. This suggests that the heightened situational pro t lem .e., In tional timize blem l more problematic and more connected) tend to trigger a non- entrepr vely l S analysis. In addition, in H19, as expected, I found that the more o m e to Relationships between CAPS and CEPS perceptions (e.g., fee eneurial mindset. It explains why our problem-solving efforts are very often ineffective even if we are very eager and pressured (i.e., ?motivated?) to work on problem resolution. In other words, a problem solver?s motivation cannot be exclusi equated with his or her adoption of the entrepreneurial cognitive strategy. In summary, I found similar support for H16-H18 for situational perceptua variables as in the SITCAP ne possesses a referent criterion, the less one becomes entrepreneurial in proble solving. However, I failed to find support for H20. The finding suggests the opposit relationship between situational motivation and cognitive entrepreneurship in problem solving. It thus suggests that from the heightened motivation, problem solvers tend mobilize available cognitive resources backwardly?to optimize the selected solution to be more conclusive and convincing. 283 I constructed two nonrecursive models that investigate the simultaneous causal influences between CEPS and CAPS. Nonrecursive models are often difficult to solve mathem 16 blem p < .001) for eliminating affirmative action. This indicates that the more active a ion es not atically. I tried both of the model specifications and found that only the Figure model (i.e., referent criterion?CEPS) converged successfully in both issues. Thus, I used the Figure 16 model to test the hypotheses regarding bidirectional causality between CEPS and CAPS. For H21, I expected a positive causal influence from cognitive entrepreneurship in problem solving to communicant activeness in problem solving. However, I found no support for this prediction in both issues: -.959 (n. s.) in war in Iraq and -.892 (n. s.) in eliminating affirmative action. For H22, I predicted a negative causal influence from communicant activeness in problem solving to cognitive entrepreneurship in pro solving. I found support for this expectation in both issues: -.919 (p < .001) for war in Iraq and -.622 ( re one?s communicative behaviors (i.e., high in information forefending, forwarding, and seeking), the less one will become entrepreneurial in problem solving (i.e., doing more cognitive retrogression in problem solving). The accumulated knowledge and experience from problem-solving efforts in dealing with informat seems to decrease an entrepreneurial mindset because of more expertise and/or confidence on the given problem. For H21, I expected that higher cognitive entrepreneurship in problem solving will lead to higher communicant activeness. However, in retrospect, I failed to consider that the low cognitive entrepreneurial problem solver could also be active in communication behavior. In other words, high cognitive non-entrepreneurship do 284 mean passiveness in communication behavior. As we found routinely, many non- entrepreneurial problem solvers tend to vigorously seek and forward information that is consist trast, the reverse causal flow seems to be logical conceptually. As a riences heightened communicant activeness, she or he is likely to develop PS seems to reflect such a negative concep d ent with their preferred conclusion. Thus, it was not conceptually and empirically valid to specify a positive causal relationship between high cognitive entrepreneurship and communicant activeness in problem solving. The nonsignificant path from cognitive entrepreneurship to communicant activeness seems to reflect such a notion. In con problem solver expe a good deal of knowledge and preference on how to solve the problem. Then, such a preference (i.e., high information forefending) would lead to more cognitive retrogression (i.e., optimizing a preferred conclusion backwardly) and less tolerance, less commitment to all the available candidate solution and ideas, and less suspension of finalizing a solution. As a result, the problem solver becomes less entrepreneurial. Therefore, the negative path from CAPS to CE tual relationship. In summary, from the H21 and H22 tests, I found that communicant activeness in problem solving tends to reduce the problem solver?s cognitive entrepreneurship eventually, while the reverse cannot be assumed. Additional Analysis As I discussed, the H21 prediction was a conceptual mistake in that both high an low entrepreneurial problem solvers could be high in communicant activeness. There is no conceptual reason that the less entrepreneurial problem solvers should be low in communicant activeness. To test the validity of this reasoning, I conducted an additional 285 analysis of how the high and low entrepreneurial problem solvers would differ in communicant activeness. Earlier, I developed a model of cognitive alpha and omega strategies (CAOS) reflecti or vey n. I then selected e p and cog formation ation e solvers fend alpha ng a problem solver?s reasoning direction: i.e., ?evidence?conclusion? f forward reasoning and ?conclusion?evidence? for backward reasoning. I named those forward reasoning problem solvers as the cognitive alpha group, and those in the backward reasoning group as cognitive omega. Using the CAOS model, I sorted sur participants into three groups: high, medium, and low in cognitive retrogressio the high and low groups and named them cognitive omega (i.e., high in cognitiv retrogression) and cognitive alpha (i.e., low in cognitive retrogression). I expected similar model structures of CAPS in both the cognitive alpha grou nitive omega groups, but I predicted that the two groups will differ in in selectivity. Specifically, I expected the cognitive alpha group would have a low path coefficient in information forefending and a high one in information permitting. In contrast, I predicted that the cognitive omega group would show the opposite pattern?a high path coefficient in information forefending and a low path coefficient in inform permitting. If the patterns developed as I expected, this should become evidence for th explanation of a nonsignificant path in H21: high or low entrepreneurial problem do not differ in their information transmission and acquisition but only in their information selectivity (i.e., non-entrepreneurial problem solvers tend to more fore information.). As I expected, information selectivity was different between the cognitive group and the cognitive omega group in both issues. Specifically, the cognitive alpha 286 groups in two issues were lower in information forefending (i.e., .488 for war in Iraq and .449 for eliminating affirmative action) and relatively higher in information permitting (i.e., .538 for war in Iraq and .524 for eliminating affirmative action). I found a reversed pattern in the cognitive omega group. Cognitive omega groups were higher in information forefending (i.e., .656 for war in Iraq and .776 for eliminating affirmat action) and relatively lower in information permitting (i.e., .496 for war in Iraq a for eliminatin ive nd .246 g affirmative action). The two groups possess similar structural path coefficients in information transmissio tion acquisition, whereas in informa tion- lvers roblem lem , I ommunicant activeness in problem solving increases, thos as information forefending, inform rmation seeking, n and informa tion selectivity between cognitive alpha and omega groups shows an interac like pattern. Thus, I concluded that the additional analysis of CAOS and CAPS reinforced my explanation of why H21 failed to get support. The low entrepreneurial problem so can also be active in communication behavior, just as those high entrepreneurial p solvers are active. Thus, this at least in part explains why my conceptual prediction in H21 failed. Overall Discussion To recap, the present study finds support for communicant activeness in prob solving. Tests showed that CAPS is a multidimensional construct. In all six dimensions found the expected positive structural path coefficients: as one?s c e subdimensions such ation permitting, information forwarding, information sharing, info and information processing increase as well. 287 I introduced a new concept, cognitive entrepreneurship in problem solving, wh consists of cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspen ich sion. Except for cognitive retrogression, all subdimensions were posited to have ons. , of issue n icant dition, as one has a stronger referent criterio e in positive relationships with their higher order construct, cognitive entrepreneurship. Cognitive retrogression captures how one would deploy a backward reasoning strategy in problem solving. In general, I found support for these predicti However, I found only partial support for cognitive commitment. In two tests, I found a positive path (in eliminating affirmative action) and a negative path (in war in Iraq). So from the current data, we cannot say definitively whether cognitive commitment generally increases as cognitive entrepreneurship increases. This may be a matter sensitivity. A future study with multi-items and comparisons across more diverse types of problems should bring clearer understanding. This study?s findings suggest that high problem recognition and level of involvement with low constraint recognition increases one?s situational motivation i problem solving. As the situational motivation in problem solving grows, commun activeness in problem solving increases. In ad n, this tends to increase communicant activeness in problem solving. However, if one has a heightened situational motivation, she or he tends to have low cognitive entrepreneurship in problem solving. The presence of a stronger referent criterion tends to lower the cognitive entrepreneurial approach in problem solving. From the bidirectional analysis, I found that problem solvers with high communicant activeness in problem solving are not necessarily entrepreneurial in their problem solving. Low and high entrepreneurial problem solvers both can be activ 288 communication behavior. Nonetheless, as communicant activeness in problem solving increases, the lower the cognitive entrepreneurship in problem solving one tends to develop. Finally, the cognitive alpha group (i.e., the entrepreneurial problem solvers) tends to be less forefending and more permitting in information acquisition and transmission. In contrast, the cognitive omega group (i.e., the non-entrepreneurial h informa e unique ry placing informa problem solvers) tend to be more forefending and less permitting in dealing wit tion during problem solving. To conclude, in the present study, I found a good amount of evidence for communicant activeness in problem solving, cognitive entrepreneurship in problem solving, and the situational theory of problem solving. These constructs describ cognitive and communicative features during problematic situations. In addition, they explain how and why those unique features arise in some situations and not in others. Yet, I found a few unexpected findings (e.g., the higher the situational motivation in problem solving, the lower the cognitive entrepreneurship) that provoke further theoretical development. In future study, such serendipity will provide a chance for theo elaboration. Implications This study has introduced two new concepts of communicant activeness in problem solving and cognitive entrepreneurship in problem solving. With the new concepts, the situational theory of publics becomes a more general theory by re tion seeking and processing (i.e., ?particular?) with six subdimensions in communicant activeness in problem solving and four subdimensions in cognitive entrepreneurship in problem solving (?the general?). The resulting theory and models 289 have implications for communication theory in general as well as for subfields such as public relations, health and risk communication, political communication, and con resolution. In the following section, I will discuss the implications of communicant activeness in problem solving (CAPS), cognitive entrepreneurship in problem solving (CEPS), and the situational theory of problem solving (STOPS). Implication of CAPS The CAPS model I have developed aims to capture the notion of intercommunication. A dominant view is that communication is an individual ac information inflow or outflow, not an individual?s act of information interflow simultaneously. Prior conceptions of communic flict t of ant activeness hinged on information learnin l ation th r ants g potential. However, to explain information interflow, we need both conceptua nuts and bolts. With the concepts of information sharing, information forwarding, inform processing, and information seeking, we can look at how communicants interlock wi each other. Active communicants who are dealing with a life problem are not only seeking information about it but forwarding information about the problem and solutions to others. They are not only active in information taking but also active in information giving. Some communicants are active but may not seek information because of their successful problem solving in past (P. K. Hamilton, 1992). With their subjective confidence that they developed in dealing with a problem, they do not actively collect information about the problem but actively forward and forefend information to othe communicants. It has been conceptually inconvenient to explain active communic who are active in information giving but not active in information taking. Previous 290 theories could not capture such type of active communicants because they did not explicitly conceptualize information giving. In addition, I have introduced another conceptual dimension, information selection, to capture how a communicant deals with information as having heightened activeness in problem solving. We routinely encounter active communicants i life who selectively n our daily share, forward, process, or seek information about some problem. rtain ways of selecting information. Our society, thus, is like a box of nuts and bolts of different sizes Communicants not only are interconnected with each other, but they have ce and shapes. Matching such a mixed pile of nuts and bolts is not a simple task to accomplish. Indeed, many information forwarders find it difficult to meet information processors (who are possessed by other life problems) who are likely to take information as the forwarders wish. Finding the right fits is more difficult than most message senders believe because of the information selectivity that problem solvers tend to develop the problem-solving periods. CAPS, Communicant Network, and Model of Meso-Level Intercommunication With CAPS, we can break down the process of trafficking information among individuals within a communication network. To explain, I introduce two terms communicants and peripheral communicants. Focal communicant refers to a central person who plays a role of information station regarding a problem. The focal communicant actively ?inhales and exhales? information in an effort to solve a problem. He or she is actively seeking, forefending, and forwarding information about the Within the boundary of the communication network in which he or she is situated, a foca communicant becomes a driving force to locate and relocate information about a prob over : focal problem. l lem 291 across one?s interconnected communicant networks. In contrast, peripheral communicants refer to those communicants who are neighbors to focal communicants within ter d communicative behavior as focal communicants have and do. ants are not necessarily bounded by geographic proxim et ts ation to other communicants in an effort to solve a problem they feel important. Notably, as shown in past research, interpersonal communication networks, rather than mass-mediated channels, such as word-of-mouth spread among people who are acquainted with each other (e.g., ?weak ties,? Granovetter, 1983) are by far the most effective method of communication (Rosen, 2000). Thus, building a model that conceptually illustrates the process of information trafficking among individual communicants would be theoretically useful. Although prior studies distinguish the communication process by the medium utilized, this often misleads our study of communication effectiveness. For example, scholars of mass media effects distinguish mass-mediated communication from interpersonal communication in building a communication model. However, such a a communicative network but who are less active on the problem. Peripheral communicants tend to passively process, permit, and share information they encoun from the communicative interactions. They may or may not cultivate a similar situational perception an Focal and peripheral communic ity. Focal communicants can be various types of people and social figures. They could be, for example, our geographically close friends and neighbors who are ups about some issue, media reporters who investigate and report publicly some new threa to readers, government officials working for a citizen safety issue, or a corporate marketing staff that is desperate to increase sales. Such focal communicants attempt to transfer inform 292 distinction is futile in studying communication effectiveness from information giver to information takers in that the mass-media cannot exclude other communicant media in its effects (Chaffee, 1982). Too often, mass-mediated communication and interpersonal communication are closely intertwined, almost inseparable from other communication networks such as mass-mediated communication flows (Chaffee, 1982). The approach in studying communication effectiveness by distinguishing the medium would only be a phenotypic account that prohibits better understanding. A better alternative puts the emphasis on ?communicants,? not the ?medium? of communication, to describe information trafficking across people. In the below model, a focal communicant can be a person (e.g., a roommate) or a mass medium (e.g., a local newspaper). In either case a person or a mass medium, plays a role of information provide, acquirer, and selector as a communicant to other communicants. Thus, without distinguishing between medium and person, we can conceptually describe information interflow with a single frame. The llowing figure describes information trafficking among the interconnected fo communicants. Figure 33: Meso-level intercommunication process among communicants. 