ABSTRACT 
Title of Dissertation: DECOMPOSING CHARISMATIC LEADERSHIP: 
THE EFFECTS OF LEADER CONTENT AND 
PROCESS ON FOLLOWER PERFORMANCE , 
ATTITUDES, AND PERCEPTIONS 
Shelley Ann Kirkpatrick, Doctor of Philosophy, 1992 
Dissertation directed by: Professor Edwin Locke, 
Chair of Management and 
Organization, College of 
Business and Management 
Leadership entails both content and process elements, 
e.g., what the leader says and how the leader says it. 
For example, charismatic leaders are theorized to 
communicate and implement a vision (content) with an 
enthusiastic communication sty l e (process). In a 
laboratory experiment with manipulated independent 
variables and a simulated task, this dissertation 
separately examined the content and process components of 
charismatic leadership on performance and attitudes. 
The content aspect was separated into two parts, 
vision (versus no vision) and implementation of the vision 
through task strategies (versus no task strategies). 
Process was manipulated as enthusiasm level (low versus 
high). Thus, a 2 x 2 x 2 design was employed. Two 
trained actors, one male and one female, played the role 
of leader, a CEO/President of a local printing company. 
Upper-level business students served as participants and 
performed a binder assembly task. Students completed 
questionnaires before each session and at the end of the 
experiment to assess how they are influenced by the 
leader. 
Results indicated that content affected performance 
and many attitudes and perceptions. Process did not 
affect performance and affected only a few attitudes and 
perceptions. Exploratory analyses showed that self-set 
goals and self-efficacy served as mediators between the 
content variables and performance. Theoretical, 
methodological, and practical implications are discussed. 
DECOMPOSING CHARISMATIC LEADERSHIP: THE EFFECTS OF LEADER 
CONTENT AND PROCESS ON FOLLOWER PERFORMANCE, 
ATTITUDES AND PERCEPTIONS 
by 
Shelley Ann Kirkpatrick 
Dissertation submitted to the Faculty of the Graduate School 
of The University of Maryland in partial fulfillment 
of the requirements for the degree of 
Doctor of Philosophy 
1992 
C \ ( u!I f) t rv1c r, cf l 
Advisory Committee: 
Professor Edwin Locke, Chair/Advisor 
Professor Stephen Carroll 
Professor Richard Guzzo 
Professor Henry P. Sims, Jr. 
Associate Professor Judy Olian 
? Copyright by 
Shelley Ann Kirkpatrick 
1992 
ii 
ACKNOWLEDGEMENTS 
This dissertation would not have been possible 
without the advice, time, and support of many others. I 
feel very fortunate to have worked with such professional 
and dedicated individuals. 
I would like to thank Robert House for his valuable 
suggestions regarding the attitude and perception 
variables as well as the experimental design. Thanks go 
also to Jane Howell. Her advice regarding actor selection 
and actor training was indispensable to this dissertation. 
I would also like to thank Bruce Avolio for allowing me to 
use the charisma scales, Tim Judge for providing the mood 
adjective scale, and Hank Sims for allowing me to use the 
intellectual stimulation items. 
I am greatly indebted to both actors, Bart Whiteman 
and Melinda Murphy. They endured endless hours of 
rehearsal, experimental sessions, and commuting to the 
University of Maryland. I could not have chosen more 
talented actors, and the experiment would not have been 
possible without their hard work and dedication. 
Bart and David Kinlein donated the binders used for 
the task, which I much appreciate. I also thank Bart for 
taking time to discuss his company and to explain the 
binder tasks to me and to the actors. This information 
iii 
was crucial to the success of the dissertation. 
I am also grateful to Steve Carroll and Ken Kovach 
for allowing the students in their classes to volunteer 
f or the experiment . 
Dean Rudolph Lamone provided funding in addition to 
the research grant I received from my Graduate School 
Fellowship, and I am especially grateful to him for this 
support. Hank Sims, in his ro l e as Director of the 
Doctoral Program, deserves many thanks for helping me 
apply for and receive the Fellowship. This allowed me to 
concentrate full-time on this dissertation. 
I cannot give enough recognition to my dissertation 
chair, Ed Locke, for his guidance on this dissertation as 
well as on every other project on which we have worked 
over the past four years. His amazing ability to provide 
fast and accurate feedback and to solve seemingly 
impossible problems are skills that I hope to emulate 
throughout my career. 
My dissertation committee, Judy Olian, Hank Sims, 
Steve Carroll, and Rick Guzzo, also deserve recognition 
for supporting me in this dissertation. Their input 
greatly improved the dissertation and final product. 
Thanks also goes to Amit Gupta for volunteering 
several evenings to get the materials ready for the next 
morning's session. 
iv 
Finally, I would like to thank my husband, Rick, for 
supporting me throughout graduate school, including 
several evenings preparing materials for the next 
morning's session as well as endless hours of helping me 
with the computer. His emotional support kept me working 
at my peak to complete this dissertation. 
V 
TABLE OF CONTENTS 
Section Page 
List of Tables viii 
List of Figures X 
Chapter 1 Introduction 1 
Charisma and Charismat i c Leadership 2 
Chapter 2 Literature Review: Charismatic Leadership 4 
Weber's Theory 4 
Transformational Leadership: Burns and Bass 5 
Charismatic Theories: House and Conger 11 
Conclusion 20 
Chapter 3 Vision, Vision Implementation, and 22 
Enthusiastic Interaction style 
Vision 22 
Vision Implementation: Task Strategies 30 
Enthusiastic Communication Style 32 
Mechanisms of Charismatic Leadership 34 
Conclusion 37 
Chapter 4 Research Propositions 39 
Vision 39 
Vision Implementation: Task Strategies 42 
Enthusiastic Communication Style 43 
Mediating Mechanisms: Goals and Self-efficacy 45 
vi 
Chapter 5 Design and Method 4 7 
Design 4 7 
subjects 49 
Task 5 0 
Procedure 52 
Manipulations 56 
Measures 60 
Chapter 6 Results: Pilot Ratings , Manipu l ation 73 
Checks, Assumptions, and Independent 
Variable Effects 
Pilot Ratings of Actors 73 
Manipulation Checks 74 
Assumptions Regarding Manipulations 76 
Descriptive Statistics and correlations 82 
Effects of Independent Variables 93 
Summary 109 
Chapter 7 Results: supplementary and Mediator 112 
Analyses 
Charisma Scale Components 112 
Goal Variable Results 117 
Chapter 8 Integration of Quantitative Results 153 
Initial Path Model 154 
Revised Path Model 159 
Final Path Model 163 
Summary: Conceptual Model 168 
vii 
Chapter 9 Qualitative Results 170 
Subjects' Questions During Experiment 170 
subjects' Comments 171 
Experimenter's Observations 179 
Actors' Comments 180 
summary 183 
Chapter 10 Discussion 184 
Limitations 184 
Implications 187 
Future Research Directions 202 
Conclusion 207 
References 208 
Appendix A Vision Script 218 
Appendix B No Vision 
221 
Script 
224 
Appendix C Task Strategy Script 
226 
Appendix D Attitude and Perception Items 
Appendix E Goal and 
231 
Self-efficacy Items 
viii 
LIST OF TABLES 
Number Page 
1 . Principal Components Analysis o f Mood Scales 65 
2. Repeated Measures Across Tr i als 80 
3 . Means and Standard Deviations for Performance, 83 
Attitude, and Perception Variables 
4a. Vision: Means and standard Deviations for 87 
Performance, Attitude, and Perception Variables 
4b. Task Strategy: Means and standard Deviations for 89 
Performance, Attitude, and Perception Variables 
4c. Enthusiasm: Means and standard Deviations for 91 
Performance, Attitude, and Perception Variables 
5 . Correlations Among Independent, Performance, 94 
Attitude, and Perception Variables 
6 . Analysis of {Co-)Variance Results 97 
7. Principal Components Analysis of Charisma Scales 114 
8. Means and Standard Deviations for Goal Scales 118 
9a. Vision: Means and Standard Deviations for 120 
Goal Scales 
9b. Task Strategy: Means and Standard Deviations for 122 
Goal Scales 
9c . Enthusiasm: Means and Standard Deviations for 12 4 
Goal Scales 
10. Correlations Among Goal Variables 126 
ix 
11. Correlations Between Goal Variables and 127 
Independent and Performance Variables 
12. Analysis of Co-variance Results for Goal Scales 128 
13. Mediation Test for Personal Goal 140 
14. Mediation Test for Self-efficacy 144 
15. Mediation Test for Goal Commitment 149 
16. Standardized Path Coefficients for Figure 8 157 
17. Principal Components Analysis of Attitude and 160 
Perception Variables 
18. Standardized Path Coefficients for Figure 9 162 
19. Standardized Path Coefficients for Figure 10 165 
X 
LIST OF FIGURES 
Number Page 
1. Research Design 48 
2. Summary of Variables 62 
3. Trials by Task Strategy Interaction 81 
4. Vision by Task Strategy Interaction on Anxiety 110 
5. Vision by Enthusiasm Interaction on Anxiety 111 
6. Vision by Enthusiasm Interaction on Goal Variables 133 
7. Vision by Task Strategy Interaction on Anticipated 135 
Satisfaction for Quantity 
8. Initial Path Model 156 
9. Revised Path Model 161 
10. Final Path Model 164 
11. Conceptual Model 169 
1 
Chapter 1 
Introduction 
Leadership has been a topic of interest since ancient 
times and remains a topic of interest to today's 
researchers and practitioners. It is more relevant than 
ever in today's rapidly changing environment. 
Organizations are relying more a nd more on leaders at all 
levels of the hierarchy to guide the organization through 
turbulent times while increasing product quality as well 
as profits. 
There are many definitions of leadership (Yukl, 
1 989) ? One definition that captures the basic elements of 
leadership is the following (Locke et al., 1991): 
''Leadership is the process of inducing others to take 
action toward a common goal." This definition implies 
that leadership is a relationship between a leader and 
followers (Kouzes & Posner, 1987); it cannot exist in 
isolation from followers. The definition further implies 
that leadership is a process. Leadership is not simply a 
title or position, but consists of actions performed by 
the leader. Finally, the definition posits that leaders 
must induce, that is, persuade or arouse, rather than 
force followers to work toward common goals. They do this 
using a variety of procedures, such as communicating a 
2 
vision using legitimate authority, role modeling, goal 
setting, rewarding and punishing, organizational 
restructuring, and team building (Locke et al., 1991). 
This dissertation will focus on one specific type of 
leadership, charismatic leadership. More specifically, 
this dissertation investigates how components of charisma 
influence follower performance, attitudes, and 
perceptions. 
Charisma and Charismatic Leadership 
Charisma and charismatic leadership are two 
distinguishable though overlapping concepts. Charisma is 
defined as magnetic charm or appeal (Webster's New 
Collegiate Dictionary). Charismatic individuals 
demonstrate more animated behaviors and facial expressions 
than non-charismatic individuals, which arouses emotions 
in othe rs {Friedman, Prince, Riggio, & DiMatteo, 1980; 
Kouzes & Posner, 1987). By strict definition, charisma is 
a personality trait or style. Thus, a leader with 
charisma is a leader who acts charismatically regardless 
of the content of the actions or statements made. 
In the current literature (e.g., Bradley, 1987; 
House, Spangler, & Woycke, 1991), the term charismatic 
leadersh ip has taken on a broader meaning than simply a 
leader who possesses a charismatic personality. 
3 
Charismatic leadership is said to encompass not only (a) 
the leader's personality traits, of which charisma is the 
core trait, but also (b) specific actions taken by the 
leader, such as communicating and implementing a vision, 
as well as (c) the leader's effects on followers, such as 
changes in followers' performance and changes in 
followers' attitudes and perceptions toward the leader. 
This dissertation deals with all three aspects of 
charismatic leadership. 
Charismatic leadership theories are discussed in 
detail in the next chapter. 
4 
Chapter 2 
Literature Review: Char ismatic Leadersh i p 
First, Weber's theory is p resented . Then, two 
transformational leadership the or i es are revi ewed. 
Finally, two charismatic leadership theories are 
discussed. 
Weber's Theory 
Weber (1947) was the first theorist to use the term 
charisma. He used the term "charismatic authority" to 
refer to a form of social authority. Charismatic 
authority, according to Weber, exists when an individual 
claims to have or is thought to be endowed with 
supernatural, superhuman, or exceptional powers. These 
powers were ?seen as divine in origin, based on God or gods 
(Tucker, 1968). such powers were not possessed by 
ordinary men. Weber (1947, p. 358) also stated that 
charismatic leaders "reveal a transcendent mission 
[vision] or course of action which may be in itself 
appealing to the potential followers, but which is acted 
on because the followers believe their leader is 
extraordinarily gifted." Still, Weber has been criticized 
as being unclear as to exactly what behaviors distinguish 
5 
charismatic leaders from non-charismatic leaders (Dow, 
1969). 
Weber also proposed that charismatic leaders emerge 
during times of distress. Charismatic leaders possess a 
divine, magical personality trait that only comes forward 
during crisis situations (Weber, 1947). However, a study 
by Willner (1984) using political and religious leaders 
found that charismatic leaders may also emerge during non-
crisis situations. 
Transformational Leadership: Burns and Bass 
Transformational leadership theories are considered 
by some to be distinct from charismatic leadership 
theories. For clarity, they will be reviewed separately 
although they are highly similar to charismatic theories 
and are even seen as the same by other theorists (e.g. 
House, Woycke, & Fodor, 1988). 
Burns' Transformational and Transactional Leadership 
Burns is well known in the leadership literature for 
advancing Weber's ideas. He made the distinction between 
transformational and transactional leadership. Burns 
(1978, p. 19) defines leaders as "inducing followers to 
act for certain goals that represent values and the 
motivations - the wants and needs, the aspirations and 
expectations - of both leaders and followers." 
6 
Transformational leaders recognize existing needs or 
demands in followers and seek ways to elevate the 
followers to satisfy these needs . They clarify followers' 
goals, causing them to set and a ccept higher transcendent 
goals (House, 1977). In contrast, transactional 
leadership is based on an exchange relationship between 
the leader and followers. For e xample, in exchange for 
pay, followers (employees) agree to meet their daily 
quota. 
Bass' Transformational Theory 
Bass (1985) has identified some specific behaviors 
that transformational and transactional leaders exhibit. 
Transactional leaders use contingent reward, which usually 
comes in the form of praise for work well done and 
recommendations for pay increases, bonuses, and promotions 
(Sims, 1977). They also rely on management-by-exception, 
only taking action when something goes wrong and use 
negative feedback and contingent aversive reinforcement 
(or punishment) when a follower fails or performs an 
undesirable behavior. 
Transformational leaders, on the other hand, rely on 
charisma. One form of charismatic leadership is what Bass 
calls inspirational leadership, which usually involves 
giving speeches and motivating pep talks. According to 
Bass' typology, transformational leaders also engage in 
7 
individualized consideration. They treat each subordinate 
independently depending on each one's unique needs and 
capabilities. Finally, they engage in intellectual 
stimulation in the form of teaching and coaching 
subordinates and arousing them to think in new ways. 
Based on the typology of transactional and 
transformational leadership, Bass developed the 
Multifactor Leadership Questionnaire. Items describing 
each of the above five lea~ership styles (contingent 
reward, management-by-exception, charisma, individualized 
consideration, and intellectual stimulation) were written 
based on a literature review (Bass, 1981) that highlighted 
the distinctive aspects of each leadership style. Then, 
he had 104 military officers complete the questionnaire, 
answering with respect to the degree to which each item 
described his or her superior. Factor analysis results 
supported the existence of the five factors, although the 
relationship between each factor and leader effectiveness 
was not examined. This factor analysis has been 
replicated in a subsequent study, providing evidence for 
the reliability of the MLQ's factor structure (Bass & 
Avolio, 1989). However, one drawback of Bass' scales is 
that they do not distinguish between leader behaviors and 
the effects of such behaviors (Yukl, 1989). For example, 
the action of giving speeches can result in followers 
8 
feeling inspired. Bass' scales include items measuring 
leader behaviors as well as the effects of those behaviors 
on the followers. 
Empirical Studies on Transformational Leadership 
Hater and Bass (1988) gathered data on employees in 
an express mail delivery organization. Managers were 
divided into two groups, top performers and ordinary 
performers. Subordinates completed the MLQ to measure 
their perceptions of their manager's transformational or 