293 This model illustrates the active communicant?s role in transmitting information about some proble m he or she is actively working on. In the model of communicant activen ., active some ot perceive . cants only when they have an opportu rmation nd ess in problem solving, active communicants?focal communicants?are seeking information from other referents. Focal communicants, as a way of problem solving (e.g effectuating), forward information (e.g., about problems and/or solutions) to other communicants who are most likely within their routine relationship network. Such forwarding would, to some extent, have effects on information processors??peripheral communicants?perception regarding the problem. If peripheral communicants find personal connection to the forwarded problem, then they would become active communicants. However, it is possible that peripheral communicants would n enough of a connection to elevate one?s problem perception and communicant activeness They might toss information, reactively, to other communi nity (e.g., ?FYI?). Importantly, this meso-level model takes information trafficking as a joint function of information transmission efforts as well as info acquisition efforts among interconnected communicants, not attributing it to the sole function of one party. Thus, it models the intercommunication process, not communication itself. CAPS provides conceptual nuts and bolts to describe how some problems a solutions are exchanged across the social pipelines. It will reserve a fresh look for communication phenomena, such as opinion leadership, agenda setting, and diffusion of innovation. CAPS and Problem-Solving Potential CAPS and its subconstructs provide a way to explain how active publics behave 294 in some social conflicts and to estimate the extent to which a conflict would bear resolution potential. In the conceptual diagram in Figure 34, I summarize how the six subdimensions in CAPS model can predict the three key problem solving potentials. Figure 34: Communication behaviors and problem-solving potential in controversial social issues. Predicting symmetrical resolution potential: Open versus closed problem solver. N personal problems evolve into social problems. Yet, many individual problems ha some potential to evolve into social problems if many people are affected by a sam problem source. For example, a construction plan by a company in a quiet neig ot all ve e hborhood would become an issue resulting in a collective problem-solving effort if neighbors 295 organize to challenge the company. Problems that attract multiple stakeholders often result in conflict. Often, one party?s proposal for problem resolution is less attractive to another. Because problem solvers apply their own evaluative lens?i.e., becoming selective?evolved from their independent inquiring phases, another party?s solution is often d osed potenti al ating). Sharing a em, its cause, and some negative consequences helps individ get any istasteful. Problem solvers in a conflict situation are tempted to refute the other party?s proposal. Because problem solvers abide by their own decisional rules, knowing how much individual problem holders have a tendency for information selection (forefending) can be a good predictor for symmetrical problem-solving potential. In other words, a problem solver?s openness to reviewing information from competing perspectives, even if distasteful, can be a litmus test for diagnosing the potential for a symmetrical resolution of a social problem. Thus, here I refer to an open problem solver as a person who is willing to use any information to increase problem-solving potential. I refer to a cl problem solver as a person who discriminates against distasteful information incompatible with his or her frame of reference. They want to increase problem-solving al by subscribing to a certain type of information. Collective action potential: Individual versus collective problem solver. As mentioned, information transmission is a critical part of problem solving (i.e., effectuating a solution). By giving information about a problem, those isolated individu problem solvers evolve into a social collectivity (i.e., collective effectu similar perception about a probl ual problem solvers increase their problem-solving potential (e.g., easier to mobilize resources to the problem). In the inquiring stage, a communicant cannot 296 297 Also, by d oblem solvers (e.g., an activist group). Thus, knowing who is likely to em n. problem solvers. In contrast, collective problem solvers not only recognize a problem but also recognize comrade problem solvers. The only way that individuals communize a similar sense of problem perception is through forwarding and sharing information about the problem with one?s neighboring communicants. Stagnancy potential: Situational versus dormant/chronic problem solver. Some problems are more enduring than others for various reasons. Problem solvers may suffer potentially useful information without communicating about his or her problematic state: i.e., problem forwarding. A communicant should talk about his or her problem. provoking other communicants to recognize a problematic state, individual problem solvers can enhance the pool of potentially useful information, can divide the costs for problem solving, and can increase their bargaining power in demanding resources from a relevant party. Information forwarding and permitting is at the heart of the locating an networking with other individual problem solvers. Giving information about a problematic situation is thus a necessary condition for a problem to produce a group of collective pr it information about a problem and a solution to other communicants explains what kind of problem would have potential for collective actio I refer to an individual problem solver as a person who is working on a problem in an independent and isolated way without knowing other problem solvers. A collective problem solver is a person who tries to solve the problem through coordinated efforts with others. The key difference between individual and collective problem solvers is whether one has a cross-awareness between problem solvers. In other words, individual problem solvers recognize a problem but may not recognize the presence of other from absence of a solution (e.g., curing can r ons may be obvious, yet problem solv er fr c esources to enact a solution; or they have difficulties mobilizing the attention of those who have necessary resources (e.g., obtaining budget for a community educational facility from government). If a problematic situation continues, problem so s be chronic problem solvers unless they leave th ion p o (i.e -o ht syndrome). To highlight such a distinction in terms of on, situational problem solver from a chronic problem solver a d t pr m so situational problem solver refers to a problem solver whose problem is solvable within a short amount of ple, a person with the flu would try to find a cure effo ly. Yet, reso f oble t uld end simply as time passes. contrast, a c pro o fers o who has procrastinated caus th sence o n or for inability to mobilize resources. For stanc iabetes patients are chronically active trying to learn and manage their sease ou ut their ime. In contrast, a dormant problem solver is a person who s found a in t a w pa s r problem-solving effort. From eir past proble olvin t ant m rs would possess some owledge. However, th re b solv othe important problems; hence, ey temporarily leave the problematic state until a new problematic state arises from it. Taking information urrent as possi o solve a problem would decrease a tagnant? state in problem ing. If competing parties are dormant or chronic in a cial conflict, the members tend to less active in information acquisition. Thus, with the acquisition tendency, we can predict how cer). O soluti ers e si suff om a la k of r lver ., f I di oble co r- uis lv me flig h a er. tuat syc pro and hol blem gic du or ally rati man ight sting A rtful time. For exam luti bl f a concluding solutio the roni on o em s th lve e pr r re ma to tic s a pr ate ble wo m solverIn be in di ha th kn th ?s so knowledge of problem ch e ab y d gho prob e of e, m thr an lifet he past g effor ey a lem m-s nd s, d usy no orm in use pro ing s hi ble or so r m he lve ore as c solv ble t solver?s information 298 likely a eful in n pes of publics derived from th situational active public, open-situational activist public, closed-situational activist public, open-chronic activist public, and closed-chronic activist public (see Table 29). given social conflict would be stagnant in conflict resolution. CAPS and Typology of Publics The situational theory of publics classifies publics into categories such as nonpublic, latent, aware, and active publics predicted by the interaction of three independent variables, problem recognition, constraint recognition, and level of involvement?i.e., 2 x 2 x 2 = 8 (J. Grunig & Hunt, 1984). However, the STP conceptualizes communication behavior only by the information taking dimension (i.e., information seeking and processing). Active and passive communication behaviors explain only when people, as information consumers, are more or less likely to take information. Hence, the active, aware, latent, and nonpublic categories are us understanding how a public is likely to consume information, but not useful in explaining how a public is likely to interpret, produce, and transmit information to other publics and how selective they would be. The CAPS model offers a more comprehensive framework that develops a new typology of publics using its general conception of communicant behavior. Using and combining six subdimensions under information selection, transmission, and acquisitio generates eight different publics. I distinguish eight different ty ree key characteristics in problem solving behavior. These are open-dormant passive public, closed-dormant passive public, open-situational active public, closed- 299 300 Table 29 Dimensions of Communicant Activeness and Types of Publics Information Transmission Forw at n A uisi arding Sharing Inform io cq tion Information Acquisition k Processing Seeking Processing Forefending Closed-Situational Activ ic ( t A e P osed-Chronic ctivist Public ( Problem Solver) A etrical Potential Rich) Closed-Situational Active Public (Individual Problem Solver) (Asymmetrical Potential Rich) Closed-Dormant Passive Public (Individual Pro Solver) (Asymmetrical P tial Rich) See ing ist Publ Collec ive Problem Solver) ( symm trical otential Rich) Cl A Collective ( symm blem oten m ectio Infor ation Sel n Permitting Open-Situational Activ ic ( t (Symmetrical Po al ) pen-Chronic ctivist Public ( Problem Solver) S etrical Potential Rich) Open-Situational Active Public (Individual Problem Solver) (Symmetrical Potential Rich) Open-Dormant Passive Public (Individual Pro Solver) (Symmetrical P al Rich) ist Publ Collec ive Problem Solver) tenti Rich O A Collective ( ymm blem otenti Dewey (1927) provided a classic definition of a public that avoids the common confusion in lay use of the term public (e.g., a general public). In Dewey?s theory, a public is a group of peopl e who 1) face a similar problem, 2) recognize that the problem and Hu lic blic. n of ) differen behavio consist a problem their ex m solver b exists, and 3) organize to do something about the problem. Following Dewey, J. Grunig nt (1984) distinguished four types of publics. They labeled a group of people who face a similar problem but do not detect the problem a latent public. When group members subsequently recognize the problem, they become an aware public. If the pub organizes to discuss and do something about the problem, they become an active pu Finally, as a logical extension, those groups of people do not meet any conditio Dewey?s notion of publics are called nonpublics. J. Grunig and Hunt (1984 tiated these four types of publics by ?the extent to which they participate in active r to do something about? a problem (p. 145). Thus, their typology of publics is ent and compatible with the present model of communicant activeness in atic situation. Specifically, nonpublic, latent, aware, and active publics differ in tent of problem solving efforts?i.e., communicant activeness. As a proble ecomes more serious in tackling a recognized problem, he or she will be more active in information acquisition, transmission, and selection. In Figure 35, I offer a three dimensional model to illustrate such conceptual correspondence between types of publics and three communicant activeness dimensions. 301 Figure 35: Three dimensions of communicant activeness and eight types of publics. Information Transmission Information Acquisition Information Sharing n. roach h ig ortance of viewing a public as having situational propert practitioners can anticipate which types of publics (e.g., closed-situational activist) emerge with what types of behavioral characteristics (e.g., collective problem solver), From the CAPS conception, a public is a problem solver who uses communication behavior as a coping mechanism to inquire and effectuate a chosen solution. A public may be individually or collectively working toward problem resolutio A public as a problem solver may have an enduring or a transient problem, may app problem resolution individually or collectively, and may take an open or closed approac in using information during problem solving process. I emphasize, following J. Grun (1968, 1989, 1997, 2003, 2005), the imp ies in terms of communicative activeness to end a problematic situation. As is easily observable, different problems produce different types of publics. Across problems, publics? compositions differ as well. Hence, if public relations (REACTIVE) Information Forwarding ROACTIVE) (P Information Processing Information Seeking Open-Situational Active Public (Symmetrical Potential Rich) (Individual Problem Solver) Information Selection Information Forefending Relevance/ Specific) (Systematic/ Information Perm Related/General) itting (Random/ Open-Chronic Activist (active) (Symmetrical Potential Rich) (Collective Problem Solver) Open-Dormant Passive Public Closed- Situational (Asymmetrical Potential Rich) Active Public (Individual Problem Solver) Open- Situational Activist (Symmetrical Potential Rich) (Collective Problem Solver) Closed-Situational Activist (Asymmetrical Potential Rich) (Collective Problem Solver) Closed-Chronic Activist (active) (Asymmetrical Potential Rich) (Collective Problem Solver) Closed-Dormant Passive Public 302 they will make a more strategic choice in dealing with publics (e.g., negotiation or information campaign). CAPS and Criticisms of the Situational Theory of Publics Criticism about failure to explain the communicative nature of publics. Vasquez and Taylor (2001) said that the STP is limited because it heavily hinges on a ?socio- /psycho-centric view of a public,? a ?tautological conceptualization of individual and public,? and a ?view of communication that is outdated or out of touch with actual observations? (p. 150). Above all, they said that the STP failed to explain ?the contemporary communicative nature of a public? (p. 150). They asserted: The situational perspective identifies communication as central to the emer after the characteristics and composition of a public have been identified does ?logic? of the situational perspective is grounded in the assumption of conceptualization and operationalization is a source of tension for the situation centric view of a public. [italics added] (Vasquez & Taylor, 2001, p. 150) CAPS grants STP a conceptual coherence that resolves this ?source of tension? by its extended and general conception of communication behavior. Specifically, CAPS defends STP against the criticisms that it uses ?social-psychological variables gence of a public but uses social-psychological variables to investigate a public. Only communication become important as an outcome effect. Yet, the underlying communication?public discussion, debate, and argument. The difference in al perspective that has the effect of orienting the researcher to a socio-/psycho- ? and treats ?comm e to . mer unication?as an outcome effect? (Vasquez & Taylor, 2001, p. 150). Opposit the criticism, STP?s use of ?social-psychological? variables and viewing communication as a dependent variable to study publics is not a problem, but a strength of the theory The conception of communication as an outcome of problem perception is indeed the core virtue that the STP uniquely has brought into the field of communication. Prior to STP and another audience-centered model such as uses and gratifications theory (Blu 303 & Katz, 1974), communicants were only described as pathologically helpless targets of message sender?s persuasive intent. They were fed or merely declined messages given by senders. As an assumption, communication was almost always an unquestioned causal ssage ns. In this vein, STP considers communication to be as ?outcome effect? corresponding to a communicant?s own problem perception in a given situation. A public?s problem perception is imperative for understanding when publics do or do not engage in problem-solving actions (e.g., debate against people with a competing perspective about a social issue). Knowing the degree of problem perception is, in turn, essential to understand when publics arise and evolve to do something about the problem. We can never understand how and why people become communicatively active without using ?social-psychological? variables such as the degree of problem recognition and the perceiv t the s to discuss the ?nature, role, and influen e created, variable that a sender does to receivers, not an outcome variable that reflects a me taker?s needs and purposes across their life situatio ed degree of involvement. For problem solvers (i.e., publics), information is no simply given but actively sought, forwarded, and selected purposefully. Few prior communication theories, if any, could dare to theorize in that way. Thus, attacking STP?s conception as problematic because it treats communication as an outcome is missing the point (ignoratio elenchi). 24 CAPS provides a framework for researcher ce of communication? by understanding how ?problematic situations ar raised, and sustained through the symbolic convergence (configuring and reconfiguring) of messages? (Vasquez & Taylor, 2001, p. 150). The information transmission variables This term refers to a missing the point ?the arguer is ignorant of the logical implications of his or her own premises and, as a result, draws a conclusion that misses the point entirely? (Hurley, 1997, p. 131). 