transactional behaviors. Top performing managers were 
significantly higher than ordinary managers on the 
charismatic and individualized consideration factors. 
Results indicated that top performers and ordinary 
managers did not significantly differ on transactional 
factors. Transformational factors (charismatic leadership 
and individualized consideration) significantly predicted 
the performance of the manager's work group but 
transactional factors did not. Although the study used 
managers and not leaders, subordinates had little trouble 
identifying transactional and transformational leadership 
characteristics in their managers. 
Avolio, Waldman, and Einstein (1988) examined the 
effects of transactional and transformational leaders on 
group performance on a management simulation game. At the 
beginning of the 3-month period over which the game took 
9 
place, each of 27 teams elected a leader. Toward the end 
of the game, group members rated their leader on the five 
MLQ scales. Regression analys i s indicated that 
individualized consideration, when entered first into the 
equation, significantly accounted for 22% of the variance 
in group performance. When charismatic/inspirational 
leadership was entered after individualized consideration, 
it added 9% variance, for a total of 33% accounted for. 
Intellectual stimulation, contingent reward, and 
management-by-exception did not significantly add unique 
variance. When charisma was entered first into the 
regression equation, it significantly accounted for 14% of 
the variance in group performance. Although this study 
shows that transformational leadership is significantly 
related to group performance, the causal direction of this 
relationship is unclear. one possibility is that group 
members may have been more likely to rate their leader as 
transformational if group performance was high. Another 
possibility is that as a result of the group performing 
well during the first part of the game, the leaders may 
have exhibited more transformational behaviors. 
Rosener (1990) surveyed men and women leaders about 
their own leadership styles and personal characteristics. 
Each female leader was matched with a male leader in a 
similar organization with similar responsibilities. The 
10 
main findings indicated that the women were more likely to 
report using power based on charisma , work record, and 
contacts as opposed to power based on organizational 
position, title, and the ability to reward and punish. 
Rosener suggests that women, therefore, may be more likely 
to use transformational leadership techniques. She 
defines transformational leadership broadly to include 
techniques such as making individuals feel important, 
sharing information and power, enhancing self-worth of 
others. Men were more likely to report relying on 
transactional techniques, using authority and the ability 
to reward and punish. However, Rosener did not obtain 
independent (e.g. subordinate) ratings of the leaders' 
styles; it may be possible that women simply describe 
their styles differently than men. 
In sum, the above studies are a first step toward 
understanding the nature of transformational leadership. 
They have shown that transformational behaviors can be 
readily observed in leaders and can be reliably measured 
through observation and questionnaires. This research is 
consistent with charismatic leadership theory and 
research, examined next. 
11 
Charismatic Theories : House and Conger 
House (1977) and Conger and Kanungo (1987) have each 
developed similar theories o f c harismatic leadership . 
Each theory is presented in turn, along with relevant 
research. 
House's Theory of Charismatic Leadership 
Based on selected literatur e from social psychology 
(as opposed to sociology or pol i tical sc i ence), House's 
(1977) theory of charismatic leadershi p posits several 
characteristics o f charismatic leaders. These include 
high self-confidence, dominance ' and a strong convic. tion 
in the moral righteousness of their beliefs. Charismatic 
leader behaviors, according to House (1977) include role 
modeling, image building, goal a r ticulati6n (a 
transcendent goal or vision), exhibiting high 
expectations, showing confidence in subordinates, setting 
challenging goals for followers, and arousing follower 
motives. The results or outcomes of these behaviors 
include the following: (a) commitment to the leader; (b) 
identification with and emulation of leader's values, 
goals, and behavior; (c) follower self-confidence; (d) 
trust in the leader. 
House views charismatic leaders in basically the same 
way as Weber and Burns view transformational leaders in 
that charismatic leaders arouse followers motives through 
12 
inspirational speeches and emotional appeals. House makes 
hypotheses about specific characteristics and behaviors of 
charismatic leaders as well as c onsequences of charismatic 
leaders. 
House and colleagues have examined the effects of 
charisma in two separate studies. First, House, Woycke, 
and Fodor {1988) studied charisma in former U.S. 
Presidents. Charisma was measured using historians' 
ratings of presidents biographies and cabinet members' 
biographies. Presidents Thomas Jefferson, Andrew Jackson, 
Abraham Lincoln, Theodore Roosevelt, Franklin D. 
Roosevelt, and John F. Kennedy were identified as 
charismatic. Results indicated that all the charismatic 
presidents were either re-elected or assassinated. But, 
only a small percentage of non-charismatic presidents were 
either re-elected or assassinated . This suggests that 
charismatic leaders arouse strong emotional feelings in 
others. Charismatic presidents were significantly higher 
than neutral and non-charismatics on effectiveness 
ratings, including general prestige, strength of action, 
presidential activeness, flexibility, and accomplishment 
of their administrations. House and colleagues also found 
that charisma accounted for almost half of the statistical 
variance in leader effectiveness, as measured by 
historians' retrospective ratings. 
13 
In a follow-up study, House, Spangler, and Woycke 
(1991) developed a general model of charismat i c 
leadership. They argued that effectiveness was a function 
of leader needs for power, achievement, a f fi liation, and 
power inhibition. Charisma and need for power were 
significantly related to rated presidential performance, 
while need for achievement and need for affiliation showed 
no significant first order corre lations. Fur ther, motives 
and charisma together predicted f rom 24% to 66% of the 
statistical variance in president ial performance. 
Conger and Kanungo's Charismatic Leadership Theory 
Conger and Kanungo (1987; see also Conger, Kanungo, & 
Associates, 1988, and Conger, 1989) have formulated their 
own theory of charismatic leadership in the f orm of 
research propositions. They identify behavioral 
components similar to those presented by House. Conger 
and Kanungo do not see charisma solely due to personality 
factors or situational factors but as a "charismatic 
relationship" between leader and f ollowers. They argue 
that "a leader becomes charismatic when he/she succeeds in 
changing his/her followers' attitudes to accept the 
advocated vision" (Conger & Kanungo, 1987, p. 64 0) ? The 
leader may be "charismatic when the vision is highly 
discrepant from the status quo yet remains within a 
latitude of acceptance for their followers" (Conger & 
14 
Kanungo, 1987, p. 642). Such leaders communicate a 
vision, try to change the status quo, are likeable, are 
trustworthy, are experts, have unconventional or 
counternormative behavior, and have strong articulation of 
future vision and motivation to lead. 
Conger and Kanungo (1987, p. 645) state that by 
isolating these behavioral components of charismatic 
leadership, "it may be possible to develop these 
attributes in managers." rt would also be possible to 
select managers on the basis of charismatic attributes. 
However, in addition to isolating the behaviors, the 
relationship between the behaviors needs to be examined. 
Conger and Kanungo only state that the behaviors are 
interrelated but do not specify how the components are 
related. For example, attempting to change the status quo 
or exhibiting unconventional behaviors may lead to the 
followers liking and trusting the leader. Clearly, the 
impact of leader behaviors on follower behaviors and 
follower perceptions of the leader needs to be examined. 
Empirical Research on Charismatic Leadership 
Yukl and Van Fleet (1982 ) conducted a field study of 
four s amples of military leaders. Using correlational 
analyse s in two of the studies and content analysis in two 
of the studies, they found that inspirational (or 
charismatic) leadership was significantly related to 
15 
leader effectiveness and high levels of follower 
motivation. Other leadership behaviors important for 
group performance were role clarification, criticism-
discipline, and performance emphasis. These findings held 
under combat, noncombat, and simulated combat conditions. 
Results also indicated that consideration was important 
for maintaining positive leader-subordinate relations in 
noncombat situations but was unrelated to group 
performance. 
House's (1977) theory of charismatic leadership was 
supported by Howell and Higgins (1990), who examined the 
differences between informal project champions (leaders) 
and non-champions across 28 organizations. To control for 
differences in organizational size, industry, and other 
variables relating to the specific innovation, they formed 
25 pairs of champions and non-champions. Questionnaires 
and interviews were used to gather data on leader 
behaviors and performance. Results indicated that 
champions scored higher than non-champions on the 
characteristic of self-confidence as well as on measures 
of charismatic behavior, including strong ideological 
conviction about goals, consideration for others, 
expression of high expectations for others, and showing 
confidence in others. 
16 
Howell and Frost (1989). A study conducted by Howell 
and Frost (1989) is particularly relevant to this 
dissertation and is examined i n detai l. Howell and Frost 
(1989) conducted an elaborate l eadership simulation . In a 
laboratory setting, they examined three styles of 
leadership: charismatic, structuring, and considerate . 
They trained two actresses, chosen based on their similar 
age and attractiveness, to disp l ay each style . 
Charismatic leaders displayed the following behaviors : (a) 
Content of the communication included articulating an 
overarching goal (vision), communicating high performance 
expectations, showing confidence in subordinates' ability 
to meet expectations, and empathizing with the needs of 
their subordinates; (b) Paralinguistic cues included 
exhibiting a captivating and engaging voice tone, 
alternating between pacing and sitting, and leaning 
forward when sitting; (c) Nonverbal behaviors included 
maintaining direct eye contact, having a relaxed posture, 
and showing animated facial expressions; and (d) 
Interaction style was projected as a powerful, confident, 
and dynamic presence. 
structuring leaders exhibited the following 
behaviors: (a) Content of their communication focused on 
factual information about the task, explaining the nature 
of the task, deciding in detail how it is to be done, and 
17 
emphasizing and maintaining definite performance standards 
(quantity of work); (b) Paralinguistic cues included 
simply reading instructions aloud, acting in a 
businesslike manner, and speaking in a moderate voice 
tone; (c) Nonverbal behaviors included sitting behind the 
desk, maintaining intermittent eye contact, and having 
neutral facial expressions (i.e . absence of smiling and 
absence of positive head nods); (d) Interaction style was 
neutral and businesslike. 
Considerate leaders exhibited the following 
behaviors: (a) Content of leaders' communication included 
a concern for the personal welfare of participants, 
participative two-way communications, and emphasizing the 
comfort, well being, and satisfaction of participants; (b) 
Paralinguistic cues consisted of speaking with a warm 
voice tone; (c) Nonverbal behaviors included sitting on 
the edge of the desk, leaning toward participants, 
maintaining direct eye contact, and having a relaxed 
posture; (d) Interaction style was friendly and 
approachable. 
Ratings of the actresses' behavior by independent 
judges, naive to the study's purpose, indicated that the 
actresses successfully portrayed each type of leadership 
style; postexperimental manipulation checks completed by 
participants corroborated their manipulations. There were 
18 
no significant differences between the two actresses in 
their portrayals of the leadership styles. 
In addition, Howell and Frost manipulated group 
productivity norms. Half of the participants under each 
leadership style were exposed to a confederate co-worker 
who expressed enthusiasm and interest in the task; the 
confederate made positive and encouraging statements about 
doing the exercise. The remainder of the participants 
were exposed to a confederate co-worker who expressed 
disinterest and boredom; the confederate made negative 
remarks about the exercise and nonverbally displayed a 
lack of interest in the task. Thus, a 3 x 2 design was 
used, with 3 levels of leadership style and 2 levels of 
group productivity norms. 
Performance on an in-basket task was the main 
dependent variable. Analyses revealed that participants 
exposed to the charismatic leader had higher quality of 
performance than the considerate leader but not the 
structuring leader. Also, participants under charismatic 
leaders had higher specific and general task satisfaction 
and lower role conflict and ambiguity than those under 
considerate leaders. There were no significant 
differences in role ambiguity between structuring and 
charismatic leaders, but individuals with charismatic 
leaders had higher specific and general task satisfaction 
19 
and lower role conflict than those with structuring 
leaders. Those in the charismatic leadership condition 
had higher adjustment to the leader than those with a 
structuring leader or a considerate leader. Adjustment to 
the leader was measured by items assessing the quality of 
the individual's relationship with the leader and co-
workers, personal commitment to the leader, and degree to 
which the individual was motivated by the leader. 
Although there were no main effects for the group 
norm variable, there were significant interactions between 
groups norms and leadership style. Unlike structuring and 
considerate leaders, charismatic leaders (Howell & Frost, 
1989) appeared to overcome group pressures for low task 
productivity and augment group pressures for high task 
productivity, thereby facilitating individuals' adjustment 
to the leader, adjustment to the task (measured by job 
satisfaction, role ambiguity, role conflict, and job-
related tension items), as well as performance in a new 
work setting. 
Overall, Howell and Frost (1989) asserted that 
charismatic leadership is multidimensional in that they 
concluded it consists of at least two components, content 
and process, e.g., what the leader said and how the leader 
said it. The verbal content of the leader's motivational 
influence included a vision that directed followers' 
20 
attention toward the desired goal and communicated high 
performance expectations. Second, the process component 
communicated to followers the leader's own personal 
excitement about the desired goals or vision . Howell and 
Frost's process components were paralinguistic cues, 
nonverbal behaviors, and interaction style. The 
charismatic leaders had, in short, an enthusiastic 
communication style. This component is one way that the 
leader arouses followers' emotions to work toward the 
content of the vision. It includes the leader's 
enthusiasm for the vision and pertains to how the leader 
communicates the importance of the vision . 
Conclusion 
The charismatic leadership theories have at least 
three broad elements or components in common. First, a 
charismatic or transformational leader communicates a 
vision (Burns, 1978; Conger & Kanungo, 1987; House, 1977; 
Weber, 1947). 
Second, the leader suggests or stimulates action to 
implement the vision (House, 1977). This can be in the 
form of intellectual stimulation and individualized 
consideration (Bass, 1985) or leader behaviors such as 
goal setting, role modeling, and image building (House, 
1977) . 
21 
Content can also take the form of task strategies, as 
suggested by Locke et al. (1991). Communicating task 
strategies is one way that the leader can intellectually 
stimulate followers (Bass, 1985) as well as disseminate 
information to followers (Locke et al., 1991). At a more 
macro or organizational level, this implementation of 
vision takes the form of a business strategy. However, at 
the level intended by this dissertation, the concept of 
task strategy is more relevant. 
Third, a charismatic or transformational leader has 
an enthusiastic communication style (Bass, 1985), which is 
the process by which a leader communicates. This includes 
projecting high self-confidence, dominance, and moral 
conviction (House, 1977) as well as being likeable, 
trustworthy, and an expert (Conger & Kanungo, 1987) as 
well as possessing the personality trait of charisma. 
When communicating, charismatic leaders are inspiring 
(Bass, 1985; Conger, 1989). 
Accordingly, this dissertation decomposed charismatic 
leadership into separate components of content (which had 
two sub-components, vision and vision implementation) and 
process, and examined the effects of each on followers' 
performance, attitudes, and perceptions. The next chapter 
presents theories and research on each of these components 
which are ?relevant to the variables in this dissertation. 
22 
Chapter 3 
Vision, Vision Implementation , 
and Enthusiastic I nteraction Style 
This chapter presents research specific to each of 
the three charismatic leadership components. First, 
vision is examined, including the definition of a vision, 
the distinction between vision and goals, and research on 
vision. Second, vision implementation is examined, 
specifically focusing on the role of task strategy as a 
method of implementing vision . Third, the role of an 
enthusiastic communication style is examined. Finally, 
some potential mediators, which may link each of these 
three leader actions to follower performance are 
discussed. 
Vision 
To date, most research on the visionary process has 
been purely qualitative and anecdotal (except for Howell & 
Frost, 1989). Vision has been proposed as a method 
effective leaders elevate followers toward high 
performance (Bass, 1985; Bennis & Nanus, 1985; Burns, 
1978; Conger & Kanungo, 1988; Conger, 1989; Tichy & 
Devanna, 1986). As stated above, the importance of the 
vision has stemmed from research on charismatic leadership 
23 
and transformational leadership (Bass, 1985; Bennis & 
Nanus, 1985; Burns, 1978; Conger & Kanungo, 1988). The 
vision is the core content of the leader's motivational 
influence. 
Definition 
A vision is a leader's ''ideal and unique image of the 
future" (Kouzes & Posner, 1987, p. 85). In theory, one 
can make the distinction between a leader's uncommunicated 
vision, or the ideas that the leader has but is unable or 
unwilling to communicate, and the communicated vision, or 
the ideas that the leader actually communicates to the 
organization. For presentation purposes, I will not 
distinguish between these two concepts; thus, I will 
assume the leader is willing and able to communicate his 
or her vision. This assumption should be tested in future 
research. If a leader does not communicate the ideas, 
then it is not a vision but simply a personal fantasy. 
The vision is often called a guiding principle, 
credo,. or aspirations statement (Locke et al., 1991; 
Howard, 1990). It is a long-term, idealistic standard of 
excellence which is realistic but is never fully attained 
in practice. The vision is a total picture toward which 
the leader wishes to work. For example, Robert N. Beck 
(1987, p. 35), head of corporate human resources at 
BankAmerica Corp. states, "it was clear from ... interviews 
24 
with the top three or four levels of management that there 
was not a shared business vision .... Individuals knew their 
organizational goals to some extent, but they did not have 
a strong feeling of commitment and understanding as to 
where the corporation was going and why it existed." 
Subsequently, Beck and BankAmerica's employees developed a 
vision that gave them a common understanding of their 
company's purpose and future directions. 
In addition to providing a common frame of reference, 
vision is a vehicle for change, allows for resolution of 
conflicts, maintains focus on larger issues, motivates 
employees, and sustains attention to excellence (Bradford 
& Cohen, 1984; Sashkin, 1988). An inspiring vision will 
energize and encourage employees to work toward the 
organization's goals and values; it will evoke emotions 
and impel individuals to work toward the vision (Berlew, 
1974; Harrison, 1987). 
The vision, when institutionalized, becomes "owned'' 
by each individual in the organization. Each employee 
should know his or her role in the organization and be 
able to find a way to achieve the vision in accordance 
with this role. 
Vision Statements. The vision is often complex and 
elaborate, containing many details. To communicate the 
vision easily to employees, leaders often summarize their 
25 
aspirations into a succinct vision statement or slogan 
(Locke et al., 1991). For the purposes of this 
dissertation, vision will be referred to as the elaborate, 
detailed long-term standard of excellence, while the 
vision statement will be referred to as the summary, often 
in the form of a slogan, of the vision. When 
communicating the vision, the leader often presents the 
reasons for and importance of the vision, the 
circumstances causing him or her to develop the vision, 
and the rationale for the vision; this may be followed and 
summarized by the vision statement. 
Often, the vision statement is in the form of a 
metaphor or analogy. For example, cosmetics businesswoman 
Mary Kay Ash uses a bumble bee to represent the women who 
work for her {Conger, 1991, p. 39): 
[A] bee shouldn't be able to fly; its body is 
too heavy for its wings. But the bumblebee 
doesn't know this and it flies very well ... They 
(women) come to us not knowing they can fly. 
Finally, with help and encouragement, they find 
their wings -- and then they fly very well 
indeed. 
Leaders also use imagery and symbols to summarize their 
visions. Phil Turner, former head of Raychem's facilities 
division uses a hot air balloon to symbolize his vision of 
26 
uplifting people's spirits through beauty, cleanliness, 
and functionality, enthusiasm, good cheer, and excellence 
(Kouzes & Posner, 1987). Other examples of vision 
statements include the following (from Bradford & Cohen; 
Harrison, 1987; Kouzes & Posner, 1987): 
"We do not sell flowers, we sell beauty." 
"Keep the lights on!" 
"Provide service that brings guests back to a 
hotel." 
"Satisfy the customer every time!" 
"Fix it right the first time!" 
"News that people want to read in a weekly newspaper" 
"Universal service at all levels" [AT&T] 
"Value at a decent price" [Sears] 
Visions and Goals 
The above theories suggest that a vision affects the 
organization's goals as well as the individual's goals 
(Burns, 1978; House, 1977). A brief discussion of the 
distinction between these two concepts is warranted 
because the relationship between goals and charismatic 
leadership will be examined in more detail below. 
Visions and goals are similar in that both imply 
standards of performance, direct an individual's attention 
toward specific outcomes, and are future oriented. 
However, some differences exist. First, goals can be 
27 
different for each employee, but visions are intended to 
be common to all. Visions are developed by a leader, 
often with input from many employees. The vision should 
suggest a specific goal for each individual and give each 
employee guidance as to what areas of the job are to be 
given the most attention and what areas are to be given 
the least attention. 
Second, goals should be very specific, containing a 
certain quantity of performance to be attained in a 
specified, usually relatively short, period of time (Locke 
& Latham, 1990). An example is "Your goal is to assemble 
30 widgets in 15 minutes." At the end of the time period, 
it is clear whether the individual attained the goal or 
not. Goals can be assigned, such as by a supervisor, can 
be set participatively, or can be self-set. 
Visions, in contrast, are more general . They apply 
to groups, including organizations, departments, and work 
teams, although individuals can perform in accordance with 
a vision. As stated above, they are ideal, long-term 
ends. There is no specific time limit nor specific 
quantities of performance explicitly stated in the vision. 
For example, Steven Jobs, founder of Apple and NEXT 
computers, does not describe NEXT's vision as building x 
number of computers by a certain date. Instead, he talks 
28 
about NEXT's role in education and leading the software 
revolution on college campuses (Conger, 1991). 
Third, there are a limited number of ways to achieve 
a specific goal, but there usually are many ways an 
individual can go about working toward a vision. For 
example, there can be many ways for a NEXT employee to 
work toward leading the software revolution besides 
production, such as executive planning, conducting market 
research, purchasing needed components, conducting 
technical research and development, and performing 
maintenance on equipment. 
Due to the lack of research on vision, only untested 
prescriptions are available for describing the content of 
a vision. Bradford & Cohen (1984) state that overarching 
goals, which are similar to visions, should: (a) reflect 
the core purpose of the organization; (b) represent a 
feasible yet challenging goal; and (c) have larger 
significance for the organization. In addition, Sashkin 
(1988) proposes visions contain three themes: (a) Dealing 
with change. Especially relevant today is dealing with 
change in the environment, in markets, and in product 
technology. (b) Ideals. Effective visions specify ideal 
conditions or processes, not clearly defined final ends. 
Achieving the vision is a never-ending process. (c) People 
working together. As can be seen by the few examples 
29 
above, vision statements (and the visions they accompany) 
are widely used by leaders in organizations. Some visions 
have even outlasted the leader who first institutionalized 
them; an example is Johnson & Johnson's Credo (Bennis & 
Nanus, 1985). Based on a qualitative study of vision, 
Collins and Porras (1991) conceptualized vision as a 
general term that includes a guiding philosophy and a 
tangible image. The guiding philosophy is composed of the 
core beliefs and values that give rise to the 
organization's purpose. The tangible image is composed of 
the organization's mission which is a specific description 
of the desired end. 
To date, there has been no known empirical research 
on whether vision and vision statements have any effects 
at all or what attributes make up an effective vision. 
Qualitative research, however, has recognized the 
importance of the vision for many years. For example, 
Bennis and Nanus's (1985) five year study with 90 
effective leaders, using unstructured interviews, revealed 
that developing a vision was common to these leaders. 
Other qualitative work by Conger (1991), Kouzes and Posner 
(1987), Peters (1987), and Tichy and Devanna (1986) has 
also stressed the importance of the vision. To sum up, 
visions have been used by organizational leaders for at 
least several decades (Bennis & Nanus, 1985) but 
30 
organizational researchers have paid little attention to 
them. 
Vision Implementation: Task Strategies 
Although a vision can be highly motivating, the 
leader must take action to implement the vision (Locke et 
al., 1991; Sashkin, 1986, 1988). That is, employees may 
understand the vision, but may have difficulty making the 
transition from the long-term, general vision to the 
shorter-term, specific demands of their jobs. By itself, 
communicating task strategies to followers, is not 
considered leadership (Kotter, 1990). However, in 
conjunction with a vision, the task strategy represents 
one way to implement the vision. 
There are many ways to implement a vision, including 
business strategies and task strategies. A task strategy 
is method for doing the work or a plan for reaching a goal 
(Campbell, 1991; Locke & Latham, 1991). For example, on a 
complex problem solving task, one task strategy would be 
to systematically vary each parameter to figure out the 
solution in a given period of time. This dissertation 
will concentrate on disseminating information in the form 
of task strategies as one method of implementing the 
vision. 
31 
Bass (1985) states that transformational leaders use 
intellectual stimulation as a means of motivating 
employees. Although transformational or charismatic 
leaders typically do not give followers specific and exact 
directions for completing a task, leaders often recommend 
some general ways the follower might carry out the task. 
Leaders show employees how to think in new ways and to 
challenge assumptions when solving problems (Bass, 1985) 
but do not give employees the answers. Leaders serve as 
teachers and coaches to employees at all levels of the 
organization (Bass, 1985; conger, 1989; Peters, 1987). 
Although Bass (1985) does not explicitly discuss task 
strategies as a means of intellectually stimulating 
followers, it is clear that one way to implement the 
vision via intellectual stimulation is to suggest some 
appropriate task strategies. Training individuals to use 
and develop relevant task strategies has a direct positive 
impact on task performance as well as an indirect effect 
through self-efficacy and goals (Earley, 1985; Locke, 
Frederick, Lee, and Bobko, 1984). 
Similarly, a leader may communicate a business 
strategy for achieving the vision. Macro-strategies 
(i.e., strategic visions), or business strategies, are 
large-scale action plans for interacting with the 
environment (Jauch & Glueck, 1988) and are implemented at 
32 
the organizational level . However, it is not the macro-
level, but the individual level task strategy, that is 
relevant to this dissertation. 
Enthusiastic Communication Style 
The enthusiastic communication style displayed by 
charismatic leaders has received little attention in the 
leadership literature compared to vision. However, 
charismatic leadership research is beginning to pay more 
attention to the process aspect as well as the content 
aspect. Enthusiasm is conveyed through nonverbal 
behaviors such as eye contact, facial expressions, 
paralinguistic behaviors (stance, movement patterns, body 
position), voice tone, and interaction style (Edinger & 
Patterson, 1983; Howell & Frost, 1989). 
A charismatic leader is said to be enthusiastic 
(Conger, 1989). It is also likely that an enthusiastic 
individual is seen as charismatic, although these 
assumptions need to be tested. The two studies examined 
below provide some information about the nature of 
enthusiasm. 
Friedman et al. (1980) developed a self-report 
measure of expressiveness, or charisma, called an 
affective communication test (ACT). Examples of items 
from the ACT are the following (Friedman et al., 1980, p. 
33 
335): "When I hear good dance music , I can hardly keep 
still," "I often touch friends during conversations," and 
"I usually have neutral facial expressions" (reverse 
scored). They showed that the 13 - item (ACT) is a reliable 
and valid measure of individual differences in charisma. 
The ACT showed adequate convergent validity by correlating 
significantly with affiliation scales, dominance scales, 
and exhibition scales. It failed to correlate with harm 
avoidance, order, and play scales, thus showing adequate 
discriminant validity. 
Smith (1982, as described by House et al., 1988) 
examined the effects of thirty leaders who had reputations 
for being charismatic and thirty leaders who were 
effective but had reputations for being non-charismatic. 
Reputed charismatic leaders were described as being 
significantly more dynamic than reputed non-charismatic 
leaders. Subordinates of charismatic leaders reported 
working longer work weeks, experiencing more 
meaningfulness in their work, and having higher trust in 
their leaders than subordinates of non-charismatic 
leaders. 
Enthusiasm may be an important part of the 
communication process. For example, Conger (1991, p. 37) 
notes that Steven Jobs, founder of Apple and NEXT 
computers, takes an "aggressive stance behind the podium" 
34 
ibing Apple's competition. Con
ger (1989), in a 
when descr
ic leaders, 
study comparing charismatic to 
non-charismat
nd that charismatic leaders are
 effective and 
has fou
powerful speakers. When listening t
o charismatic leaders , 
ening, becoming 
followers reported enjoying list
 stimulated, and becoming emoti
onally 
intellectually
excited. In contrast, non-charisma
tic leaders were 
ut not dynamic . on 
described as monotone and inform
ative b
ther hand, Bennis and Nanus (19
85), in a qualitative 
the o
study of 90 leaders, concluded 
that a charismatic style 
essential to a leader's 
(i.e. personality trait) was no
t 
eness. So, although an enthusi
astic communication 
effectiv
style may be associated with ch
arismatic leadership, it 
unication style itself 
may be that an enthusiastic com
m
does not affect performance. 
p 
Mechanisms of Charismatic Leade
rshi
The next logical step after hav
ing identified three 
c leadership is to examine 
components comprising charismat
i
s of effects may 
the effects of these components
. Two type
ponents could have direct effect
s on 
occur. First, the com
followers' performance, attitude
s, and perceptions. 
fects on 
Second, the components could ha
ve indirect ef
erformance, attitudes, and perc
eptions. Due 
follower's p
 of these 
to the fact that both direct an
d indirect effects
35 
components have not been widely examined in the leadership 
research, both types of effects will be examined in this 
dissertation. Some possible mediating variables, or 
causal mechanisms, that may explain how a leader affects 
followers are examined next. 
The literature implies several mediating variables 
through which a charismatic leader's behavior could affect 
follower performance (for example, conger & Kanungo, 1988; 
House, 1977). That is, in addition to direct effects of 