24 304 in CAPS provide conceptual links that enhance the STP?s contribution to balancing a sender-biased conception and upsetting the traditional communication assumption Through the concepts of information forwarding and information sharing, CAPS allows researchers to set communication behavior as independent variables as well. That is, active communicants who are publics tend to forward information about a problemati state as well as their own solution proposals to other communicants. Forwarding or sharing information about their problem is a means to effectuate their preferred solution Thus, by setting up communication as an independent variable, we can understand how active publics increase problem perception among others by providing them . c . with informa ted that ations ing model can be tested to see the reversa m tion. Then, we can illustrate a process through which a communicant, a member of public, begins to perceive a problem and a social collectivity arises in the social mesh of communicants to deal with their common problems. Hallahan (2001) lamen ?issue process? and ?issue dynamics? received ?too little attention? in public rel research (p. 49). He (2001) thus posited a question, ?what role does communication play in creating problem recognition, involvement, and constraint recognition?,? and requested a conceptual effort to put communication as ?independent variable? (p. 49). I believe the new model of CAPS answers Hallahan?s call for new research. Although in the present study I did not test it, the follow l effect from communication behavior as a cause to the other communicants? perceptions in a problematic situation. In the following model, I predict the effect of communication as an independent variable will reduce perceptual gaps among interacting communicants. Thus, as focal and peripheral communicant activeness increases, the perceptual gaps between two interacting communicants will decrease about the proble 305 and the ?accuracy,? ?understanding,? and ?agreement? (two concepts in the Coorientation model, McLeod & Chaffee, 1972) about the problem will increase. Figure 36: Effects of communicant activeness to perceptual and communicative behavioral coorientation among problem solvers. ow individual problem holders begin to realize commonality among rmed into co informa r cross meta-perception about problem perception among individual problem solvers (e.g., at isolated publics turn into a social collectivity or a system to coordinate their behavior Importantly, CAPS advocates the conception of ?collective consciousness? in STP as a viable explanation of how an isolated individual member of a public transforms into a cohesive collective public in some problems. The communicant interaction captured in CAPS explains h themselves. Then, it explains how the individual problem solvers are transfo llective problem solvers. Information forwarding by activist communicants, which is parallel to tion seeking by active communicants, increases the chance of meta-perception o individual problem solvers are aware that other individual problem holders are aware th I am making an effort to solve this problem.). This is the process and the point at which 306 toward problem resolution. Thus, CAPS rebuts the criticism of ?the tautological conceptualization of individual and public? by explaining how a collective conscio arises. By CAPS, we can conceptually differentiate publics as individual problem solvers (e.g., open/chronic active publics) and collective problem solvers (e.g., open/chronic activist publics). Further, the conceptual dimensions of CAPS, such as information selection, transmission, and acquisition, explain not only the detailed evolving process from individual to collective problem solving, but also the extent to which a problem the potential for a collective problem solving. Vasquez and Taylor (2001) attac usness has ked the situational theory of publics as ?outdated ndence validity?whether or not concepts and relations specified can be 9- ling ns the ocial on, the homo narrans perspective, the situatio t, raised, and lacking in correspo seen in the observations of everyday life? (p. 150). They said that STP ?marginalized the role of communication process and dynamics? because of its ?socio/psycho-centric conceptualization of a public? (Vasquez & Taylor, 2001, pp. 14 150). Although most attacks against the situational perspective have come as straw man arguments (e.g., the ?tautology? claim of public concept against the STP), their cal for emphasizing ?communication process and dynamics? seems to be a constructive criticism for advancing our understanding of publics. By extending the conceptual scope of communicant behavior beyond information acquisition, as CAPS does, I believe the situational theory of publics better explai ?communication process and dynamics? inherent in the emergence of publics and s issues. Just as they claim for their favored positi nal theory of publics with CAPS is better capable to explain the process tha ?Symbolic reality and individual knowledge of a problematic situation are created, 307 and sustained through the symbolic convergence (configuring and reconfiguring) of messages? (Vasquez & Taylor, 2001, p. 150). Vasquez and Taylor (2001) requested a refinement such that: ?One could question whether or not the STP information- processing and information-seeking behaviors are appropriate variables of concern f explicating the communicative nature of a public. A single coherent approach is needed? (p. 150). CAPS would be a single coherent approach to respond to their request for further refinement of the situational theory and even for their ?rhetocentric? homo narrans perspective (Vasquez & Taylor, 2001). CAPS and Active Public or s without Information Seeking hat some ow sufficiently avoid negative communications, will indeed reduce their level of a respondent who continues to be highly involved and low in constraint but has in essence decided he or she recognized. The question is would situational theory continue to count this individual as being high in communication activity? any uncertainty, will be contrary to situational theory?s prediction. [italics added] The situational theory of publics predicts active publics will seek information about a problem. However, P. K. Hamilton (1992) posed one intriguing question regarding the predictive power of the situational theory of publics. He found t active publics who are high in problem recognition and involvement perception with l constraint recognition are not actively seeking information about the problem. P. K. P. K. Hamilton (1992) explained: ?individuals who have made up their mind regarding the election, and can information seeking. Indeed this specific case will find how to solve the problem If so, then the reduced level of information seeking, because there is no longer (p. 144) CAPS provides a conceptually coherent answer to P. K. Hamilton?s question. With its general conception of communicant activeness other than information acquisition (e.g., information seeking), CAPS predicts a special active public that would be low in 308 information seeking or processing, will be high in information transmission (i.e., information forwarding) and high in information selectivity (i.e., information forefendin Conceptually, such an active public is, CAPS predicts, in the effectuating phase who ?active? in translating their favored solution (e.g., actively promote their preferred candidate to other voters), while having little need for more information. In other words, the ?anomaly? public is active in information transmission and selectiveness afte passed the inquiring phase. Thus, P. K. Hamilton?s paradox of active publics is not a paradox from the CAPS point of view. CAPS and Campaigns Often we observe that apparently well thought-out mass-mediated campaigns a not successful. The model of communicant activeness, CAPS, illustrates this in two ways. Active communicants are likely g). is r it re to be selective in their informational activities. Thus, campai ts who gners who are selective in their information forwarding would be frustrated by the active information seeker?s information forefending. In contrast, those less active communicants are less selective (i.e., permissive) and thus would be less resistant to the attempts of an information forwarder?s selectivity. However, their permissiveness makes it difficult to catch and hold their attention. They are processing information, not seeking. Further, even if they pay attention, they are not likely to withhold information as information forwarders intend. They neither refute the informational contents nor withhold it. They are nonchalant and transient. At the same time, the new model of communicant activeness can enhance campaign effectiveness. Some communication theories encourage campaigners to segment their audience. Taking an example, the situational theory of publics predic 309 are likely to seek information with its three independent variables, problem recognition, constraint recognition, and level of involvement (J. Grunig, 1989, 1997). This theory suggests that public communicators should not attempt to communicate with those passive publics (who are engaged in information processing) and whose communicatio behaviors are almost random. Instead n , the situational theory recommends segmenting the to seek, keep, and use the information. Thus, campai s in n der ers are individuals knowledge by others. The CAPS framework considers such ?reactive leadership? as passive communicant behavior by the variable active communicants who are likely gners can increase the communication effectiveness and decrease the communication costs. In addition to information seeking and processing, the four additional variable CAPS (i.e., information forefending, information permitting, information forwarding, and information sharing) allow information campaigners to predict, first, when a communicant would refute or receive information from the campaigner and, second, when a communicant would voluntarily disperse information that he or she learns to other communicants (c.f. word-of-mouth campaign). With CAPS, public communicators can predict better when their information campaign would be more successful and less costly by allowing a more comprehensive audience segmentation. CAPS and Opinion Leadership Opinion leadership has been defined as the act of being sought out for informatio and advice on a topic (Troldahl & Van Dam, 1965). CAPS explains the opinion lea concept in a fresh way. Despite its conceptual appeal, opinion leadership concept hinges more on a reactive notion of leadership. In literature, the opinion lead who are ?being sought out? for their 310 informa ?proactive role arying degrees of opinion leadership using the information selectio ait. opts a e tion sharing. I suggest expanding the opinion leadership concept to leadership.? Opinion leaders (active communicants) at times offer information even without a request. They may voluntarily disperse opinions, issues, and solutions to neighboring communicants. CAPS extends reactive opinion leadership to a proactive with its generalized notion of communicant activeness. It explains how opinion leadership might differ across problems. For example, proactive opinion leaders are likely to seek information, forefend information, and forward information. Reactive opinion leaders are more likely to permit, process, and share information. We can conceptually distinguish v n, transmission, and acquisition dimensions in a more comprehensive way than information acquisition alone. In addition, CAPS extends opinion leadership beyond a non-varying personal tr The opinion leadership literature treats opinion leadership as an enduring personal characteristic. Yet, we encounter many situations in which an unlikely individual plays the role of opinion leader (e.g., uneducated, less prone to think). Because CAPS ad notion of ?situational? activeness, any individual can be an active opinion leader in som problem he or she is seriously trying to solve. Indeed, almost everyone is an active problem solver at some times for some issues. It is not surprising, therefore, that some unlikely people transmit or are sought for information by other communicants even when they deviate from a typical opinion leader profile (e.g., old, experienced, and/or highly educated). CAPS lifts the conceptual bar for lay people to become opinion leaders at some situation by relieving rigid assumption of enduring personal traits in opinion leadership. Any problem solver can be an opinion leader when he or she develops enough 311 confidence and competence through solving a problem. Extraordinary problem sit make even laypeople situational opinion leaders who forward/share their soluti uations ons with other c things is theory e ation seekers and processors and information forwarders and PS postulates that active commu n n theories (McComb & Shaw, 1972). It cleverly solved the puzzle of ommunicants. CAPS conceptually welcomes and resolves such an anomaly in previous opinion leadership studies. CAPS and Diffusion of Innovation The theory of the diffusion of innovations uses two simple variables, time and number of adopters. In brief, the increased ?number of adopters? of new ideas or a function of time (Rogers, 1995). Despite the theory?s parsimony, it is interesting to conceptually partial out the variance of the independent variable of ?time.? From a building perspective, time is neither interesting nor useful because time engulfs many possible causes. Hence, conceptually partialing out the variance in ?time? advances diffusion theory. CAPS can segment significant variance from the independent variable, ?time.? Specifically, CAPS explains the key dependent variables in the diffusion theory such as the adoption rate, the number of adopters, or the shape of adoption curve of som innovation (e.g., Chaffee?s, 1977, three curves of diffusion). These can be determined by the number of inform sharers regarding a given innovation/things. Also, because CA nicants not only learn something new but also educate others about it, the changing rate of communicants from information seeker to information forwarder of a innovation can explain a portion of the variance contained in time. CAPS and Agenda Setting Theory Agenda setting theory has become one of the most influential mass communicatio 312 empiric ior. hl, e e communicants? or at le mple, ly. g, al failure to find a strong mass communication effect on attitudes and behav By shifting its focus from ?what to think? (i.e., ?attitude change?) to ?what to think about? (i.e., audience?s ?learning? public issues and their priorities), McComb and Shaw (1972) found a strong mass media effect on audience thinking (McQuail & Winda 1993). In essence, the agenda setting hypothesis postulates that the mass media influenc public opinion by ?paying attention to some issues and neglecting others? (McQuail & Windahl, 1993, p. 104). Using the definition of communicant activeness in CAPS, active communicants are those who seek, forward, and forefend about some problem. When speculating on the role of media and actors in mass media institutions (i.e., reporters and editors), we can consider the mass communication actors (c.f., gatekeepers) to be ?activ ast quasi-active communicants who seek, forward, and forefend information about social problems. If we extend the conception of individuality or personality to the organizational persona of a mass medium, we can connect the CAPS framework to the agenda-setting hypothesis. Active media communicants are selective in sampling social problems/issues and prioritizing them by applying their own decision rules. For exa active media communicants (a reporter or editor who subjectively defines social issues) whose ?problem? is to fulfill a social watchdog responsibility select information routine Their communicant activeness may be triggered not by a personal problem but by a quasi-problem perception (e.g., ?although I am not personally interested in it, the side effects of this new medical pill seem to be a serious problem to my readers?) (J. Gruni 1983). They think some issues are more worthwhile to seek and forward to other communicants (audiences) while others are not: i.e., information forefending. 