leader behavior on follower performance, there may also be 
indirect causal mechanisms through which leader behavior 
affects followers' performance and attitudes. Examining 
such mediating relationships will provide insight into how 
leaders influence and motivate followers. 
One plausible set of mediating variables are goal-
setting variables (including goal level, goal commitment, 
and self-efficacy). Both goal level and self-efficacy 
have been shown to strongly affect performance variance 
(Bandura, 1986; Locke & Latham, 1990; Locke, 1991). 
According to Locke (1991), goals, self-efficacy, and 
performance constitute the motivation "hub." He argued 
that goals and self-efficacy are closest to action, thus 
implying that other variables must affect goals and self-
efficacy in order to affect performance (Locke, 1991). 
This has already been found with respect to feedback 
36 
(Locke & Latham, 1990). That is, feedback affects 
performance through goals and self-efficacy; goals and 
self-efficacy serve as mediators of the feedback-
performance relationship. Therefore, it makes sense to 
examine followers' goals and self-efficacy as mediators 
between leadership and followers' performance. These 
variables are examined below. 
House (1977) specifically posits that charismatic 
leaders communicate and model high self-set goals, which 
can result in followers' setting and achieving high goals 
(Conger & Kanungo, 1988). Further, charismatic leaders 
may influence goal commitment (House, 1977). Goal 
commitment is one's acceptance of the goal and is 
necessary for goals to affect performance (Locke & Latham, 
1990). House (1977) proposes that followers identify with 
and emulate the leader's values, goals, and behavior. 
Thus, when the leader c ommunicates a vision, it may be 
that followers' accept the vision and become committed to 
the vision as well as to the specific goals implied by the 
vision. 
House (1977) also argues that leaders have high 
expectations of followers. Leader communication of such 
expectations could result in followers feeling confident 
that they can perform the task well; this is referred to 
as the Galatea effect (Eden, 1990). The feeling of self-
37 
confidence on a specific task, or self-efficacy (Bandura, 
1986), may positively affect follower performance (Avolio 
& Gibbons, 1988). 
So, if through the vision, the leader communicates an 
ideal goal as well as the confidence that such a goal can 
be reached, this may affect followers' personal (self-set) 
goals and self-efficacy and in turn affect followers' 
performance. 
Conclusion 
Two main components of charismatic leadership are 
content (vision and vision implementation, in this case, 
task strategy) and process (enthusiastic communication 
style). Visions are becoming more and more commonly used 
by organizational leaders. Visions and vision statements 
can be found in large organizations, such as Apple, AT&T, 
Chrysler, General Electric, Johnson & Johnson, Mary Kay 
Cosmetics, McDonald's, NEXT, and in small organizations, 
such as local electric companies, departments of larger 
organizations, branch offices of banks, and even 
neighborhood grocery stores. An effective leader must 
take steps to implement or carry out the vision such as by 
suggesting suitable strategies. By having an enthusiastic 
communication style, the leader is thought to energize and 
enthuse others. These components may have direct effects 
38 
on followers' performance, attitudes, and perceptions as 
well as indirect effects, through goals and self- efficacy, 
on followers' performance. By closely examining these 
components in a controlled sett i ng we can isolate the 
effects of each component. 
39 
Chapter 4 
Research Propositions 
Based on the research in Chapters 2 and 3, 
propositions about the effects of each component on 
followers' performance, attitudes, and perceptions are 
presented as well as propositions regarding the mediating 
variables. 
Vision 
Many different types of visions can be developed, 
such as a vision for quality, customer service, or market 
dominance. Due to the importance of quality in today's 
global economy (Peters, 1987), this dissertation focused 
on the effects of a quality vision. 
Due to ?he fact that vision can be seen as a type ' of 
goal, it seems likely that a quality vision will inform 
followers to aim for high quality when performing their 
tasks, resulting in high performance quality. It is 
proposed that: 
Proposition 1: A quality vision communicated by 
the leader will result in higher performance 
quality than a leader who communicates no 
vision. 
40 
Charismatic leaders also are said to affect followers 
attitudes and perceptions about the task and about the 
leader. First, charismatic leaders have an impact on 
several types of attitudes. It is widely espoused that 
charismatic leaders affect followers' emotions or affect. 
Charismatic leaders inspire and uplift followers (Bass, 
1991; Conger, 1991; House, 1977) . Thus, the vision may 
uplift followers' mood state and increase followers' task 
satisfaction. A motivating and inspiring vision may carry 
over to affect follower motivation in the form of 
enthusiasm for the task and enthusiasm for the vision. 
Also, the literature posits that charismatic leaders 
compel followers to pursue the goals of the group and 
organization (Bass, 1985; House, 1977). In this 
dissertation, this was measured as the followers' 
willingness to work for the leader in the future for low 
pay or as a volunteer. This represented forgoing one's 
self-interest for the good of the company. 
second, charismatic leaders are perceived differently 
than non-charismatic leaders. Vision may result in 
increased congruence between followers' and leaders' 
beliefs and values (House, 1977; McClelland, 1985; 
Sashkin, 1988). Trust in the leader may be an important 
outcome of a communicating a vision (House, 1977). 
Similarly, followers may like a leader who communicates a 
41 
vision more than a leader who does not communicate a 
vision. The vision is also an important component in 
followers perceiving the leader as charismatic (Bass, 
1985; House, 1977). Bass (1985) states that 
transformational leadership is associated with increasing 
followers' intellectual stimulation and inspiration. 
Charismatic leaders may also increase task clarity for the 
followers. 
Based on the above theories, the impact of 
charismatic leaders are proposed to affect follower 
attitudes and perceptions in the following manner: 
Proposition 2: Communication of a vision, compared to 
no vision, by the leader will result in positive 
effects on each follower attitude and perception 
(namely, task satisfaction, mood state, enthusiasm 
for task, enthusiasm for vision, willingness to work, 
congruence of beliefs, trust in the leader, liking 
the leader, charisma, intellectual stimulation, 
inspiration, and task clarity). 
Although a global, multivariate test could be 
conducted this dissertation focuses on the effects of 
' 
each separate dependent variable. Thus, the proposition 
will be tested using univariate methods. 
42 
Vision Implementation: Task Strategies 
As one way of implementing the vision, the leader can 
disseminate information in the form of a task strategy. A 
task strategy would give specific information to followers 
regarding how to attain the vision in practice. Assuming 
that the task strategy is suitable to the task, the 
strategy should positively affect performance. It is 
proposed that: 
Proposition 3: The task strategy presented by the 
leader will have positive effects on performance 
quality and/or performance quantity, as compared to 
no task strategy. 
By itself, task strategy is not considered leadership 
unless it occurs in combination with a vision. Simply 
providing task strategy information by itself should not 
affect followers' attitudes or perceptions that are 
related to leadership, except for intellectual 
stimulation. A significant effect of task strategy on 
intellectual stimulation would provide a manipulation 
check that the task strategies were, in fact, suitable to 
the task. Thus, no main effects on follower attitudes and 
perceptions are predicted. 
However, the interaction of task strategy and vision 
could affect follower attitudes and perceptions because 
the task strategy could supplement the vision. This 
43 
proposition is exploratory and tests the possibility that 
charisma cannot be decomposed but rather is only perceived 
as a whole, or gestalt, as compared to separate elements. 
It is proposed that: 
Proposition 4: A leader who communicates a 
vision and communicates a way to implement the 
vision will have a disproportionately positive 
effect (i.e., a significant interaction) on each 
follower attitude and perception (namely, task 
satisfaction, mood state, enthusiasm for task, 
enthusiasm for vision, willingness to work, 
congruence of beliefs, trust in the leader, 
liking the leader, charisma, intellectual 
stimulation, inspiration, and task clarity) as 
compared to a leader who only communicates a 
vision or only displays enthusiasm. 
Enthusiastic Communication Style 
Leaders can have different but equally effective 
content and yet have different styles of communicating the 
vision. No proposition is made regarding the effects of 
enthusiasm on performance because, unlike content, 
enthusiasm does not give followers any specific standards 
or suggestions regarding how to perform. Regardless of 
vision content, an enthusiastic leader may affect 
44 
followers' attitudes and perceptions more than a non-
enthusiastic leader. That is, the leader's activity 
level, which is used to communicate the leader's own high 
motivation, may become contagious (Conger, 1988). 
I t is proposed that: 
Proposition 5: A leader who displays enthusiastic 
behavior will more positively affect each attitude 
and perception (namely, task satisfaction, mood 
state, enthusiasm for task, enthusiasm for vision, 
willingness to work, congruence of beliefs, trust in 
the leader, liking the leader, charisma, intellectual 
stimulation, inspiration, and task clarity) than a 
leader who displays unenthusiastic behavior. 
When enthusiasm is paired with a compelling vision, 
the effects on follower attitudes may be even greater than 
a vision alone or enthusiasm alone. This could occur 
because a high enthusiasm level may show that the leader 
is enthused by the vision which may then get the followers 
excited about and committed to the vision. Again, this 
proposition is exploratory and tests the possibility that 
charisma is perceived as a whole rather than separate 
components. Thus, an interaction between vision and 
enthusiasm is proposed that: 
Proposition 6: A leader who communicates a 
vision and displays highly enthusiastic behavior 
45 
will have a disproportionately positive effect 
(i.e., a significant interaction) on each 
follower attitude and perception (namely, task 
satisfaction, mood state, enthusiasm for task, 
enthusiasm for vision, willingness to work, 
congruence of beliefs, trust in the leader, 
liking the leader, charisma, intellectual 
stimulation, inspiration, and task clarity) as 
compared to a leader who only communicates a 
vision or only displays enthusiasm. 
Mediating Mechanisms: Goals and Self-efficacy 
Goal-related variables (personal goal, anticipated 
satisfaction, goal commitment, and self-efficacy) may 
mediate the leadership-performance relationship. 
Specifically, it is expected that goal-related variables 
will mediate between vision and task strategies (but not 
enthusiasm), and follower performance. Goal-related 
variables are not expected to mediate the enthusiasm-
performance relationship because neither high nor low 
enthusiasm communicate specific information to followers 
regarding the task. Accordingly, it is proposed that: 
Proposition 7: Goal-related variables (namely, 
personal goal, goal commitment, and self-efficacy) 
46 
will serve as a mediator between vision and follower 
performance. 
Proposition 8: Goal-related variables (namely, 
personal goal, goal commitment, and self-efficacy) 
will serve as a mediator between task strategy and 
follower performance. 
The study design and method are presented in the next 
chapter. 
47 
Chapter 5 
Design and Method 
Design 
This dissertation empirically examined the effects of 
2 levels of vision (no vision and vision), 2 levels of 
task strategy (no task strategy and task strategy), and 2 
levels of enthusiasm (low and high) on performance, 
attitudes, and perceptions in a completely crossed, 
between subjects design in a laboratory setting using a 
simulated task. The vision variable and the task strategy 
variable represent two aspects of the content of 
charismatic leadership. Enthusiasm represents the process 
aspect of charismatic leadership. The design is shown in 
Figure 1. 
One-half of the subjects within each cell were 
exposed to the male leader, and one-half were exposed to 
the female leader. Each leader conducted two sessions, or 
"runs", of each cell in order to examine the consistency 
with which the manipulations were carried out. No 
differences between the male and female leader were 
expected, and no differences between session 1 and session 
2 were expected. These assumptions were tested and 
results are presented in chapter 6. 
48 
Figure 1 
Research Design 
Content Variables 
No Vision Vision 
Process No Task Task No Task Task 
Variable Strategy strategy Strategy Strategy 
Low 
Enthusiasm 
High 
Enthusiasm 
49 
The laboratory setting was considered justifiable for 
two reasons. First, in practice, charismatic leaders 
exhibit both content and process components. In a field 
setting, it would be virtually impossible to observe an 
equal proportion of leaders for each experimental cell. 
Second, previous research has shown that it is possible to 
successfully manipulate charismatic leadership in a 
laboratory situation (Howell & Frost, 1989). Although 
trusting the leader and commitment to the vision must be 
built up over time in an organization, it has been shown 
that individuals are able to immediately perceive a 
leader's charisma and observe charismatic behaviors 
(Friedman et al., 1980; Hater & Bass, 1988; Howell & 
Frost, 1989). 
Subjects 
Students in two third-year business classes 
participated in the study in exchange for a class bonus. 
A total of 282 subjects participated. Fifty-two percent 
of the subjects were male and 48% were female. Fifty-nine 
percent were currently employed full- or part-time. 
Ninety-eight percent had been previously employed. Thus, 
it was assumed that the subjects had enough work 
experience to understand the leader and the ?content of the 
manipulations. 
50 
Task 
The task was a simulated production task. Subjects 
inserted pages into sections of notebook binders according 
to task instructions. For realism, the task was derived 
from an actual binder assembly task done in printing 
companies and was adapted for use in a laboratory setting. 
The task had a quantity (number of pages inserted) and a 
quality component (number of errors). For practical 
reasons, three binders were used: binder 1 for the 
practice session, binder 2 for trial 1, and binder 3 for 
trial 2. 
A pilot test with 12 subjects from the above business 
classes revealed that subjects found the task difficult 
but not impossible. For a 10 minute session, subjects had 
a mean of 5.0 errors and inserted a mean of 16.3 pages 
across the three binders. There was considerable 
variability in performance, with errors ranging ?rom o to 
22 (s.d.=6.0) and pages ranging from Oto 42 (s.d.=9.9). 
The difficulty level of each binder was approximately 
equal. The mean number of errors (and respective standard 
deviations) for binder 1 was 4.1 (s.d.=5.8), for binder 2 
was 6.0 (s.d.=7.1), and for binder 3 was 2.3 (s.d.=2.1). 
The mean number of pages for binder one was 9.4 
(s.d. =5.7), for binder 2 was 22.5 (s.d.=5.3), and for 
binder 3 was 27.7 (s.d.=18.3). (Note: Due to an unequal 
51 
number of subjects across binders, the mean number of 
errors and pages presented here are weighted means) ? 
Post-experimental interviews with the subjects confirmed 
that they found the task difficult . Although their 
performance increased as they went through the binders, 
indicating a practice effect, there was considerable 
variability in performance for all three binders. 
In addition, post-experimental interviews revealed 
the types of work methods the subjects used and which of 
these work methods were more or less effective. This 
information was used to develop the task strategy 
manipulation. The pilot test confirmed that the most 
effective way to perform the task was to write out the 
order of the pages before inserting them. The following 
additional work methods were also determined to be 
effective: (a) using adjacent desks to place the stack of 
pages and binders or somehow organizing the materials so 
they are easy to reach, (b) double checking the order of 
the pages before inserting them, (c) reading the 
instructions carefully, (d) paying close attention to the 
titles of the pages due to the similarity in many of the 
titles (e.g. "equations and inequalities" and "systems of 
equations and inequalities"), and (e) collecting all the 
pages needed for a section before inserting them. Two 
task strategies that subjects thought were effective were 
I 
52 
in fact, not effective at all. Spreading out the pages on
 
the floor and taking the pages out of alphabetical order 
led to more errors and fewer pages inserted than not using
 
these strategies. Thus, performance was a function of 
proper work organization as well as a function of 
motivation due to the considerable effort and attention 
required to assemble the binders correctly (Locke & 
Latham, 1990). 
Procedure 
Experimental runs were conducted in groups of 4 to 12 
Per session by condition. The order of the runs was 
determined randomly. 
Each subject performed the task on an individual 
basis with work areas separated as much as possible to 
Prevent communication between the subjects. 
Sessions lasted about 2 hours. Introductory 
instructions and consent form completion was approximately
 
lo minutes in length, the practice trial was 10 minutes, 
the trials were 30 minutes total, scoring was 20 minutes 
total, questionnaire completion was 30 minutes total, and 
the manipulations were 8-9 minutes. The vision and no 
Vision manipulations were the same length (8 minutes), but
 
the task strategy manipulation was approximately 1 minute 
longer than the no task strategy manipulation. 
53 
iving for the experiment, e
ach subject 
Upon arr
orm. Then, the experimente
r read a 
completed a consent f
at 
rief set of instructions and
 cover story, explaining th
b
working with 
the College of Business and
 Management is 
y 
BKc, a printing company in Col
umbia, Maryland, to stud
e told that the leader 
"task design." The subjects wer
til then, they would be 
would be arriving shortly a
nd un
ing on a practice binder. T
hey then performed a 10 
Work
 
s the same difficulty as the
minute practice trial which
 wa
actice trial 
other trials. By purposefu
lly making the pr
possible to examine the effe
cts of the 
difficult, it was 
leader "uplifting" the subj
ects' mood. 
tice binder, the 
After subjects scored the p
rac
 
erimenter called the leader 
into the room. From this
exp
 experimental session, gave e
Point on, the leader ran th
 
 directions to subjects, an
d the experimenter remained
a11
xperiment (i.e. distributing
 
to facilitate running the e
materials). The leader briefl
y gave some background 
rmed according to 
information on the company 
and then perfo
? t 
ript and the appropria e man
.i pu l ati?o n. 
th e appropriate sc
There were two trials of 15
 minutes each. 
corded in order 
Performance for each of two
 trials was re
he effects of learning throu
ghout the 
to analyze t
nce in one 
experiment. rt may be that s
ubjects' performa
54 
ter rate than for those 
in 
condition increases at 
a fas
another condition. 
s recorded the number of
 