313 With the CAPS framework, those ?institutional elites? (e.g., members of the domina ommunicators, activist publics, and active audiences are functio o entation? and ferent n for problem resolution. Then, how can commu orienting effort between compet scribe nt coalition or public relations managers) or activist groups who are the ?information sources? for the media are also active or activist communicants. These institutional elites or organizational rebels actively promote their problem as a critical issue for society and demand public attention. Just as media reporters are active communicants in that they tend to forefend, seek, and forward information in dealing with information to readers, these organizational elites or rebels actively seek and forward information selectively to media and audiences. Although the routes or sources of problem recognition would be different, the resulting communicant actions of media reporters, active organizational c nally identical. CAPS conceptualizes media reporters as active communicants wh perceive a set of possible problems (agendas), who place problem priorities on that list (agenda setting), and who inquire and effectuate their problems for their problem (social watchdog) solving. CAPS and Coorientation: From Competitors to Cooperators Many social problems lead to social schisms because of a lack of ?coori among problem solvers (e.g., different ?orientation? between pro-choice activists pro-life activists). Achieving a cooriented state between problem solvers with dif definitions of the problem and its solutions is therefore a necessary conditio nication moderate a co itors? In a conflict, the potential for resolution depends on the extent of cognitive overlap between the competing parties. The extents to which conflicting parties sub 314 to similar sets of information and to the extent to which they have similar interpretatio of information, that is, ?coorientation,? will be keys to predicting a likelihood of collaboration and conflict resolution. By seeking, consuming, and generating a certain subset of information that fits with their desired end state (information forefending), people construct their own situational reality regarding the problem they face. In terms o the coorientation model (McLeod & Chaffee, 1973), rampant social schisms on many issues result from a lack of coorientation. CAPS explains the lack of coorientation (e.g., lacking understanding and/or agreement) as originati ns f ng from an unabated information ve communicants around problems. . beliefs ded that thy to liberate suppressed Iraq civilians. On the opposite side, some anti- war act f forefending tendency among acti For example, the U. S. military and government prohibited publicity about dead U S. soldiers from the war in Iraq (since March 2003). Active communicants become selective in information giving. The U. S. authorities in charge of the war were active communicants themselves who wanted to define, interpret, and construct the problem and solution in a way consistent with their interest and beliefs. They were forefending, seeking, and forwarding information to illustrate and define reality as their strong lead (referent criteria). They referred to the things, events, and segments of reality in a way consistent with their constructed solution (referent criterion). They conclu the war was wor ivist groups selectively looked for how many U. S. soldiers and civilians died from the war (information seeking and forwarding with strong forefending). An activist group?s strong belief of ?stop the war? (referent criterion) leads them to selective information seeking and forwarding. People reduce their conflict resolution potential because of the dissimilar sense o 315 reality, which is caused and reinforced by their subscription to filtered information. The degree of information forefending and permitting, therefore, becomes a critical indica to diagnose resolution potential in m tor any social problems held between competing parties. nformation forefending with different definitions of problem in a on rs). ng information forefen at I used to distinguish two groups in problem solving. I introduced a model of When problem holders are high in i s and solutions proposed, they tend to construct and become entrenched with certain aspect of reality?constructing a sense of reality by sampling only some porti of information?over the other alternatives. Not surprisingly, the potential of reaching an integrative solution decreases. Often, such an issue devolves to a chronic social problem. This, in turn, creates chronic rebellions against each other. Correspondingly, active or activist publics become habitual in their communicative efforts. They experience a chronic problematic situation and become chronically active publics. In many cases, such chronic problems breed closed mindsets in reviewing available information (e.g., a solution proposal from othe Having competing entities (e.g., activist groups) routinizes stro ding in their communicative actions. Then, a chain reaction, seeking and forwarding the forefended information to others, sets in motion. Each faction of collective problem solvers mirror what the other did previously thereafter. Implication of CEPS Cognitive entrepreneurship in problem solving consists of four underlying subfactors: cognitive retrogression, multilateralism, commitment, and suspension. Among four dimensions, I paid special attention to cognitive retrogression with its novelty and potential utility to other research. Cognitive retrogression is the primary concept th 316 cogniti n: i.e., lpha tion, ory dissonance theory explains. When a person faces d m ion), ade n. , & Guerrero, 1969). Such an instance is a cognitive alpha strategy whereby the problem vein, ve alpha and omega strategies (CAOS) in terms of their reasoning directio ?evidence?conclusion? for forward reasoning and ?conclusion?evidence? for backward reasoning. I referred to those forward reasoning problem solvers as the cognitive a group, and the backward reasoning group as the cognitive omega group. The CAOS model helps us explain some important phenomena and theories. In the following sec I will review the implications of CEPS and CAOS. CAOS and Dissonance The CAOS explains typical cases of what issonance (?perceived discrepancy? in terms of the situational theory of proble solving) with the joint conditions of other situational antecedent variables (e.g., high in problem recognition, level of involvement, constraint recognition, and referent criter the person would use a ?backward cognitive strategy? (i.e., cognitive omega approach) about the dissonance state. She or he may effortfully justify his or her choice by selectively seeking or avoiding some information that can fortify his or her readym decision. The cognitive omega approach is a fair reflection of the dissonance phenomenon. However, CAOS does not exclusively associate with activeness and backward reasoning. It is possible that problem solvers would dare to seek ?dissonant? informatio At times, we observe that problem solvers do not fear or shy away from dissonance arousing information or situations (Chaffee, Saphir, Graf, Sandvig, & Hahn, 2001; Carter, Pyszka solver pursues problem solving with a more entrepreneurial mindset. In this 317 CEPS o theory loses its explanatory power. CAOS and Information Selectivity Cognitive alpha and omega strategies explain why selectivity arises?one?s inclusion and exclusion of information to devise a set of solution alternatives. In the cognitive alpha strategy, one?s communicative efforts are more likely to be used to revise and reform the prior maladaptive solution, rather than to reinforce an old solution. Thus, people with high cognitive entrepreneurship in problem solving are more likely to be selective in the sense that she or he adopts an ?except this? approach?i.e., excluding an old ill-suited solution. Besides, as one?s information acquisition becomes comprehensive, certain guiding principles emerge with subjective confidence in evaluative tasks. This makes communicants more selective. Such selectivity takes the form of ?preference? rather than ?avoidance? as in the cognitive omega approach. In contrast, in the cognitive omega strategy, which is a default human cognitive strategy in problem solving, one?s communicative efforts are more likely to be used to reinforce and reconfirm a preferred solution. But, it tends to make one more maladaptive ntrepreneurship in problem ffers a more general conceptual frame beyond the point at which dissonance to the problematic situation. Again, people with low cognitive e solving are also likely to be selective in a different sense from the cognitive alpha approach. Now, problem solvers adopt an ?only those? approach?i.e., including information that reconfirms a preferred outcome state and increase confidence in the preference. Such selectivity takes the form of ?avoidance? as in the dissonance theory account. 318 In the present study, I found that the cognitive omega group had stronger selectivity (i.e., more information forefending and less information permitting), whereas the cognitive alpha group showed relatively low selectivity (i.e., less information forefending and more information permitting). Such a finding helps us understand how subdimensions of CAPS. With this understanding, we can introduce a normative knowledge about a better problem-solving practice. CAOS and Theories of Behavioral Intention Social psychologists have proposed several theories to explain why and how one?s attitude affects behavior. They have been puzzled by the so-called attitude-behavior inconsistency problem (e.g., the weak empirical evidence of attitude as a cause of a behavior). Such efforts resulted in the theory of reasoned action (the TRA) (Fishbein & Ajzen, 1975). In tackling the puzzle of strong attitudes but low behavioral interpretation, Fishbein (1967) devised a clever conceptual bridge, namely behavioral intention. Fishbein (1967) reasoned that the proximal cause of a behavior is one?s intention. Intention is antecedent to an action. The intention is, in turn, determined by one?s attitude toward the behavior and subjective norms (e.g., compliance to a specific norm of reference regarding the behavior). Attitudinal influence on a behavior is mediated by behavioral intention, which is a preliminary decision to act in a certain way. Eagley and Chaiken (1993) praised Fishbein?s approach to the attitude-behavioral inconsistency such that: The scientific and philosophical issue of how the mental event of holding an ing another psychological event, the formation of an intention, between the attitude information selectivity arises and how it specifically differs in six attitude is transformed into observable action was thus resolved by interpos and the behavior. Intention, a psychological construct distinct from attitude, 319 represents the person?s motivation in the sense of his or her conscious pla n to exert effort to carry out a behavior. (p. 168) Figure 37 summarizes the conceptual process that Fishbein and Ajzen (1975) proposed to explain the linkages attitude to a behavior. Figure 37: Conceptual model of the theory of reasoned action. *This figure is adapted from Ajzen and Fishbein (1980, p. 84). However, the TRA has a conceptual limitation that CAOS does not. In the CAOS conceptualization, our action or effectuating--the last step of problem solving--does not require our intention. For many problems, we take the cognitive omega strategy in which we skip much cognitive effort in order to solve the problem quickly. Thus, we may not need intention to behave. Indeed, Ajzen, Timko, and White (1982) found this deficiency to be a severe limitation of the theory of reasoned action. They criticized the theory because it was founded on the assumption that behavior is always volitional. And with 320 that assumption, the TRA only explains behaviors that are consciously thought ou beforehand. T t o fix the problem, Ajzen proposed a theory of planned behavior (the TPB, Ajzen, 1985) to explain non-volitional behaviors in the TRA (see Figure 34 for a summary of the TPB). Figure 38: Conceptual model of the theory of planned behavior. *This figure is adapted from Ajzen (1991, p. 182). Ajzen (1985) reformulated the theory of reasoned action to explain the cases in which behaviors can occur without a person's volitional control (i.e., intention). He introduced a new variable of ?perceived behavioral control??self-perception of how easy or difficult it is to perform the behavior. Perceived behavioral control is analogous to ?self-efficacy,? which refers to the ?conviction that one can successfully execute (a given) behavior? (Bandura, 1977, p. 193). Thus, when one has a sufficient level of perceived confidence, she or he would not necessarily go through ?intention? to execute an action (the dotted path from perceived behavioral control to intention in Figure 38). The inclusion of perceived behavioral control released the burden of the intention assumption (i.e., volition?action) for the reasoned action model. Especially, when we 321 explain well-rehearsed or trivial acts (e.g., easily executable actions or habitual behaviors), the intermediate conception of intention is of no use to explain behavior. However, Pavitt (2003) pointed out the weakness of theories that use behavioral intention in general: Other forces might prevent a person from behaving consistently with an intention, Nonetheless, the mental state has the power to direct behavior in an attempt to intentional explanations by their very nature cannot be causal because goals explanations rely on and some behavior is performed without any relevant intentions at all. achieve a person?s goals?. This is a good time to dismiss the argument that [intention] occur after the behavior requiring explanation. Intentional entpres conceptions of goals to explain future behaviors. [italics added] (Pavitt, 2003, p. 8) As Pavitt pointed out, we can easily identify situations in which people go from behavior to intention. Our choice of a behavior can precede our perceived intention when we do something in response to a strong urge and then look for justification: e.g., one might shoot another person and then reconstruct and subsequently elaborate one?s intention as self-defense. At times, we cognitively retrogress to reconstruct our intention to increase our confidence, satisfaction, or ?consonance? with our preceding behavior. Under such circumstances of reconstruction of intention, we need nonrecursive paths between behavior and intention. In Figure 34, I drew a nonrecursive path from behavior to intention to illustrate the point. Although Ajzen (1985) conceived of cases in which intention is absent for a certain behavior, it is questionable whether the theory of planned behavior can explain the reversal of temporal order between intention and behavior. As mentioned, the cognitive alpha and omega model suggests that people in some instances can rationally and successfully choose a backward cognitive approach in which they reach a conclusion 322 (e.g., behavior) before they find evidence (e.g., reason/intention) to back up that conclus ?s or two g refers ing refers to when people hold one belief and re ging and wedgin d how e a nce? th ion. With a serious problem, people hastily decide (behave) what to do and reconstruct reasons (behavioral intention) backwardly. The motivation to select the backward cognitive approach could be to explain an action to others (e.g., answering a doctor question about a risky personal behavior) or to justify and reinforce an a priori choice behavior. The reverse temporal order between intention and behavior highlights the conceptual limitations of some theories, such as the theory of reasoned action, that rely on the concept of intention to explain a behavior. CAOS reveals such a limitation in a few popular social psychological models using the concept of behavioral intention. Cognitive Multilateralism and Hedging and Wedging The studies on environmental issues, Stamm and J. Grunig (1977) and J. Grunig and Stamm (1979) found that people have situationally different cognitive strategies? ?hedging? and ?wedging.? They found that environmental publics often hold incompatible beliefs and frequently change their beliefs across situations. Hedgin to when people hold two conflicting beliefs; wedg ject other (J. Grunig & Hunt, 1984). J. Grunig (1997) used the hed g concepts in combination with the concept of ?attitude? to explain when an publics? attitudes differ from situation to situation. Hedging and wedging concepts ar useful alternative to illustrate how communication effects happen without the ?vale prediction of attitudes. I created a new concept, cognitive multilateralism, which is implied in hedging and wedging. Cognitive multilateralism in problem solving refers to cognitive bread 323 during a given problem solving task?i.e., the number of alternatives one generates and one?s tolerance of rival information during the problem-solving process. I speculated that a more entrepreneurial mindset in problem solving will have more cognitive breadth. To expand a viable candidate solution for a given problem requires ?tolerance.? Expansion of the candidate solutions inherently has a problem solver sustaining incompatible ideas and opinions. Thus, one?s situational need and willingness to withhold incompatibl and competing perspectives can e ideas be a good yardstick of the extent of an entrepreneurial problem l in applied communications. For nt of cognitive multilateralism may associate with the style of polit ld nitive multilateralism and cognitive entrepreneurship) rather than changing their problem solving approach. Cognitive multilateralism has much potentia instance, knowing voters? exte ical information processing and electoral decision making. A researcher cou study how voters with different level of cognitive multilateralism and cognitive entrepreneurship would vary in forefending (permitting), seeking (processing), and forwarding (sharing) election-related information. It also explains how effective a political candidate and supporters might be in managing conflicting social issues. In addition, assuming that cognitive multilateralism can better prepare one in health problem solving, a public health intervention program would aim at rather modest but realistic objectives, such as incrementally changing a risky group?s cognitive strategy (to have more cog atic behaviors immediately. Considering the difficulties in changing audience behaviors as a communication objective in health campaigns, enhancing cognitive entrepreneurship (e.g. reducing retrogression, increasing multilateralism) as an intervention objective must be a better alternative for health communicators. 