After each trial, subje
ct
 and were given a scorin
g key to record 
Pages inserted
d 
 errors for that trial. 
Subjects were debriefe
Pages and
ection process was comp
leted. 
after the entire data c
oll
cored 
sure consistency in scor
ing, the experimenter s
To en
bjects were dismissed fro
m the 
each subject's work afte
r su
used for data 
experiment. The experimen
ter's scores were 
analysis purposes. 
Actor Selection 
ssional male actor and .o
ne professional 
One profe
 all manipulations by 
female actor were hired 
to carry out
ited through 
Portraying the leader. 
The actors were recru
ce. The male actor was 
a local professional ac
tor's servi
 selected 
selected from 16 inquir
ies, and the female was
quiries about the positi
on. The basis for 
from 21 in
 
g the actors was the fol
lowing: (a) ability, as
choosin
, 
ion and reading of the s
cript
determined through an a
udit
enthusiasm behaviors and
 
to portray both the high
 and low 
 and times. The actors 
(b) availability for des
ired days
ractiveness 
chosen were similar in a
ge (39 and 43) and att
y the author)? 
(as subjectively perceiv
ed b
 actors had previous lea
dership 
In addition, both
nder of a local 
experience. The male a
ctor was the fou
55 
theater which had produced hundreds of productions over 
the past 10 years. He also had some work experience in a 
printing company. The female actor founded and operated 
her own antique store. 
To reduce the possibility of contamination if 
subjects told each other about the experiment, the male 
actor played the role of CEO and the female actor played 
the role of President of the company. Howell and Frost 
(l989) used two female actors with similar physical 
features and found no differences between them, but this 
dissertation varied the leader gender while choosing the 
actors of similar attractiveness a~d age. No differences 
between the actors were predicted. 
Actor Training 
Approximately 30 hours of training was given to the 
In addition, the actors 
actors over a 5-week period. 
spent their own time memorizing and rehearsing the 
scripts. About 30% of the training time covered the 
following topics: (a) an overview of the research process 
emphasizing the importance of experimental control; (b) an 
explanation of the study design and variables without 
(c) positive and 
ind?i cation of hypotheses or expectati?o ns; 
negative examples of charismatic leaders via articles and 
v?l deotapes; (d) review of possible questi.o ns that subjects 
may ask and review of appropriate responses. The 
56 
remaining 70% of the training was spent rehearsing the 
scripts and behaviors and providing feedback to the 
actors. 
After all experimental sessions were completed, the 
actors were debriefed and given the results of the 
experiment. 
Manipulations 
Vision Manipulation 
The vision manipulations were longer and more 
detailed than the vision expressed by the leader in Howell 
a nd Frost's (1989) charismatic leadership manipulation. 
The vision manipulations of the present study were 
Conger's 
expected to be more realistic and informative. 
(l 9 S9) research found that actual charismatic leaders 
express similar types of visions, including describing the 
way the vision came about, its importance, key examples of 
employees acting in accordance with the vision. 
The leader told two stories that illustrated 
Vision. 
the vi? sion and its importance. These stori? es were 
examples of how the employees have acted in accordance 
With the vision, showing their commitment to the vision. 
The leader then read a formal vision, ending with the 
Vision statement (BKC is bound to quality!), which 
Finally, to arou
s st er  e f ollower motives to ssed quality. 
57 
perform well, the leader told the subjects that their 
performance would later be compared to that of two 
competing printing and binding companies. The leader then 
inspired them and increased their self-confidence by 
saying that although the task is difficult, he/she really 
thinks that they will do a super job. Each aspect of the 
vision manipulation is an important part of a charismatic 
leader's content (House, 1977). This script is shown in 
Appendix A. 
No Vision. No vision was communicated to the 
ii 
"I  
subjects. To allow the leader the opportunity to display ' 
the appropriate enthusiastic behaviors and to equalize the 
t'im e the leader talked to the subjects, the leader 
presented factual information about the process of making 
paper. This was so that the leader would appear credible 
without communicating a vision. This script is shown in 
Appendix B. 
T_ask Strategy Manipulation 
Task strategy. The leader suggested some possible 
task strategies for achieving the vision. He/she told the 
SUbJ' ects that employees in the trai?n i.n g d epar t men t h ave 
developed some techniques for doing the task. The leader 
then gave suggestions about which methods (as presented 
above) were effective and which were ineffective. This 
58 
nformation was in accordance with the visi
on. This 
i
manipulation is shown in Appendix c. 
No Task Strategy. No task strategy or wor
k methods 
were suggested by the leader. 
Enthusiasm Manipulation 
siasm manipulations were similar to those The enthu
used by Howell and Frost (1989) in their c
harismatic 
leadership condition. Also, the enthusiasm
 components 
 here (paralinguistic and nonverbal behavi
ors) 
manipulated
have been identified by conger (1989) as e
xhibited by 
actual charismatic leaders. 
High Enthusiasm. The leader showed enthus
iasm toward 
the task and the subjects' performance but
 did not show 
differential attention or approval to indi
vidual subjects. 
He/she displayed nonverbal charismatic beh
aviors, 
ence. 
including a powerful, confident, and dynam
ic pres
Upon entering the room, the leader shook h
ands with the 
 with the subjects. As in the Howell and experimenter and
Frost study, the leader alternated between
 pacing around 
the front of the room (he/she did not pace
 up and down the 
, as subjects could have perceived this as
 a type of 
aisles
attention or reward) and leaning forward a
nd sitting up 
de eye 
straight in the chair when sitting. The l
eader ma
contact with the subjects and used hand ge
stures when 
speaking. The leader had animated facial 
expressions when 
59 
speaking. The leader spoke with a c
aptivating and 
f 
engaging voice tone, alternated the 
pace and loudness o
used pauses to emphasize points. Th
e leader 
speaking, and 
continued to display these behaviors
 throughout the 
experiment. 
nthusiasm. In this condition, the lead
er did 
Low E
lay any of the above enthusiastic be
haviors, but 
not disp
simply carried out his/her role with
 neutrality toward the 
hen entering the 
task and the subjects' performance. 
W
room, the leader simply greeted the 
experimenter but did 
not shake hands with the experiment
er or subjects. While 
 simply sat relaxed 
appearing friendly and polite, he/sh
e
in the chair behind the desk, did no
t pace around the 
ow interest in the subjects' work on
 
room, and did not sh
the task. The leader spoke clearly 
but in a monotone 
Voice and spoke with a constant pace
 and medium voice 
tone. His/her presence was uninvolv
ed, including 
ent and appearing 
remaining seated for most of the exp
erim
relaxed in the chair when sitting. 
The leader continued 
t. 
to portray these behaviors throughou
t the experimen
Realism of Manipulations 
Realistic manipulations and control 
over the 
in several 
experimental setting was built into 
the study 
First, professional actors were used
 to reduced any 
ways. 
oss 
random inconsistencies in behavior t
hat may occur acr
60 
ns. To reduce experimente
r expectancy 
experimental sessio
study's hypotheses 
effects, the actors were b
lind to the 
y 
enthal & Rosnow, 1969), alt
hough it is possible the
(Ros
may have guessed the hypo
theses . 
pulations via 
Second, rather than presen
t the mani
ve. 
V1' d eotape, the actors carried
 out the manipulations li
ncrease the power of the e
ffects as well as 
This was to i
increase realism. 
pulations were framed aroun
d an actual 
Third, the mani
sk and manipulations were 
PUblishing company. The t
a
e 
developed in consultation 
with Mr. Bart Kinlein, th
 of the actual company . P
articipants were told 
founder
th  BKC Publishing, was worki
ng with 
at the company,
study "task 
researchers at the univer
sity of Maryland . to 
design." 
ximately 30 hours of traini
ng sessions 
Fourth, ? appro
 the appropriate 
increased the actors' abil
ity to exhibit
ma ? s. All actions and commu
nications during the 
nipulation
experiment were specified 
in scripts. 
Measures 
ry mediating variables, 
Dependent variables, explo
rato
ation, and 
manipulation checks, gene
ral background inform
to leader were measured as
 indicated 
5 Ubjects' questions 
shown in Figure 2. 
below. The summary of the
 variables is 
61 
Qependent Variables 
g the practice session 
Performance. Performan
ce durin
 as a 
served as a measure of 
ability, which was used
ble during the data ana
lysis. Quality and 
control varia
 of 
tity of performance were
 measured as the number
quan
ages, respectively. (Note: 
A 
f p
errors and the number o
o examined, but 
ratio variable of error
s per page was als
 correlated, K=.90, ?<.0
01, with performance 
Was highly
 same results as perform
ance 
quality and thus yielde
d the
quality.) I ! 
I' ' 
ariables 
I 
Attitude and Perception
 V
erception variables were
 
The following attitude 
and p
measured via questionna
ires: 
was measured using a moo
d 
(a) The subjects' mood 
e Table 1 for the compl
ete list of 
adjective list (se
). Subjects indicated 
adjectives) (Judge & Hu
lin, in press
th hich they felt each of v
arious emotions 
e extent to w
iment using a 5-point sc
ale (1 = very 
during the exper
ry much). To assess the
 
s1 l? ghtly or not at all to
 5 =ve
ood, 
to which the leader can
 change the subjects' m
degree 
checklist was completed 
directly after 
a mood adjective 
der manipulations) and the Practice session (bef
ore the lea
een used 
after the second sessio
n. This measure has b
ood (Judge & Hulin, in p
ress) and 
Previously to measure m
 reliability (g = .92).  consistency
has shown good internal
62 
Figure 2 
summary of variables 
INDEPENDENT 
I. Vision (content) 
II Task strategy (content) 
II. Enthusiasm (process) 
DEPENDENT 
I. Performance 
A. Quality 
B. Quantity .11 ' ''  
' 
II. Attitudes 
A. Mood State 
1. Anxiety 
2. Energy 
B. Task Satisfaction 
C. Enthusiasm for Task 
D. Enthusiasm for Vision 
E. Willingness to work 
Irr. Perceptions 
A ? c ongruence o f Beli' efs and Values 
B. Trust in Leader 
C. Liking the Leader 
D. Charisma 
E. Intellectual stimulation 
6 3 
(Figure 2 , cont . ) 
F. I ns p i r ation 
G. Tas k Clari ty 
MEDI ATING (exp l oratQry) 
I . Goa ls 
A. Goa l L~v~l (personal) 
B. Anticipat ed satisfaction 
c. Goal Commi t ment 
. Self - e ffi cac y 
OTHER MEASURES 
I. Mani _ul ation Check 
A. Vi s i on 
B. T s k trate gy 
C. Enthus ias m 
II . G_eneral Back roun~ 
A . Age 
B. Gender 
C. Work e xper i ence 
III. Qualitative Data 
A. Subjects ' Questions Our ' ng Experiment 
B. Subj e cts; Comments 
C. Expe rimenter 's Observations 
D. Actors ' comments 
64 
ts a nalys i s o f the scale 
using var i max 
Princip al componen
mponents , anxiety and 
r otation revealed t wo pr
'ncip a l co
he r e s ults of the princip
al components a nalysis 
e n e r gy . T
Thus, two mood scale scor
es were 
is shown in Table 1. 
 
computed , on e for e nergy 
and one f or a nxiety .
nd c haracteristic effects
 or social 
To avoid a ny d e ma
irab i l'ty effects , t h e fo
llowing attitude a nd 
des
sing a 7- point scale (1 ~
 
d , u
Perception items were me
asure
nd 
gly disagree , 7 = s t r ongl y a g r e e
), after the seco
stron
ssion on l y . Th e c omp lete
 list of attitude a nd 
se
i n Appe ndix D (except for
 
Perception items are pres
en ted 
Table 1 ) . An example 
the mood items , which are
 s hown i n 
le a nd the i nternal c onsis
tency 
item from each sca
re l 'ab' l ' t y are presented
 below. 
on (e. g . " I n ge nera l , I wa s (b) Task satisfacti
satisfied whe n doing t h e 
task ") . Q. == . 92 . 
his task 
(c ) Enthusiasm for task (e
. g . "I t hought t
r eally fun") . g = . 88 . 
as t ic 
(d) En t husiasm for v ision
 (e.g . " I was e nthusi
ut assembling high qualit
y b i nder s" ) . Q = . 75. 
abo
f ut ure for t h e leader 
( e ) Wi l ling ness to wor k i
n the 
 p a y (e . g. " Ev e n 
lid r c on d ' t'on of below 
market pay or no
e l O% below averag e , I wo
uld like to work 
if t h e p ay wer
") ? Q = ?69 ? 
for the [ leader ] in a f u t
ure job
65 
Table 1 
Principal Components Analysis of Mood sca l es 
post-experiment 
Mood Pre- experiment 
EnergY Anxiety Energy Adj_ective Anxiety 
.76 .03 
Distressed ..e...11,_ .02 
- .54 = . 10 
At e s t - . 47 - . 31 
. 68 .09 
Fearful .69 .07 
.73 - .16 
Host i l e .67 - . 16 
- .32 -._69 
- .12 
Calm - . 65 
.67 . 11 
Jittery .62 .19 
..:.1,2 . 17 
Nervous . 69 .09 
- . 67 - .03 
Relaxed - . 69 - .21 
.13 .! 6,i 
-. 20 
Scornful . 5 6 
. 23 ~56 
Activ~ .28 .:. 51 
. 12 - . 69 
Drowsy .09 - .!.-7 6 
.49 
Elated . 23 . 3,1. 
.08 
.06 - .6 6 
D_ u 1 - . 0 4 - .58 
- . 73 
Sl ee y .00 - ? 7,1., 
. 04 
- .2 5 - .3 1 
Pl acid - .28 - .!.-3_2. 
. 06 . 68 
Enthusiast ,09 .i 5c 2  
.17 - . 68 
Sl uggish - .70 ? 1 7 
.2 7 .l5 8 
Peppy .50 . 18 
- . 02 . 39 
Strong ..!,...2_Q .01 
66 
j ects' belie f s a nd va lue
s 
Congru e nce betwe en s u
b
(f ) 
n (e . g. ''Qua lity 
ahd those conunun ? c ate
d through the vis io
. g = . 60 . 
s the most important 
aspect o f t h i s job" )
i
r (e . g . " I have compl
ete trust 
(g ) Trust in t h e l ead
e
i n the [ l e a d e r] ") . . 84
. 
 (e.g. " Th is [lea d e r] 
is a 
r
(h ) Liki n o f the l ead
e
ally likeable person "
) . _g = . 84. 
re
ed to determi ne wh i ch 
(i ) Two charisma sca l e
s were us
 , lated to the s ubjects
' p ercept i ons o f t he 
V
a r iables are re
s cale was from r elevan
t 
t 
leader ' cha risma. T
he firs
s from the 
from Bass ' (1 985 ) MLQ 
- Form st . Two item
items 
i rrelevant and were omit
te d. 
eme d 
For m BY s ca l e we re d e
sion of th i s scale , th
e 
'l'h cond wa s a mor e rece n
t ver
s e
 - Form SR , which is p ub
lished by consu t i ng 
MLQ
m 8Y , a = . 92 and f or For For
Psychologists Pr e ss, 
I n c . 
Form SR, Q = . 92 . 
rre la t e d 
Th e t wo c h ar i sma scal
es were h i gh l y co
tte r n of relations hips
 
he same pa
(r ~ .83 , < .0005 ) a nd 
had t
bles of inte rest . Thu
s , the Form BY 
With the other var i a
sca l e was used f o r al
l ana l yses. 
n (e . g. ''The [ lea d e r ] 
( j ) I nte llectua l stimu
latio
binder 
e d me with new ide as 
about a s semb i ng the 
Provid
secti ons " ). Q = .8 7 . 
 [leader ] heightened my "The
(k) Inspirat i o n (e . g . 
lno t . Q = .95 . l?. Va t i on t o s u cceed" ) . 
67 
do on 
ask clarity (e.g. "I k
new exactly what to 
(1) T
this task"). g = .86. 
Mediating Variables 
ing variables were me
asured 
The exploratory media
t
e manipulations by the
 
before each trial but
 after th
bjects were 
r had been given. Du
e to the fact that su
leade
s against which to jud
ge their goals 
not given any standar
d
o not 
formance and the fact
 that these measure d
or per
mpleting these measur
es 
 leader, co
involve evaluating th
e
to yield any reactive 
before each trial was
 not expected 
Appendix E presents th
e goal items. 
effects. 
isfaction, goal comm
itment, 
The goal, anticipated
 sat
 were combined into sc
ales 
and self-efficacy mea
sures
nt pattern between tri
al 1-
cause they showed a c
onsiste
be
mance 
Performance correlatio
n and trial 2-perfor
s that 
tion .for each goal me
asure. This indicate
correla
did not change across lationship 
the goal-performance 
re
d high 
ls. Also, the goal c
ommitment items showe
tria
ere used 
 consistency reliabil
ity. These scales w
internal
 The following items
 
ysis.
for purposes of all d
ata anal
uantity before 
ere completed with res
pect to quality and q
W
ch of the two trials:
 
ea
-set) goal level. The
 quality goal 
(a) Personal (self
er of 
easured by the item "W
hat is the most numb
Was m
 to make, that is, 
errors that you would
 find acceptable
68 
Your minimum goal for quality?" Quantity goal was 
measured by the item "What is the least number of pages 
that you would find acceptable to insert, that is, your 
minimum goal for quantity?" The two personal goal 
measures were significantly correlated across the two 
trials (K=.60, 2<.0l for quality and K=.44, Q<.01 for 
quantity), and thus were combined across trials into 
personal goal for quality and personal goal for quantity 
scales. 
(b) Anticipated satisfaction or valence is a related, 
Yet independent measure of goal level (Locke & Latham, 
19 90). Anticipated satisfaction was measured by having 
individuals indicate the degree to which they "anticipated 
being satisfied with each level of performance with 
respect to errors" (on a 9-point scale, performance levels 
ranging from o to 20 errors in units of 2) and "with 
respect to pages" (on a 9-point scale, performance levels 
ranging from o to 50 pages in units of 5). The measure of 
anticipated satisfaction has been shown to have construct 
Validity and is negatively related to goal level (Mento, 
Locke, & Klein, in press). Anticipated satisfaction for 
quality and for quantity were combined across trials 
because they were significantly correlated across the two 
trials (K=. 56 , Q< .0l for quality and K=.45, p < .01 f or 
quantity). 
69 
itment for quality an
d quantity was 
(c) Goal comm
s to rate five items 
on a 7-
red by asking individ
ual
measu
uality 
e (i.e., "It is quit
e likely that this q
Point scal
 strongly committed 
to 
goal may need to be 
revised,'' "I am
quantity goal"). Th
ese items have been 
pursuing this 
dity (Hollenbeck, Kl
ein, 
vali
shown to have constr
uct 
These items showed hi
gh 
8). 
O'Leary, & Wright, 1
98
= g y for quality (g .85) a
nd quantity (
internal consistenc
lated across the two tly corre
= an-88) and were signific
 K=.78, Q<.01 for 
trials (K=.72, Q<.01
 for quality and
items for quality and
 for 
 
quantity). Goal com
mitment
ales. 
quantity were each c
ombined into sc
ength and self-effica
cy 
(d) Self-efficacy str
elated to performance
 
-m=a 
n to be r
-gn?i tude have been sh
ow
trength was 
Locke & Latham, 1990
). self-efficacy s
(
10 scale) 
by asking subjects to
 indicate (on 1 to 
measured 
umber of errors, from
 o 
th g each neir confidence in ma
kin
 or each number 
 20 (in units of 2 e
rrors), for quality,
to
(in units of 5 pages)
, for 
of Pages, from o to 50
 
 was measured by aski
ng 
agnitude
quantity. Self-effic
acy m
hey thought they 
bjects to indicate (
yes or no) whether t
su
centage of errors, fr
om Oto 20 (in 
could correct that p
er
pages, 
 errors), for quality
, or each number of 
Units of 2
r quantity. Both of ges), fo
from o to 50 (in units 
of 5 pa
& Latham (1990). 
these measures are re
commended by Locke 
70 
 reliability for the 
self-
Internal consistency
cales respectively w
as 
efficacy for quality
 and quantity s
be noted that due to 
the 
= . 81 and Q = .80. It shoul
d 
Q 
trength 
t correlation between
 the self-efficacy s
significan
s (mean ?=.71, Q<.01
), 
gnitude scale
and self- efficacy ma
gle 
th re co
mbined into a sin
e z - scores for these
 scales we
correlation between t
hese 
-efficacy scale . Hi
gh 
Self
, 1990). 
measures is common (
Locke & Latham
-M=a .D? lpulation Checks 
ks determined the ext
ent to which 
Manipulation chec
The 
ended manipulations. 
Subjects' perceived 
the int
ed into scales based re combin
manipulation check i
tems we
he coefficient alpha d because t
on an a priori basis
 an
ity. 
howed high internal 
consistency reliabil
measure s
after the second 
These were completed
 on a 5-point scale 
session: 
d the extent to which
 the 
Vision items assesse(a) 
ting BKC's vision 
s Ubjects perceived th
e leader communica
 [lea d er J prov1
.  d e d a vi. si?o n 
and 1 eong-term goals (e.g. 
"Th
s all about"). Q = .B 9 . of What BKC wa
y items assessed the 
extent to which 
(b) Task strateg
 by 
th n tip
s and suggestions
e subjects felt they
 were give
 the task (e.g. "The 
[leader] 
 perform
the leader on how to 9
ing the task"). g_ = . 2 . 
gave specific tips f
or perform
71 
(c) Enthusiasm items assessed the extent to which 
subjects perceived the leader as enthusiastic in general 
and the extent to which the leader displayed enthusiastic 
behaviors (e.g., pacing, eye contact, voice) (e.g "The 
[leader] was very enthusiastic when talking," "The 
[leader] paced around the room for a good part of the 
experiment").~ = . 85. 
General Background Information 
These items were completed after the second session 
and included the subject's age, gender, and whether the 
subject has been employed or is currently employed. 
Qualitative Data 
Qualitative data from several sources were collected 
in order to gain insight into the results of the 
experiment. The number and content of subjects' questions 
to the leader were written down by the experimenter 
immediately following each session. Rather than denying 
subJ? ects the opportunity to ask the lea d er questi?o ns, the 
effects of each condition were allowed to emerge 
For example, highly enthusiastic leaders, 
naturally. 
through their nonverbal behaviors, may invite more 
Due to the 
questions than a less enthusiastic leader. 
fact that it may be impossible to stop the subjects from 
asking questions, the leader in each condition asked the 
subjects if they had any questions. BY recording what was 
72 
asked, it was possible to qualitatively analyze the 
effects of the process and content on subjects' 
Willingness to interact with the leader. As indicated 
above, prepared responses to the questions were given by 
the leader when possible. 
Also, subjects were asked to write comments at the 
end of their experiment regarding their impressions of the
 
experiment and the leader and regarding whether the leader
 
affected their performance or attitudes toward the leader.
 