324 CEPS and Characteristics of An Excellent Organization The IABC funded study, excellence in public relations and effective organizations (L. Grunig, J. Grunig, & Dozier, 2002; J. Grunig, 1992), found that a set of factors (e.g. roles and models of public relations preferred by CEO) affect the ?excellence? and ?effectiveness? of public relation and its hosting organizations. Treating cognitive entrepreneurship in problem solving as an enduring personal trait (e.g., the extent of cognitive retrogression), it is interesting to study how the problem-solving characteristics held by the CEO and members of the dominant coalition would affect , public relations excelle ation. re a ong organizational elites would have on the communication excelle ion structures) as a promising new research direction. In this vein, I proposed two new nce (e.g., the likelihood of giving public relations access to the decision-making process), types and quality of relationships with key stakeholders and publics (e.g., the likelihood of holding a symmetrical worldview), and the effectiveness of the organiz The IABC study found that the values of CEOs and dominant coalition members we critical factor for excellence in public relations and organization effectiveness. In this vein, it is worth investigating what kind of influence the extent of cognitive entrepreneurship am nce and the organizational effectiveness. Implications of Situational Theory of Problem Solving (STOPS) Researchers have tried to expand the situational theory by introducing new dependent variables such as ?cognitive response? (Slater, Chipman, Auld, Keefe, & Kendall, 1992) and ?message retention,? pro- or anti- ?cognition,? ?attitude,? and ?behavior, (Major, 1993; J. Grunig, 1982). J. Grunig (1997) posited a need for ?extens of the theory to new outcomes of communication? (e.g., ?breadth and depth? of cognitive 325 concepts of CAPS and CEPS and introduced them as dependent variables. In a nutshell, the new concepts are more general in describing the unique communicative and cognitive process n, cursor, has become a crucial compon ; ions (L. use , ners eight types of publics: open-dormant passive public, closed-dormant passive that a problem solver takes during a problematic situation. I tested how well the situational antecedent variables account for the two new communicative and cognitive variables. Findings suggest that the situational theory of problem solving explains whe why, and how we communicate and how and why we take a differential cognitive approach in problem solving. In the next section, I will discuss implications of the new situational theory to the theory and practice in the field of communication. STOPS and Public Relations The situational theory of publics, STOPS? pre ent of a general theory of public relations (L. Grunig, J. Grunig, & Dozier, 2002 J. Grunig, 1992). It provides a conceptual ground for strategic management of public relations (J. Grunig & Repper, 1992) and the two-way models of public relat Grunig, J. Grunig, & Dozier, 2002; J. Grunig & L. Grunig, 1992). In this vein, the situational theory should be considered as an essential theoretical foundation for the field of public relations. For that reason, I consider the influences and implications of this situational theory of problem solving to be the same as those of the situational theory of publics. First, an immediate implication of STOPS relates to segmenting publics. Beca of CAPS and CEPS, practitioners can classify publics in more useful ways. For example using the new typology of publics that I derived from the CAPS dimensions, practitio can identify 326 public, c, and closed- d of problem-solving potential there is for the lving tude, oners grams, public relations practitioners can apply the probabilities of informa e n open-situational active public, closed-situational active public, open-situational activist public, closed-situational activist public, open-chronic activist publi chronic activist public. Then, practitioners can predict when the information about the given problem or issue would be sought or just processed, forwarded and/or shared, and forefended or permitted. Furthermore, after practitioners identify the profiles of the publics in a problem, they can anticipate what kin issue?i.e., the symmetrical problem-solving potential, collective problem-so potential among publics, and stagnant potential for a given problem. Related to segmentation, J. Grunig (1982) studied the probabilities of a public?s information seeking and processing and other useful outcomes such as cognition, atti and behavior. The probabilities of communication behavior, cognition, attitude, and behavior guide public relations practitioners to prioritize publics in relation to their monetary and time budget constraints. Such knowledge of probabilities help practiti make a more strategic decision in implementing communication programs with the identified publics. In a similar way, the situational theory of problem solving can generate a probability table for the subdimensions in CAPS and CEPS. Thus, in preparing communication pro tion forwarding, information sharing, information forefending, information permitting, cognitive retrogression, cognitive multilateralism, cognitive commitment, and cognitive suspension. These probability estimates can be used in the ?expected-valu analysis? (J. Grunig & Hunt, 1984) that is a critical guideline in budgeting and decisio 327 making for public relations programs. STOPS can improve public relations pract formative and evaluative research in practice. Second, public relations researchers have proposed several typologies of publics. For example, Chay-Nemeth (2001) offered ?circumscribed,? ?co-opted,? ?critical,? an ?circumventing? publics. Hallahan (2001) outlined ?active (high knowledge and high involvement),? ?aroused (high involvement and low knowledge),? ?aware (high knowledge and low involvement),? and ?inactive (low knowledge and low involvement).? Contrary to the common misconception of ?general public,? J. Grunig Hunt (19 itioners? d and 84) classified publics as ?active,? ?aware,? ?latent,? and ?nonpublic.? Each type is conn enting, describ h onships ected with eight types of publics segmented by three independent variables of STP (e.g., high problem facing, low fatalistic behavior). Following its mother theory, STP, STOPS generates a new typology of publics that hinges on the dimensions of communication behavior. It is not intended to compete with existing typologies, but to complement them by emphasizing different communicative features among publics. It brings much conceptual utility in segm ing, and predicting publics? actions. Also, the new typology invites more researc about publics. For example, researchers can investigate extent and kind of relati that each type of public would form with organizations and what kind of behaviors each public would engage in (e.g., joining activist group, contacting organization/government official, changing one?s behavior). Also, with a finer distinction between active and activist publics (e.g., closed-chronic activist public vs. open situational activist public), researchers who study social activism and conflict resolution can form new research questions. 328 Third, as discussed earlier, with the inclusion of CAPS, the situational theory n describes in more detail how a group of publics would arise and the role of communication in the emergence of publics. In short, the information forwarding and forefending of focal communicants increase cross-awareness among isolated individu problem solvers. The cross-awareness about problem solving turns isolated ind ow al ividual problem a t PS ts s nce lematic situations are created, raised, and sustained g and reconfiguring) of messages? solvers into collective problem solvers. Provoking other communicants to recognize a problematic state, individual problem solvers can enhance the pool of potentially useful information, divide the costs for problem solving, and increase their bargaining power in demanding resources from relevant party. Information forwarding and permitting is at the heart of locating and networking with other individual problem solvers. Giving information about a problematic situation is thus a necessary condition for a problem to produce a group of collective problem solvers (e.g., an activist group). Thus, knowing who is likely to transmit information (i.e., problem and solution) to other communicants explains wha kind of problem would have more potential for resulting in a collective action. CA describes such processes with new dimensions of communication behaviors. It highligh the ?intercommunication? process (see Figure 33 in this chapter) among communicant (problem solvers) Such understanding of the intercommunication process highlights the role of communication behaviors in the emergence of public and the ?nature, role, and influe of communication? in the social contexts (Vasquez & Taylor, 2001, p. 150). Thus, STOPS better explains how ?prob through the symbolic convergence (configurin 329 (Vasqu e roposed of support? should be added to the situational theory. I second J. Grunig?s (1997) 9). tive em ? or al in ez & Taylor, 2001, p. 150). With CAPS, in which publics are defined as ?communicants? who work on problem solving, the new situational theory explains intercommunication better. Publics are conceived not only as the information takers, but also information givers and selectors who create perceptions around a problem, rais issues, and sustain their problem-solving efforts via symbolic interactions. Now the situational theory responds directly to the request of ?communication process and dynamics? (Vasquez & Taylor, 2001) and ?a framework that shifts the locus of analysis to the public?s communicative practices in interactional settings? (Cozier & Witmer, 2001). Fourth, the new situational theory addresses the issue of valence prediction with the earlier situational theory. Cameron and Yang (1990) and Slater et al. (1992) p that ?valence rejection of such request for ?philosophical? and ?pragmatic? reasons (pp. 38-3 Interestingly, a redefined referent criterion and new dimensions such as information forefending ensures advancement for this issue. Although I opposed adding ?valence? to the theory, it is possible and useful to know when and why a problem solver becomes more selective in dealing with information. I explained that problem solvers become selective because of the presence of a strong referent criterion. I defined referent criterion as ?any knowledge or subjec judgmental system that exerts specific influence on the way one approaches probl solving??this includes any ?decisional guidelines or decision rules perceived as relevant to a given problem.? In other words, a referent criterion can be either ?objective ?subjective.? However, both types of decisional referents are functionally identic 330 problem solving. Selectivity is divorced from the ?valence? of one?s beliefs. In other words, ew situational theory has resolved an issue of multicollinearity among s. In wnie, and De Stefano (2005) studied the source of d the ir f em ng in the em of multico predicting when a person becomes ?selective? is different from predicting ?valence? of support. Although I oppose the prediction of valence with the situational theory, I alternatively suggest predicting the communicant?s selectivity. STOPS now opens a way to explain and predict such recurrent selectivity among publics. Finally, the n independent variables (Kim, Downie, & De Stefano, 2005; J. Grunig, 1997). J. Grunig (1997) posed a question about multicollinearity among independent variable one study using the situational theory of publics, J. Grunig and Childers [a.k.a. Hon] (1988) found a few standardized path coefficients greater than 1.00. Such unusual standardized coefficients often result from the multicollinearity problem among independent variables (J?reskog, 1999). Kim, Do of multicollinearity particularly from the conceptual overlapping between level involvement and problem recognition. Kim et al. (2005) conceptually explicate definition of problem recognition as more of ?detect something is missing? in the situation and isolated the ?stop to think? tendency as the situational motivation. The empirical analysis reduced multicollinearity with the refined conceptual definition o problem recognition. In the present study, I redefined and used measures of probl recognition to be more about a perceived problem (i.e., detect something is missi situation). Analyses showed that situational theory now has little, if any, probl llinearity. Thus, I conclude the refined conceptual definition can be recommended in future research and practice. STOPS and Health and Risk Communication 331 Researchers have applied STP to enhance the effectiveness of health and ris communication (e.g., Aldoory, 2001; Chay-Nemeth, 2001). Similarly, STOPS with CAPS (i.e., SITCAPS) helps risk and health communicators better identify critical agents for health/risk information diffusion. Focal communicants who are active problem solvers tend to disperse problems and solutions proactively. They are sophisticated information takers and givers who regularly and selectively upd k ate information related to ts tend to be in an effectuating stage, they are likely t a first ely. For if a ation previously drawn conclusions. Tests with the situational anteced ive the problem. Because focal communican o forward information. Further, focal communicants, even reactively, act as stop referent (cf. opinion leader) when neighboring communicants are identifying a problem and entering into the inquiring stage. STOPS can prescribe and trigger better information trafficking regarding health problems among target segments of the population. For health care experts frustrated with groups that continue to engage in risky behaviors, STOPS with CEPS (i.e., SITCEPS) explains how and why some chronic problem holders fail to use information adequately and fail to behave appropriat example, the theory predicts that most information dissemination efforts will be futile risky behavior group possesses a strong cognitive omega approach. As discussed earlier, a cognitive omega strategy in problem solving can worsen one?s problematic situ when one refuses to revise ent variables indicated that as one has higher problem recognition, lower constraint recognition, higher involvement, and higher referent criterion, one?s cognit entrepreneurship decreases. SITCEPS thus requests differential intervention in accordance with the level of entrepreneurial mindset in one?s problem solving. 332 STOPS and Political Communication Communication effects studies originated from the early studies about voters and election (Katz & Lazarsfeld, 1955; Lazarsfeld, Berelson, & Gaudet, 1944). The main interest was to understand how voters were influenced and influenced other voters. In essence, such studies inquired how political information was dispersed and us what effectiveness. First, STOPS with CEPS explain how voters and citizens make political dec and actions in some situations. The variables cognitive retrogression and cognitive ed with isions multila re more f a nd th ion and behavioral changes are the prime ?communication objectives.? However, researchers have found that these objectives are rarely met. J. teralism are especially interesting in this context. For example, a voter with mo cognitive multilateralism and less cognitive retrogression would make a decision slowly, would consider more aspects before decision making, and would have more o tendency to cross-vote than to vote a party line. Next, STOPS with CAPS fits well with the ?two-step flow model? of personal and mass media influence. As discussed in opinion leadership and meso-level intercommunication models, communication by opinion leaders in political issues a elections is a special case of focal communicants (high in CAPS). They sometimes reinforce and compete with other focal communicants such as mass media or rival candidate supporters. Thus, SITCAPS will provide a conceptual framework that illustrates how voters? information behavior will be different and how focal and peripheral communicants shape the electoral outcomes in the political arena. In summary, in most applied communications contexts (e.g., political and heal campaigns), attitude format 333 Grunig n s in and around ?doing science.? For example, philosophy provide s s? (Kuhn, how CAPS, CEPS, and STOPS fill t , CAPS conceptualizes unique features of a problem solver?s information selectivity, transmission, and acquisition. Problems can be any life problems that lay people experience routinely (e.g., a health problem, buying a car, or job hunting), whereas they can be scientific problems that and Hunt (1984) introduced alternative communication objectives such as ?coorientation? and ?symmetrical? communication approaches such as ?accuracy,? ?understanding,? and ?agreement.? With the new situational theory of problem solving, I also introduced a new set of communication objectives: e.g., increasing ?informatio permitting,? and ?information forwarding,? making less ?retrogression,? having more ?cognitive multilateralism,? and more ?cognitive suspension.? STOPS and Communication of Science Several disciplines contribute to a better understanding of the antecedents, processes, and consequence s essential understanding of how scientific epistemics differ from non-scientific lay thinking: i.e., philosophy of science (Popper, 1963). History of science elucidate how, why, and what factors drive scientific advancements or ?paradigm shift 1962). Sociology explains how scientific knowledge and technology are dispersed, scientists gather, connect, compete, and advance knowledge in terms of social connection (Kraut, Egido, & Galegher, 1990; Bijker, Hughes, & Pinch, 1987). Psychology investigates how scientific inference and method is different from and similar to the process of na?ve lay inference and lay epistemics (Kruglanski, 1989). However, there is little research on the ways in which lay and scientific thinkers differ in terms of communication behavior. his gap. First 334 researchers and scientists work to resolve professionally (e.g., curing cancer with stem cells, explaining the causes of the Great Depression, decoding particles consisting of d matter in the universe). Regardless of the type of problems, life or scientific mysteries, human activities around information play the key role in bringing solutions. Thus, studies should ask, ?How are lay and scientific epistemics different in terms of their communication behavior?? CAPS provides a way to distinguish the lay problem sol and scientific problem solvers with 1) the extent of their information permitting to different or competing ideas and 2) their distinct ?causes? of information forefending. Types of referent criterion, whether subjective beliefs (e.g., wishful or willfu thinking) or a more objective knowledge (e.g., carried solution from experts or experience of the successful problem solving in past), affect the types of information forefending, such as ?reinforcing selectivity? (e.g., avoidance in the diss ark vers l onance theory) or coming effective in distinguishing relevant information as a result o from ?revising selectivity? (e.g., be f extended problem solving). In reinforcing selectivity, the purpose is to enhance cognitive confidence in the preferred solution, whereas in revising selectivity, the purpose is to enhance ?efficiency? in problem solving stemming from information saturation. In addition, it is interesting to see whether problem solvers will continue to permit information even when it conflicts with or refutes one?s current beliefs in the problem-solving outcome. I reason that the scientific problem solver would be more permitting than the lay problem solver (cf. ?Scientific theories are distinguishable myths merely in being criticizable, and in being open to modifications in the light of criticism,? Popper, 1963). 335 CEPS will add knowledge to why and how a communicant?s information ity arises and how, in turn, selective communicant activeness would reduce the eneurial mindset in problem solving. Then, it is interesting to investigate selectiv entrepr whether ard reasoning) led one being equally capable in problem ing). In addi subdim how sc entrepreneurial mindset would hat there are differences between scientific and lay problem solvers in edent variables will explain what causes such differen r willful ss n permitting, less cognitive multilateralism, or more cognitive retrogression), lf-awareness about their problem-solving approach (e.g., tr for Validity of Findings In general, CAPS and its subdimensions had close to or more than 50% construct validity in terms of variance extracted. Construct reliabilities for CAPS and its six dimensions were also more than the minimum values of .70. However, CEPS (i.e., about cognitive retrogression (i.e., backw solving as a less retrogressive problem solver would be (i.e., forward reason tion, it is interesting to investigate whether cognitive entrepreneurship and its ensions can also be conceptualized as an enduring personal trait. If so, I can test ientific problem solving and the traits of an correlate with each other. Assuming t their approach, those situational antec ces and how we encourage positive aspects and discourage negative characteristics in approaching both scientific and lay problems. For example, with knowing what factors (e.g., subjective type of referent criterion such as wishful o thinking) causes an undesirable problem-solving approach for some problems (e.g., le informatio we can cultivate problem solvers? se ain children which cognitive and communicative approach one should take some special problems). Limitations 336 20%) and its two subdimensions (i.e., 30-40%) and low variance extracted. For cons reliability, two subdimensions of CE truct PS in two issues were close to .70 or higher. For ower H and construct validity seemed e CAPS case. With this notion in mind, it is n of the CEPS concept. Future research should rly tests of nascent theories? adopt non- probability student samples. Such a l, but they are useful for initial theoriz ). CEPS, itself, .612 and .627 which is lower than .70. The l to originate from the smaller pool of CEPS items. The sheer number of question items for three problems reached 100 questions and up to 300 responses. In order to minimize participant fatigue, I reduced the number of items from what I originally proposed. I set more priority on the CAPS items, with its six dimensions, and thus the number of items for the CEPS dimensions was smaller. As a result, it was difficult to identify good items, as I could do in th necessary to be more conservative in interpretatio address this limitation by using more items and a multiple-item approach for the cognitive multilateralism and the cognitive commitment dimensions. 25 Generalizability of Findings In the present study, I chose the convenient snowball sampling method. This is a non-probability sample that has great limitations if the purpose of the study is statistical generalization. For example, the However, the purpose of this study was to develop new variables and new theory. In practice, ?most ea strategy is not idea ing and hypothesis testing with ?multivariate relationships? (Caplan, 2005, p. 732 Besides, I delimit the scope to ?theoretical generalizability? rather than statistical generalizability. Calder, Philips, and Tybout (1981) pointed out: 25 However, the multiple-item approach can lead to a more participant fatigue. Thus, narrowing-down the research focus either to SITCAPS or SITCEPS, not both, will be necessary to reduce the number of questionnaire items. 337 [to make] theoretical generalizations, a representative sample is not required is applied beyond the research setting. The research sample need only allow a sample), not just a representative one, can provide such a test. (p. 200) In addition, Shapiro (2002) said: [the issue of] rejecting a study that seeks to expand theory and that detects a misunderstanding of the nature of generalizability. If a study detected important people, in some setting, at some time, for some message. The next step ma applies and to whom it does not. (p. 499) Thus, I delimit the generalizability of this study only as theoretical generalizability, no because statistical generalization of the finding is not the goal. It is the theory that test of the theory. And, any sample within the theory?s domain (e.g., any relevant potentially important effect on the basis of a nonrandom sample usually reflects a effect, no matter what kind of sample is used, it is clearly true for some group of y be to conduct a theoretically driven boundary search to determine to whom the effect further e Nonrec ed at ariable than that. Cross-validation with Single Sample Related to convenience sampling, in an ideal study cross-validation tests for th variables should be tested using different samples. In the present study, I used a sub- sampling strategy (randomly select subsamples from the total sample) for cross- validation tests. In future research, this limitation can be addressed by using multiple samples. ursive Model Identification It was interesting to test an integrated model that contained nonrecursive relationships between CAPS and CEPS while specifying situational antecedent variables as exogenous variables. However, such model is either hard to identify (just identifi best) or hard to converge. As a realistic alternative, I used the referent criterion v only for identifying the nonrecursive model. However, it was still an interesting model 338 that integrated all the variables into a single model. Future study should continue to attempt such an integrative model. Similarity between CAPS and CEPS Survey Instruments Although cognitive and communicative characteristics are conceptually distinct, the measurement items for CAPS and CEPS bear some similarities. Such similarity among rsive ). In ay depend on the problem chosen. For example, the structural path coeffic h survey instruments was unavoidable, but it is still a limitation from the survey instrument approach. Thus, it is necessary to be cautious interpreting the nonrecu model findings between CAPS and CEPS, although the nonrecursive models converged and reached to acceptable model fit. Demographic Variables and Problem Sensitivity The current study did not include demographic variables such as sex, age, and income. However, it would be interesting to examine whether there are some mediating effects from demographic variables (e.g., gender difference in situational perception addition, the theory m ients in SITCAPS could be different with other types of problems (e.g., Tuition Increase) for the current respondent group. Future study should test whether there is suc problem sensitivity using different problems. Future Research Findings from the current study generate interesting new studies. First, CAPS is an apt model whereby I can test reversal effects from communication behaviors to situational antecedent variables. The CAPS model captures information transmission from the focal communicant to other peripheral communicants in their communicant networks. It would be interesting to test how the activeness of focal communicant?s 339 communicant activeness in problem solving would influence perceptual gaps between focal and peripheral communicants (e.g., problem recognition gap). In addition, it is interesting how peripheral communicant?s own communicant activeness would affect percept CAPS and olving potential, and stagnant problem-solving potential. The relationship betwee n or as nal budgeting and decision making of public relations programs. It helps to predict how a ual gaps regarding a problem. Finally, I can test the effect of communication on ?accuracy,? ?understanding,? and ?agreement? in the coorientation model (McLeod & Chaffee, 1973). This is a way of empirically testing the coorientation model within the situational theory framework. Also, this study answers Hallahan?s (2001) call for research on using communication as an independent variable. Snowball sampling can provide a way to test such a hypothesis using the clusters within a snowballed sample. Second, I conceptualized relationships between the six dimensions in problem-solving potential in socially conflicting issues. In doing so, I proposed three types of problem-solving potential symmetrical problem-solving potential, collective problem-s n a public?s communication behavior and issue resolution potential has rarely bee studied. The study of CAPS and problem-solving potential in conflicting social issues will contribute to a body of knowledge in public relations as well as conflict resolution. Third, J. Grunig (1982) studied how the situational variables explain the occurrence of cognitive, attitudinal, and behavioral effects of communication behavi well as the occurrence of communication behavior. This study expanded the situatio theory of publics not only theoretically but also practically. In his study, J. Grunig presented a table of probabilities of communication behaviors and effects of 16 behavioral situations. As discussed in the implication section, this table is useful in 340 given communication program would pay off by expected-value analysis. Thus, it wou be interesting to revisit ld what J. Grunig did in his 1982 study with CAPS, CEPS, and STOPS practice as ound rograms (J. Grunig, 1975, 1977, 1978, 1979, 1982, 1983). In addition to continu o anizational n and curring types o I bles, I hose types of publics (e.g., all-issue publics, hot-issue publics tc). If replicated, I can study how those publics differ and how each type of public would differ in the six dimension of CAPS and four dimensions of CEPS. For example, we may (e.g., probabilities of information forwarding, sharing, permitting, and cognitive multilateralism). The resulting probability tables can enhance public relations the 1982 study has done. Fourth, J. Grunig has developed a methodology to identify publics formed ar particular issues. The method is useful for public relations practitioners to target communication p ous testing and refinement of the new situational theory, it will be necessary t devise specific steps of how public relations practitioners can segment publics with the new situational theory. Replication of the STOPS with different problems, org settings, and societies should be done. Those accumulated findings from validatio replication studies will provide examples for practitioners to apply in actual public segmentation tasks. Fifth, studies using the situational theory of publics have found four re f publics. They are all-issue publics, apathetic publics, single-issue publics, and hot issue publics. J. Grunig (1997) noted that these types of publics seem to have theoretical regularity, which he found repeatedly from canonical correlation analysis. consider a canonical correlation study with different issues and with the new situational theory interesting. With CAPS and CEPS and the newly revised antecedent varia may or may not replicate t e 341 inquire how single-issue publics and hot-issue publics differ in terms of their information forefending and permitting and information forwarding and sharing. The findings will result in better understanding of different kind of publics that are critical in practice. Sixth, I deployed survey methodology for testing CAPS, CEPS, and STOPS in owe Specifically, a validation study setting experimental manipulation of antecedent variables will test how and PS and CEPS mental des and the extent of rmation permitting, cognitive multilateralism a pension as dependent variables can help clarify causal interpretations be qualitative interv exploratory study for CEPS and CAPS. I conducted a SEM analysis for a non helpful in unders ts. However, ative study develop rich desc d communicative behaviors are interconnected. Future study with focus groups, elite interviews, and in-depth interviews this study. H ver, I found a need for validation studies with experimental design. to what extent CAPS and CEPS would differ. In addition, I believe CA constructs are useful and promising on their own. For example, an experi ign that manipulates cognitive retrogression as an independent variable information forefending, info nd cognitive sus tween cognitive strategies and communicative behavior in problem solving. Seventh, in the developmental stage of the CAPS and CEPS, I conducted a few iews to understand communicants? characteristics in social issues. Such lead me to propose the propositions in CAPS and CEPS. However, I found a need conducting qualitative study on recursive relationship between CAPS and CEPS. The finding was tanding the interrelationship between the two concep quantit using survey methods is quite limiting as a way to test in-depth and to ription about how cognitive features an 342 with different types of publics (e.g., open-situational activist public) would provide a better understanding. the present study. However, I am interested in whether these concepts can be ore enduring cognitive and communicative styles in dealing with Ninth, assuming CEPS as a personal trait, it is interesting to test whether the solving would affect the styles in ethical tionships between CEPS and ethical decision-making style will enhance our ess management. Further, it akers to be -making style and problem-solving approach. Finally, I asked what kind of relationship exists between the extent of cognitive Grunig, & Dozier, 2002). Such study as a continuous theory-building for strategic management of public relations will improve our anagement excellence. Eighth, CAPS and CEPS have been conceptualized as ?situational? constructs in conceptualized as enduring personal traits. From informal studies, I found that people tend to develop m problems. Hence, in future study, I will reconceptualize and test CAPS and CEPS as personal traits. tendency of entrepreneurial mindset in problem decision making (e.g., teleological or deontological approach). Testing of the rela understanding regarding ethics in public relations and busin will result in some normative knowledge, such as how we train decision m aware of their ethical decision entrepreneurship of dominant coalition members and excellence in public relations and organizational effectiveness (L. Grunig, J. understanding of what factors explain and affect communication excellence and m 343 Appendix A INFORMED CONSENT FORM Identification of Project/Title Situational Theory of Publics Statement of Age of Participant (parental consent needed for minors) I state that I am 18 years of age or older and wish to participate in a program of research being conducted by Jeong-Nam Kim and Dr. James E. Grunig in the Department of Communication at the University of Maryland, College Park, MD 20742-7635. Purpose The purpose of the research is to test the influence of personal traits on perception of issues and communication behavior in the United States and Korea. Procedures The procedures involve answering a number of questions. I understand my participation will require approximately 20 minutes. Confidentiality All information collected in the study is confidential. The data I provide will not b e linked to my name and, furthermore, will be grouped with data others provide for reporting and presentation. Risks I understand that there are no foreseeable personal risks associated with my participation. Benefits I understand that the study is not designed to help me personally, but that the investigator hopes to learn more about communication behaviors in the United States and Korea. Freedom to Withdraw, & Ability to Ask Questions I understand that this participation is voluntary and free to ask questions whenever I feel like to do. I understand that I can withdraw from participation at any time without penalty and/or decline to answer certain questions. I understand that alternative assignments are available if I decide not to participate and it is possible to earn equal extra credit I would get from participation of this study. Finally. I understand that any record of my participation will be destroyed, if I withdrawal from the study. Principal Investigator Dr. James E. Grunig Department of Communication 2118 Skinner Building University of Maryland, College Park, MD 20742-7635 Phone: (301) 405-6525; Email: jgrunig@umd.edu Obtaining a copy of the research results I understand that I may obtain a copy of the results of this research after July 2005 by contacting Jeong-Nam Kim, at the Department of Communication, 2118 Skinner Building, University of Maryland, College Park, MD 20742-7635. Phone: (301) 405-6533; E-mail: jnkim@umd.edu.. << Disagree >> << Agree >> Powered by CreateSurvey.com 344 Please Read Each Statement and Pick a Number that Best Reflect Your Feeling, Thoughts, and Opinion Regarding Three Problems. Please enter the last four digits of SSN and the first letter of your last name ("S6789", "D3212", ...) 1. In your mind, how much of a connection do you see between yourself and this problem? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 2. To what extent do you think there is something missing in this problem? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 3. I continue to think about the pros and cons of possible solutions regarding this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 4. How much does the current situation differ from your expectation? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 5. I have found enough support for the position I take in this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 6. How strong do you feel that something needs to be done to improve the situation for this problem? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 7. From time to time, I contact people about this problem to learn what kind of solutions there are. 345 Please rate War in Iraq Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 8. I hesitate to make up my mind about what should be done for this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 9. I have made efforts to justify my decision on this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 10. To what extent do you believe this problem could involve you or someone close to you at some point? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 11. I want to take more time before making up my mind for this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 12. I know what people around me think about this problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 13. I have found much evidence that reinforces my decision regarding this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 14. Sometimes I find I am engaging in aggressive conversations on this problem. War in Iraq Please rate 346 Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 15. Regarding this problem, I regularly meet and chat with likeminded people whose views about the problem are similar to my own. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 16. It is one of my top priorities to share my knowledge and perspective about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 17. If it is possible, I take time to explain this problem to others. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 18. How much does the current situation deviate from what you think it should be? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 19. Others respect my perspective about this problem because it is simple and clear. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 20. To what extent would you say you are curious about this problem? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 21. I am frustrated because there is too much information available about this problem. War in Iraq Please rate Losing Weight Please rate 347 firmative Action in American Higher Education Please rate Elimination of Af 22. There are many misleading but widely accepted opinions about this problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 23. To make better decisions regarding this problem, I listen to opposite views and information as long as they are related to the problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 24. Some publicized statements about this problem are worthless. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 25. I have spent too much time on this problem to change my position now. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 26. I have invested enough time and energy so that I understand this problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 27. I know where to go when I need updated information regarding this problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 28. I have studied this problem enough to judge the value of information. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 348 29. I have found counter evidence that rejects the positions different from mine. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 30. I am confused with what is going on when I hear something about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 31. To what extent would you say that this problem is more difficult for you to understand than other problems? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 32. I am confident about my knowledge about this problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 33. I paid attention to a news report about the problem recently. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 34. It is too late to change the position I now have on this problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 35. It is worthy spending some time to persuade others about this problem. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 36. I feel like resisting some persuasive efforts around this problem. War in Iraq Please rate 349 Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 37. I will keep my current position even if someone challenges it with contradictory evidence. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 38. Please, indicate how much you would like to understand each of these problems better. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 39. I express my opinions confidently about what should be done to deal with this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 40. I feel it is costly to change my mind on this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 41. How much do you believe this problem affects or could affect you personally? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 42. Although I am knowledgeable about this problem, I rarely speak up. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 43. I look for chances to share my knowledge and thoughts about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 350 44. I actively seek out opportunities to participate in public opinion polls about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 45. For this problem, I welcome any information regardless of where it comes from. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 46. It is important to learn the latest information around this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 47. To what extent do you believe this problem is a serious national or social problem? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 48. I am shy in expressing my opinions publicly about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 49. I strongly support a certain way of resolving this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 50. I have a preference for how the problem should be settled. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 51. I am careful in accepting information about this problem because of the vested interests of those who provided the information. 351 War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 52. I love to start a conversation on this problem with others. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 53. How strong would you say your opinions or thoughts are about this problem? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 54. I am sure that I will be quite active in passing on information related to this problem in the near future. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 55. I know how I should behave for this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 56. I regularly visit websites relevant to the problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 57. I am a person to whom my friends and others come to learn more about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 58. I volunteer to inform others about the problem. War in Iraq Please rate 352 Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 59. To what extent do you think this problem is too complicated for you to do anything about? War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 60. Listening to an opponent's view about this problem is a waste of time. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 61. It is too late to shake the conclusion I have drawn for this problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 62. I regularly check to see if there is any new information about this problem on the Internet. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 63. I would request booklets containing relevant knowledge about the problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 64. Please think of whether you, personally, could do anything that would make a difference in the way these problems are handled. If you wanted to do something, would your efforts make a difference? War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 65. I believe there is no need to buy or read books or brochures about this problem. War in Iraq Please rate Losing Weight Please rate 353 Please rate Elimination of Affirmative Action in American Higher Education 66. I feel happy when I provide new information about this problem to others. War in Iraq Please rate Losing Weight Please rate Please rate Elimination of Affirmative Action in American Higher Education 67. I pay attention to the problem when a news report appears on TV news. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 68. I visit an online or regular bookstore to find useful information about the problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 69. My friends think that I take too much time for learning about this problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 70. I feel like I am suffering from information overload about this problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 71. To what extent do you believe this problem is a problem that you can do something about? War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 72. I often play a leadership role in initiating conversation about the problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 73. In the past, I researched about this problem seriously. 354 War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 74. I am pretty sure, I know how to solve this problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 75. I don't want waste my time trying to persuade others about this problem. War in Iraq Please rate Losing Weight Please rate firmative Action in American Higher Education Please rate Elimination of Af 76. Past experience has provided me with guidelines for solving this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 77. I may take some time listening if someone tries to give information about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 78. I listen to even opposite views on this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 79. At times, I find that I have accepted conflicting information about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 80. At times, I am asked to give advice regarding this problem. War in Iraq Please rate Losing Weight Please rate 355 Elimination of Affirmative Action in American Higher Education Please rate 81. How often do you stop to think about each of these three problems? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 82. To what extent do you believe that you could affect the way this problem is eventually solved if you wanted to? War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 83. People around me know clearly what I think about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 84. I am willing to write a letter, email, or fax to express my concern about this problem to a relevant organization. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 85. I hesitate to share my knowledge about this problem with others. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 86. I have never participated in a public opinion poll related to this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 87. I don't want share my ideas and opinions with other people regarding this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 356 88. I am picky in choosing information sources when I think about this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 89. I need more time to think before I finalize my position on this problem. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate 90. For this problem, I will try to suspend any judgment until all the evidence is in. War in Iraq Please rate Losing Weight Please rate Elimination of Affirmative Action in American Higher Education Please rate Please read the statements below and pick a number that best reflects the extent to which you agree with the statement. 91. The war in Iraq can be justified because the cost of controlling Saddam Hussein was in while he power was higher than that of war. Please rate 92. With the economic and domestic security problems the United States was facing, it was a bad time to go to war in Iraq. Please rate 93. A pre-emptive attack by the U.S. gives credibility to those who describe the U. S. as an aggressive nation. Please rate 94. The war in Iraq has increased anti-American sentiment. Please rate 95. Saddam was connected with terrorists. Please rate 96. Saddam's human rights record, among the worst in the world, was enough justification to go to war. Please rate 97. Affirmative action demeans true minority achievement because success is labeled as a result of 357 affirmative action rather than hard work or ability. Please rate 98. Affirmative action levels the playing field because minority students, generally speaking, start out at a disadvantage. Please rate 99. Diversity is desirable yet won't always occur if left to chance. Please rate 100. Because of affirmative action, a wealthy minority student who doesn't put in much effort could be chosen over a poor white student who works harder. Please rate Thank you! You finish the questionnaire! Submit Powered by CreateSurvey.com 358 References E. A., (1952). Flatland. A romance of many dimensions. New YAbbot, ork: Dover. uhl & J. 11-39). New York: Springer-Verlag. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall. Ajzen, I., Timko, C., & White, J. B. (1982). Self-monitoring and the attitude-behavior relation. Journal of Personality and Social Psychology, 42, 426-435. t ublic Relations Research, 13, 163-185. Bab Ban ry of behavioral change. Psychological Review, 84, 191-215. Bargh, J. A. (1996). Automaticity in social psychology. In E. T. Higgins & A. W. -183). Afifi, W. A., & Weiner, J. L. (2002). Introduction: Information-seeking across contexts. Human Communication Research, 28, 207-212. Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. K Beckmann (Eds.), Action control: From cognition to behavior (pp. Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52, 317-322. Aldoory, L. (2001). Making health communications meaningful for women: Factors tha influence involvement. Journal of P Ashcraft, M. H. (1994). Human memory and cognition. New York: Harper Collins. bie, E. (1998). The practice of social research (8 th ed.). Bermont, CA: Wadsworth. dura, A. (1977). Self-efficacy: Toward a unifying theo Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 169 New York: Guilford. 359 Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical consider Journal of Personality and Social Psychology, 51, 1 ations. 173-1182. Beniger, J. R. (1986). The control revolution. Cambridge, MA: Harvard University Press. Bentler, P. M. (2005). EQS 6 structural equations program manual. Encino, CA: Multivariate Software. al T Press. Biocca, F. A. (1988). The breakdown of the canonical audience. In J. Anderson (Ed.), Communication Yearbook 11 (pp. 127-132). Newbury Park, CA: Sage. ed. pp. 43-50). New York: Free Press. Bobrow, D. G., & Norman, D. A. (1975). Some principles of memory schemata. In G. B. Bobrow & A. Collins (Eds.), Representation and understanding: Studies in cognitive science (pp. 131-149). New York: Academic Press. Bogdan, R., & Biklen, S. (1998). Qualitative research in education: An introduction to theories and methods. Needham Heights, MA: Allyn & Bacon. r: , 10, 113-117. Bollen ons with latent variables. New York: John Wiley & Sons. Bijker, W. E., Hughes, T. P., & Pinch, T. (1987). The social construction of technologic systems: New directions in the sociology and history of technology. Boston, MA: MI Blumer, H. (1966). The mass, the public, and public opinion. In B. Berelson & M. Janowitz (Eds.), Reader in public opinion and communication (2 nd Bohner, G., & Siebler, F. (1999). Paradigms, processes, parsimony, and predictive powe Arguments for a generic dual-process model. Psychological Inquiry , K. A. (1989). Structural equati 360 Booth, A., & Babchuck, N. (1972). Informal medical opinion leadership among the middle aged and elderly. Public Opinion Quarterly, 36, 87-94. of Brock, ormation Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Brownstein, A. L. (2003). Biased predecision processing. Psychological Bulletin, 129, 545-568. Caciop ersonality and Social Psychology, 42, 116-131. Cacioppo, J. T., Petty, R. E., Feinstein, J. A., & Jarvis, W. B. G. (1996). Dispositional differences in cognitive motivation: The life and times of individuals varying in need Cald W., & Tybout, A. M. (1981). Designing research for application. Journal of Consumer Research, 8, 197-207. Caplan, S. E. (2005). A social skill account for problematic Internet use. Journal of Communication, 55, 721-736. Carter, R. F. (1965). Communication and affective relation. Journalism Quarterly, 42, Brehm, J. W. (1956). Postdecision changes in the desirability of alternatives. Journal Abnormal and Social Psychology, 52, 384-389. T. C. (1965). Commitment to exposure as a determination of inf receptivity. Journal of Personality and Social Psychology, 2, 10-19. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136-132). Newbury Park, CA: Sage. Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123-152. po, J. T., & Petty, R. E. (1982). The need for cognition. Journal of P for cognition. Psychological Bulletin, 119, 197-253. er, B. J., Phillips, L. 203-212. 361 Carter, R. F. (1966, December). On defining communication. Paper presented at a session on ?Human Communication Theory Building,? Sponsored by the National Carter, R. F. (1972, September). A general system characteristic of systems in general. Carter, R. F. (1973, July). Communication as behavior. Paper presented to the Theory and Methodology Division, Association for Education in Journalism, Fort Collins, CO. Carter, R. F., Pyszka, R. H., & Guerrero, J. L. (1969). Dissonance and exposure to Case king for information: A survey of research on information Cha Chaffee S. H. (Ed.), Cha channels: Competitive, convergent, personal communication in a media world (2 nd ed.) (pp. 57-77). New York: Oxford University Press. Chaffee, S. H., Saphir, M. N., Graf, J., Sandvig, C., & Hahn, K. S. (2001). Attention to counter-attitudinal messages in a state election campaign. Political Society for the Study of Communication, Chicago. Paper presented to the Far West Region Meeting, Society for General Systems Research, Portland, OR. aversive information. Journalism Quarterly, 46, 37-42. , D. O. (2002). Loo seeking, needs, and behavior. San Diego, CA: Academic Press. ffee, S. H. (1977). The diffusion of political information. In Political communication (pp. 85-128). Beverly Hills, CA: Sage. ffee, S. H. (1982). Mass media and interpersonal or complementary? In G. Gumpert, & R. Cathcart (Eds.), Inter/Media: Inter Communication, 18, 247-272. 362 Chaffee, S. H., Stamm, G., Guerrero J. L., & Tipton, L. P. (1969). Experiments on cognitive discrepancies and communication. Journalism Monographs, 14. Chaiken, S., & Eagly, A. H. (1989) Heuristic and systematic information processing . New York: Guilford Press. Cha ). Revisiting publics: A critical archaeology of publics in the Thai Chwe, University Press. Some new strategies Colema ocial organizations with survey Dawes orld. San Diego, CA: Harcourt Dennet onsciousness explained. Boston, MA: Little, Brown & Co. Dewey . Dewey, J. (1929). The quest for certainty. New York: Minton, Balch. within and beyond the persuasion context. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 212-252) Chaiken, S., & Eagly, A. H. (1993). The psychology of attitudes. Orlando: Harcourt Brace College Publishers. y-Nemeth, C. (2001 HIV/AIDS issue. Journal of Public Relations Research, 13, 127-161. M. S. (2001). Rational ritual: Culture, coordination, and common knowledge. Princeton, NJ: Princeton Clarke, P., & Kline, F. G. (1974). Mass media effects reconsidered: for communication research. Communication Research, 1, 224-270. n, J. S. (1975). Relational analysis: The study of s methods. In G. H. Lewis (Ed.), Fist-fights in the kitchen: Manners and methods in social research (pp. 97-107). Pacific Palisades, CA: Goodyear. , R. M. (1988). Rational choice in an uncertain w Brace Jovanovich. t, D. C., (1991). C DeVellis, R. F. (1991). Scale development: Theory and applications. Newbury Park, CA: Sage. , J. (1927). The public and its problems. Chicago: Swallow 363 Dewey, J. (1938). Logic: The theory of inquiry. New York: Holt. . C. (1969). Matrix algebra: A programmed iDorf, R ntroduction. New York: Wiley. . Su Ps dis Feather al of Festing sity Pr Festing nance. Stanford, CA: Stanford Fink, A Fishbei gs in attitude theory and measurement (pp. 389-400). New York: Wiley. Eagley, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcourt Brace. Earle, T. C., & Cvetkovich, G. T. (1995). Social trust: Toward a cosmopolitan society Westport, CT: Praeger. Evans, J. St. B. T. (1989). Bias in human reasoning: Causes and consequences. East ssex, U. K.: Erlbaum. Evans, J. St. B. T. (2002). Logic and human reasoning: An assessment of deduction paradigm. Psychological Bulletin, 128, 978-996. Evans, J. St. B. T., & Over, D. E. (1996). Rationality and reasoning. East Sussex, U. K.: ychology Press. Feather, N. T. (1962). Cigarette smoking and lung cancer: A study of cognitive sonance. Australian Journal of Psychology, 14, 55-64. , N. T. (1963). Cognitive dissonance, sensitivity, and evaluation. Journ Abnormal and Social Psychology, 66, 157-163. er, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford Univer ess. er, L. (1964). Conflict, decision, and disso University Press. . (2003). The survey handbook (2 nd ed.).Thousand Oaks, CA: Sage. n, M. (1967). Attitude and the prediction of behavior. In M. Fishbein (Ed.), Readin 364 Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley. Freedm an Freedm ive exposure. In L. Berkowitz (Ed.), rk: Ac Frey, D. (1986). Recent research on selective exposure to information. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp. 41-80). New York: Academic Press. Gigeren Gollwi n ns & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic Grunig ism Monographs, 3. Grunig, J. E. (1968). Information, entrepreneurship, and economic development: A study d Fowler, F. J. (2002). Survey research methods (3 rd ed.). Thousand Oaks, CA: Sage. an, J. L. (1965). Preference for dissonance information. Journal of Personality d Social Psychology, 2, 287-289. an, J. L., & Sears, D. O. (1965). Select Advances in experimental social psychology (Vol. 2, pp. 57-97). New Yo ademic Press. George, F. H. (1980). Problem solving. London, U. K.: Duckworth zer, G. (1991). From tools to theories: A heuristic of discovery in cognitive psychology. Psychological Review, 98, 254-267. tzer, P. M., & Moskowitz, G. B. (1996). Goal effects on action and cognition. I E. T. Higgi principles (pp. 361-399). New York: Guilford. , J. E. (1966). The role of information in economic decision making. Journal of the decision making processes of Colombian Latifundistas. Unpublishe doctoral dissertation, University of Wisconsin Madison. 365 Grunig, J. E. (1969). The minifundio problem in Colombia: Development alternatives. Inter-American Economic Affairs, 23, 2-23. , J. E. (1971). Communication and the economic decision making processes of Colombian peasants. Economic Development and Cult Grunig ural Change, 19, 580-597. Grunig, J. E. (1976). Communication behaviors in decision and nondecision situations. Grunig search, 9, 163-200. y Park, CA: Sage. Ver?i? (Eds.), U. K.: Grunig and practice. In B. Dervin & S. ton Press. Grunig, J. E. (1972). Communication in community decisions on the problems of the poor. Journal of Communication, 22, 5-25. Journalism Quarterly, 53, 232-263. , J. E. (1982). The message-attitude-behavioral relationship: Communication behaviors of organizations. Communication Re Grunig, J. E. (1983). Washington reporter publics of corporate public affairs programs. Journalism Quarterly, 60, 603-615. Grunig, J. E. (1989). Publics, audiences and market segments: Models of receivers of campaign messages. In C. T. Salmon (Ed.), Information campaigns: Managing the process of social change (pp. 197-226). Newbur Grunig, J. E. (1997). A situational theory of publics: Conceptual history, recent challenges and new research. In D. Moss, T. MacManus, & D. Public relations research: An international perspective (pp. 3-46). London, International Thompson Business Press. , J. E. (2003). Constructing public relations theory Chaffee, with L. Foreman-Wernet (Eds.), Communication, another kind of horse race: Essays honoring Richard F. Carter (pp. 85-115). Cresskill, NJ: Hamp 366 Grunig s , Portland, OR. Grunig, J. E., & Grunig, L. A. (1996, May). Implications of symmetry for a theory of ethics and social responsibility in public relations. Paper presented to the International Communication Association, Chicago. Grunig, J. E., & Grunig, L. A. (1997, July). Review of a program of research on activism: Incidence in four countries, activist publics, strategies of activist groups, and ental Issues, Lake Bled, Slovenia. Grunig, J. E., & Grunig, L. A. (2000). Introducing the ethics of pr research. Jim & Lauri Grunig?s Research, a supplement of PR Reporter, No. 9, August 28, 2000. inston. issues: A second study. Journalism Quarterly, 56, 715-726. , J. E. (2005). Situational theory of publics. In R. L. Heath (Ed.), Encyclopedia of public relations (pp. 778-780). Thousand Oak, CA: Sage. Grunig, J. E., & Childers (a. k. a. Hon), L. (1988, July). Reconstruction of a situational theory of communication: Internal and external concepts as identifiers of public for AIDS. Paper presented to the Association for Education in Journalism and Mass Communication Grunig, J. E., & Disbrow, J. B. (1977). Developing a probabilistic model for communication decision making. Communication Research, 4, 145-168. organizational responses to activism. Paper presented to the Fourth Public Relations Research Symposium, Managing Environm Grunig, J. E., & Hunt, T. (1984). Managing public relations. New York: Holt, Rinehart and W Grunig, J. E., & Stamm, K. R. (1979). Cognitive strategies and the resolution of environmental 367 Grunig, L. A. (1992). Activism: How it limits the effectiveness of organizations and how excellent public relations departments respond. In J. E. Grunig (Ed.), Excellence in public relations and communication management: Contributions to effective organizations (pp. 503-530). Hillsdale, NJ: Erlbaum. ganizations: A study of communication management in three countries. Gudyk ty, uncertainty, and perceived Hallaha arch, 13, 27-59. o in theory, methods, process, and research Hamilt theory: A replication, application, and Hancock, G. R. (2006). Power analysis in covariance structure modeling. In G. R. urse Hancoc Grunig, L. A., Grunig, J. E., & Dozier, D. M. (2002). Excellent public relations and effective or Mahwah, NJ: Erlbaum. unst, W. B., & Nishida, T. (2001). Anxie effectiveness of communication across relationships and cultures. International Journal of Intercultural Relations, 25, 55-71. n, K. (2001). The dynamics of issue activation and response: An issues processes model. Journal of Public Relations Rese Hamilton, M. A., & Nowak, K. L. (2005). Information systems concepts across tw decades: An empirical analysis of trends domains. Journal of Communication, 55, 529-553. on, P. K. (1992). Grunig?s situational extension. Journal of Public Relations Research, 4, 123-150. Hancock & R. O. Mueller (Eds.), Structural equation modeling: A second co (pp. 69-115). Greenwich, CT: Information Age. k, G. R., & Mueller, R. O. (2001). Rethinking construct reliability within latent variable systems. In R. Cudeck, S. du Toit, & D. S?rbom (Eds.), Structural 368 equation modeling: Present and future?Festschrift in honor of Karl J?reskog (pp. 195-216). Lincolnwood: Scientific Software International. Sons. Hempel, C. G., & Oppenheim, P. (1953). The logic of explanation. In H. Feiger & M. k: Higgins, E. T. (1996). Knowledge activation, application, and salience. In E. T. Higgins p. ew York: Guilford. . Hyman J?resko mn2.htm Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley & Brodbeck (Eds.), Reading in the philosophy of science (pp. 319-352). New Yor Appleton-Century. & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (p 237-275). N Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariances structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55. Hurley, P. J. (1997). A concise introduction to logic (6 th ed.). Belmont, CA: Wadsworth , H. H., & Sheatsley, P. B. (1947). Some reasons why information campaigns fail. Public Opinion Quarterly, 11, 412-423. Johnson-Laird, P. N., & Bara, B. G. (1984). Syllogistic inference. Cognition, 16, 1-61. g, K. G. (1999). How large can a standardized coefficient be? http://www.ssicentral.com/lisrel/colu - Katona d economic behavior. Psychological Review, 60, 307-318. Katz, E., & Lazarsfeld, P. F. (1955). Personal influence. Glencoe: Free Press. Katona, G. (1951). Psychological analysis of economic behavior. New York: McGraw Hill. , G. (1953). Relational behavior an 369 Kim, J.-N., Downie, M., & De Stefano, H. (2005, March). Resolving multicollinearity in situational theory of publics: Conceptual explication of problem recognition. , eling. New York: Guilford. wbury Park, CA: Sage. Kraut, R. E., Egido, C., & Galegher, J. (1990). Patterns of contact and communication in scientific research collaboration. In J. Galegher, R. E. Kraut, & C. Egido (Eds.), Intellectual teamwork: Social and technological foundations of cooperative work (pp. 149-171). Hillsdale, NJ: Erlbaum. Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cognitive and Krugla ging social psychology. Mahwah, NJ: Erlbaum. Kruglanski, A. W., & Thompson, E. P. (1999). Persuasion by a single route: A view from Krugla o, A., Mannetti, L., & Spiegel, S. gnition from a single?mode perspective. In J. Forgas, B. von Hippel, & K. Williams (Eds.), Motivational affective and cognitive processes of human judgment (pp. 137-161). Cambridge, U. K.: Cambridge University Press. Paper presented to the 8th International Public Relations Research Conference Miami. Kline, L. B. (1998). Principles and practice of structural equation mod Kraemer, H. C., & Thiemann S. (1987). How many subjects? Statistical power analysis in research. Ne motivational bases. New York: Plenum. nski, A. W. (in press). Theories as bridges. In P. van Lange (Ed.), Brid the unimodel. Psychological Inquiry, 10, 83-109. nski, A. W., Chun, W. Y., Erb, H. P., Pierr (2003). A parametric unimodel of human judgment: Integrating dual-process frameworks in social co 370 Krugman, H. E. (1965). The impact of television advertising: Learning without involvement. Public Opinion Quarterly, 29, 349-356. . S. (1962). The structure ofKuhn, T scientific revolutions. Chicago: University of Chicago Press. Lasswell, H. D. (1948). The structure and function of communication in society. In L. Bryson (Ed.), The communication of ideas (pp. 37-51). New York: Harper and Brothers. k: Lewis, G. H. (1975). Topic II: Tumbling dice and city blocks: Probability and sampling. In G. H. Lewis (Ed.), Fist-fights in the kitchen: Manners and methods in social research (pp. 83-93). Pacific Palisades, CA: Goodyear. Libet, B., Gleason, C. A., Wright, E. W., Jr., & Pearl, D. K. (1983). Time of conscious intention to act in relation to onset of cerebral activities (readiness-potential): The unconscious initiation of a freely voluntary act. Brain 106, 623-642. MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). variable effects. : Journal of Public Relations Research, 5, 251-268. Marcus, G. E., Neuman, W. R., & MacKuen, M. (2000). Affective intelligence and political judgment. Chicago: University of Chicago Press. Lazarsfeld, P. F., Berelson, B., & Gaudet, H. (1944). The people?s choice. New Yor Free Press. A comparison of methods to test mediation and other intervening Psychological Methods, 7, 83-104. Major, A. M. (1993). Environmental concern and situational communication theory Implications for communicating with environmental publics. 371 Marshall, C., & Rossman, G. B. (1995). Designing qualitative research (2 nd ed.). Thousand Oaks, CA: Sage. McCombs, M. E., & Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36, 176-187. McCul T. Press. McDonough, A. M. (1963). Information economics and management systems. New York: McGraw-Hill. McLeod, J. M., & Chaffee, S. H. (1973). Interpersonal approaches to communication research. American Behavioral Scientist, 16, 469-500. McQuail, D. (1997). Audience analysis. Thousand Oaks, CA: Sage. McQuail, D., & Windahl, S. (1993). Communication models (2 nd ed.). London, U. K.: Mellma ver, Matthews, Smith. on Neteme a, S. (2003). Scaling procedures: Issues and Newco municative acts. Psychological McCracken, G. (1998). The long interview. Beverly Hills, CA: Sage. loch, W. S. (1965). Embodiments of mind. Cambridge, MA: M. I. Longman. n Group (1999, September). Socially engaged Internet users: Prospects for online philanthropy and activism. Arlington, VA: Cra Muth?n, L. K., & Muth?n, B. (2002). How to use a Monte Carlo study to decide sample size and determine power. Structural Equation Modeling, 4, 599-620. yer, R. G., Bearden, W. O., & Sharm applications. Thousand Oaks, CA: Sage. mb, T. M. (1953). An approach to the study of com Review, 60, 393-404. 372 Oakhill, J., & Johnson-Laird, P. N. (1985). The effect of belief on the spontan production of syllogistic conclusions. Qu eous arterly Journal of Experimental Psychology, 37A, 553-570. Obhi, S. S., & Haggard, P. (2004). Free will and free won?t. American Scientist, 9, 358- 365. Pavitt, collective action and social collectivity formation. Unpublished manuscript. Perrow, C. (1989). On not using libraries. In B. P. Lynch (Ed.), Humanists at work: ity Pinch, T., & Bijker, W. (1984). The social construction of facts and artifacts: Or how the r, Popper s and refutations. New York: Harper. Popper, K. R. (1999). All life is problem solving. New York: Routledge. Radnitzky, G., & W. Bartley, W. W. III. (1987). Epistemology, rationality and the Reeves tion M. E. Roloff & C. R. Berger (Eds.), Social cognition and communication (pp. 287-326). Beverly Hills, CA: Sage. Over, D. E., & Evans, J. St. B. T. (1997). Two cheers for deductive competence. Current Psychology of Cognition, 16, 255-278. C. (2003). Toward the scientific explanation of Disciplinary perspectives and personal reflections (pp. 29-42). Chicago: University of Illinois at Chicago, Institute for the Humanities and the Univers Library. sociology of science and the sociology of technology might benefit each othe Social Studies of Science, 14, 399-441. , K. R. (1963). Conjecture sociology of knowledge. Chicago: Open Court. , B., Chaffee, S. H., & Tims, A. (1982). Social cognition and mass communica research. In 373 Richins, M. L. (1983). Negative word-of-mouth by dissatisfied consumers: A pilot study. Journal of Marketing, 47, 68?78. Rothchild, M. L., & Ray, M. L. (1974). Involvement and political advertising effect: An exploratory experiment. Communication Research, 1, 264-285. s , pp. 243-269). Norwood, NJ: Ablex. avior. Journal of Personality and Social Psychology, 49, 843-851. Schwarz, N., Groves, R. M., & Schuman, H. (1998). Survey methods. In D. T. Gilbert, S. T. Fiske, & L. Gardner (Eds.), The handbook of social psychology (4th ed.), (Vol. 1, pp. 143-179). Boston, MA: McGraw-Hill. Sears, D. O., & Freedman, J. L. (1965). The effects of expected familiarity with ality Sears, D ). Selective exposure to information: A critical review. Public Opinion Quarterly, 31, 194-213. Shannon, C.E., & Weaver, W. (1949). The mathematical theory of communication. Rosen, E. (2000). The anatomy of buzz: How to create a word-of-mouth marketing. New York: Doubleday. Salmon, C. T. (1986). Perspectives on involvement in consumer and communication research. In B. Dervin & M. J. Voigt (Eds.), Progress in communication science (Vol. 7 Schifter, D. E., & Ajzen, I. (1985). Intention, perceived control, and weight loss: An application of the theory of planned beh arguments upon opinion change and selective exposure. Journal of Person and Social Psychology, 2, 420-426. . O., & Freedman, J. L. (1967 Urbana: University of Illinois Press. 374 Shapiro, M. A. (2002). Generalizability in communication research. Human Communication Research, 28, 491-500. Signorlie, M. (1995). Outing yourself: How to come out as lesbian or gay to your family, friends, and coworkers. New York: Random House. Simon, H. (1957). Administrative behavior (2 nd ed.). New York: Free Press. Simon, H. (1994). Bottleneck of attention: Connecting thought without motivation. In W. D. Spaulding (Ed.), Integrative views of motivation, cognition, and emotion (pp. 1-21). Lincoln: University of Nebraska Press. Simon, H. A. (1967). Motivational and emotional controls of cognition. Psychological Review, 74, 29-39. Slater, M. D., Chipman, H., Auld, G., Keefe, T., & Kendall, P. (1992). Information processing and situational theory: A cognitive response analysis. Journal of Public Relations Research, 4, 189-203. Slovic, P. (1993). Perceived risk, trust, and democracy. Risk Analysis 13, 675-82. Sobel, M. E. (1982) Asymptotic intervals for indirect effects in structural equations models. In S. Leinhart (Ed.), Sociological methodology (pp. 290-312). San Francisco: Jossey-Bass. Spector, P. E. (1992). Summated rating scale construction: An instruction. Newbury Park, CA: Sage. Steiger, J. H., & Lind, J. C. (1980, May). Statistically-based tests for the number of common factors. Paper presented at the meeting of the Psychometric Society, Iowa City. 375 376 Tewksbury, D. (2003). What do Americans really want to know? Tracking the behavior of news readers on the internet. Journal of Communication, 53, 4, 694-710. Thayer, L. (1987). How does information ?inform?? In B. D. Ruben (Ed.), Information and behavior (Vol. 2, pp. 13-26). New Brunswick, NJ: Transaction Books. Troldahl, V. C., & Van Dam, R. (1965). Face-to-face communication about major topics in the news. Public Opinion Quarterly, 29, 626-634. Vasquez, G. M., & Taylor, M. (2001). Research perspectives on ?the public.? In R. L. Heath (Ed.), Handbook of public relations (pp. 139-154). Thousand Oaks, CA: Sage. Veyne, P. (1988). Did the Greeks believe in their myths? Translated by P. Wissing. Chicago: University of Chicago Press. Wegner, D. M., & Wenzlaff, R. M. (1996). Mental control. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 466-492). New York: Guilford. Weisberg, H. F., Krosnick, J. A., & Bowen, B. D. (1996). An introduction to survey research, polling, and data analysis (3P rd P ed.). Thousand Oaks, CA: Sage. West, S. J., Finch, J. E., & Curran, P. J. (1995). Structural equation models with nonnormal variables. In R. Hoyle (Ed.), Structural equation modeling (pp. 56-75). Thousand Oaks, CA: Sage. Wilson, P. (1995). Unused relevant information in research and development. Journal of the American Society for Information Science, 46, 45-51. Wilson, P. (1996). Research and information overload. Library Trends, 45, 194-203. 377 Zaller, J., & Feldman, S. (1992). A simple theory of the survey response: Answering questions versus revealing preferences. American Journal of Political Science, 36, 3, 579-616.