The experimenter's observations of the leader's 
effects on the subjects and the general atmosphere of the 
experimental sessions was expected to be helpful when 
interpreting the quantitative results. Finally, the 
actors' impressions of the training sessions and 
experimental sessions was expected to yield similar 
insight. 
73 
Chapter 6 
Results: Pilot Ratings, Manipulation Checks, 
Assumptions, and Independent Variable Effects 
Pilot Ratings of Actors 
Thirty-nine independent observers, blind to the 
conditions, rated the extent to which they perceived each 
actor portraying the process manipulations and vision 
manipulation; due to the fact that these observers did not 
Perform the task, vision was the only content condition 
manipulated. For this pilot test the actor and the 
Process manipulation was between-subjects, and the content 
manipulation was within-subjects. That is, each subject 
saw either the male or female actor and observed either 
high or low enthusiasm. Each subject heard both the 
Vision and no vision manipulation, counterbalanced across 
th e Pl? lot sessions. subjects then completed the 
manipulation check items for these variables. 
Twenty subjects observed the male actor. The 
subjects' ratings revealed that they were able to 
distinguish between the vision and no vision scripts 
(t~s.ss, Q<.OOOS). They were also able to determine 
Whether the male actor displayed low or high enthusiasm 
(t~B.78, Q<.0005). 
74 
Nineteen subjects observed the female actor. Again, 
the subjects distinguished the vision from the no vision 
sc ri. pts (t=9.59, Q< .0005) and determined whether the 
female actor displayed high or low enthusiasm (t=4.66, 
Q< .0005) . 
Thus, it was confirmed that each actor was portraying 
the process behaviors as intended and that the content of 
the communication was as intended. These variables were 
perceived by the observers as intended. No changes were 
made to the process behaviors or to the vision 
manipulation. 
Also, this pilot session revealed that the subjects 
did not have any significant questions to ask the leader, 
probably because the scripts were not intended to raise 
quest Sio un bs j. e cts were asked to write down any 
questions that came to mind as the leader spoke, but less 
than 10 questions were recorded. After the pilot, these 
questions were reviewed with the actors to assist them in 
preparing for the actual experiment. 
Manipulation checks 
The means and standard deviations of the vision 
4 45 
ma ni? pulation check scale were 2.36 ( . BO) an d . (.50) 
for the no vision and vision conditions respectively. 
75 
These means were significantly different (1=26.24, 
2<.0005). 
The means and standard deviations of the task 
strategy manipulation check scale were 1.57 (.81) and 3.86 
(.91) for the no task strategy and task strategy 
conditions respectively. These means were significantly 
different (1=22.16, 2<.0005). 
The means and standard deviations of the process 
manipulation check scale were 2.49 (.75) and 3.59 (.78) 
for the low enthusiasm and high enthusiasm conditions 
respectively. These means were significantly different 
(1=11.94, Q<.0005). 
Although the independent variables were orthogonally 
manipulated, the independent variables may not have been 
perceived as such. The three manipulation check scales 
were all significantly intercorrelated (r=.17 for vision 
with task strategy, r=.36 for vision with enthusiasm, 
r=.25 for task strategy with enthusiasm; all 2<.0l). In 
addition, the vision condition and the task strategy 
condition (dummy coded) were each positively related to 
the enthusiasm manipulation check scale (r=.26, 2<.0l and 
r=.18, 2<.0l, respectively). So, subjects who were in the 
vision or task strategy conditions were more likely to 
report that the leader was enthusiastic than subjects in 
no vision or no task strategy conditions. 
76 
This indicates that the manipulations were not 
Perceived as totally orthogonal. Still, the manipulations 
Were regarded as successful due to the fact that the 
manipulations were orthogonally manipulated and that each 
variable was perceived as intended (i.e. subjects 
correctly indicated whether they were in the vision or no 
Vision condition, task strategy or no task strategy 
condition, and high or low enthusiasm condition). 
Assumptions Regarding Manipulations 
~ession 1 versus Session 2 
Each cell was run by each actor two times, thus 
comprising two sessions per actor per cell. To test 
Whether the actors consistently carried out the 
manipulations for session 1 and session 2, 1-tests were 
conducted. For the vision manipulation check scale, the 
means (standard deviations) were 3.39 (1.24) for session 1 
and 3.45 (1.25) for session 2. These means were not 
significantly different (1=.44, ns). The means (standard 
deviations) for the task strategy manipulation check scale 
Were 2.76 (1.50) for session 1 and 2.64 (1.36) for session 
2 ? These means were not significantly different (1=.66, 
!1..?.) ? The means (standard deviations) for the enthusiasm 
manipulation check scale were 3.15 (l.02 ) for session 1 
and 2.92 (. 84 ) for session 2, which were not significantly 
77 
different (~=2.00, ns). Thus, subjects present at the 
first session of a given cell did not perceive the 
manipulations as significantly different than the subjects 
present at the second session of that cell. This finding 
provides support that the actors consistently portrayed 
the desired behaviors. All analyses conducted were based 
on combining across the experimental sessions within a 
given cell. 
Male versus Female Actor 
To test whether the male actor carried out the 
manipulations differently than the female actor, an 
analysis of variance was computed for each manipulation 
check scale. These computations were also used to test 
for the presence of an interaction of actor with subject 
gender. 
There were no significant differences between the 
actors for the vision [E (1, 274) = .89, ns) or task 
strategy [E (1, 275) = .24, ns) manipulation check scales. 
This indicates that the vision and task strategy variables 
were perceived the same regardless of actor. 
There was a significant difference between the actors 
on the enthusiasm manipulation check scale [E (1, 274) = 
11.2, 2<.0l]. The means, 3.22 for the male actor and 2.86 
for the female actor, indicate that the male actor was 
perceived as significantly more enthusiastic than the 
78 
 were conducted 
to test 
ctor. Addition
al analyses
female a actors' effects  
differences in t
he
Whether there w
ere any 
ion 
formance, attitu
de, or percept
on any of the p
er
the male actor d
id not 
t 
bles. These te
sts showed tha
varia ctor 
ects from the fe
male a
v~ significantl
y different eff
ha ) = .07, ns] or ance [E (1, 274
regarding quali
ty of perform
 
 (1, 274) = .37, ns
]. For the
ormance [E
quantity of perf the 
ined below, none
 of 
13 attitudinal var
iables exam
statistical sign
ificance. So, 
effects for act
or reached 
e on the 
re was a signifi
cant differenc
although the to 
this did not app
ear 
enth Usi? asm manipula
tion check, 
 dependent varia
bles. 
impact on any r's behavior was
 
acto
e subjects' perc
eptions of the 
Th the 
 related to the 
interaction of 
ly
not significant
er. For the vis
ion 
's gend
SUbJ'
 he actorects' gender wi
th t
=  E (1, 274) .73, . (n
s), for the
man? . ipulation check
 items, 
E (l, 275 ) = 3.0 
ipulation check 
items, 
task strategy m
an
 check items, E 
or the enthusias
m manipulation
(11?.), and f rs, all analyses
 
se facto
274) = . 67 (ns). B
ased on the
(l,  across the two 
ere based on com
bining measures
conducted w
actors 
Meas-u
res 
~A~c~r!:..!o~s~s___:!:T:..!rc..:i~a? :!c
l~s~ 
= rial to test for
 
each t
Performance was
 measured for 
n. Repeated me
asures 
ct? ing by conditio
lfferential lear
n
ty of 
ables for qualit
y and quanti
analysis of var
iance t
79 
For quality of n Table 2. 
Performance are
 shown i
r trial 1 
ndard deviation
 fo
Performance, th
e mean and sta
5.4 
pectively was 6
.4 (6.03) and 
anct t ri? al 2 errors re
s
 
hese means were
 significantly
6 1 Table 2 shows t
hat t
( ? 0). 
6, ,2<.0l], indi
cating a 
different [E (1
, 271) = 7.1
r all subjects. 
Subjects 
Sign1' fl' eant learning 
effect fo
 probably due t
o 
 2 than trial 1
,
made fewer erro
rs on trial
icant four-way gnif
lear ? e task. There 
was also a si
ning th
dent variables 
and 
epen
raction between
 the three ind
inte ? 
xam?i na 
t ion of the 
 [F 2<
.05]. An e
trials _ (1, 271) = 4.80
, ,
f e ff ec t s. Due 
? etable pattern 
o 
r
means r
p
revealed no int
e
 crucial to the
 
raction is not
to the fact tha
t this inte
 fact that 4-wa
y 
is dissertation
 and the
main focus of t
h
interpret, an i
n-
s are extremely
 difficult to 
interaction
was not conduct
ed. 
 
 analysis of th
is interaction
depth
, Table 2 shows
 that only 
ty of performan
ce
For quanti s s1? gn1' f'i can 
t [E 
interact 
'i on wa 
the task ategy by t
rials str re 3 
The means plotte
d in Figu
(l, 271) = 4.13, .Q<.05]. 
itions improved 
across 
task strategy c
ond
i nd icate that t
he 
 not improve at
 
roup did
 While the no t
ask strategy g
trials
a11. 
they 
subjects had fe
wer errors as 
In general, all
 
n while 
riment regardle
ss of conditio
e expe
Went through th xception 
constant, with 
the e
the number of pages
 remained 
ttle support fo
r 
There was li
Of th e task strategy
 group. 
80 
Table 2 
Repeated Measures Across Trials 
Dependent 
Variable Source df MS .E 
Performance Within Cells 271 17.84 
7.16 .. 
Quality Trials (T) 1 127.72 
Vision (A) X T 1 3.28 .18 
Task Strategy (B) X T 1 12.13 .68 
Enthusiasm ( C) X T 1 3.23 .18 
A X B X T 1 46.53 2.61 
A X C X T 1 2.66 .15 
B X C X T 1 6.55 .37 
A X B X C X T 1 85.72 4. so? 
Performance Within Cells 271 24.60 
Quantity Trials (T) 1 26.85 1.09 
Vision (A) X T 1 2.09 .08 
Task Strategy (B) X T 1 101. 59 4. 13 ? 
Enthusiasm ( C) X T 1 1.12 .05 
A X B X T 1 52.71 2.14 
A X C X T 1 4.32 . 18 
B X C X T 1 30.38 1. 23 
A X B X C X T 1 58.73 2.39 
. .. 
:Q<.05, :Q<.01 
81 
Figure 3 
Trials by Task strategy Interaction 
30 
Task Strategy 
- 29 co 
Cl) 
Cl 
-ia Q. >. 28 
:-;:: 
C 
ia 
::, 
0 
Cl) 27 
CJ 
C 
ia 
E...  
..2..  26 
Cl) 
Q. 
25 No Task Strategy 
24 
Trial 2 
Trlal 1 
82 
fferential learning
 occurred across 
th proposition that d
i
e  
 purposes, perform
ance
trials. 
lysis
Thus, for data ana
ance quantity vari
ables were each 
quality and perform
. 
combined across tr
ials
ve statistics and 
Correlations 
Descripti
ions of 
shows the means an
d standard deviat
Table 3 
ent 
ent variable for th
e three independ
each depend
d 4c show means an
d 
an
V . 
, 
ariables or cells.
 Tables 4a, 4b
nt variable 
st ndard deviations f
or each independe
a
ent variables. 
apsed across the o
ther two independ
Coll
ning the means for hen exami
Caution should be 
taken w
use these means do
 
rformance, anxiety
, and energy beca
Pe
relevant pre-experi
mental measures. 
not control for th
e 
t f f  or th e per orman
ce 
nt to no e
This i's especially importa
(i.e. higher 
riable. Table 3 sh
ows fewer errors 
quality va
w enthusiasm cell 
lity) for the visio
n/task strategy/lo
qua l 
ng cells have appr
oximately an equa
While the remaini of a 
uld infer the prese
nce 
number of errors. 
One co
e means, but, as sh
own 
ed on th
three-way interact
ion bas
y 
e ability is contr
olled, no three-wa
nc
below in Table 61 
o
lthough ability was
 not 
icant. A
interaction is sig
nif
 independent 
ificantly correlate
d with any of the
sign ? 
 was slightly high
er ability for 
Va 1? ables (Table S), 
there
r
i?a sm ce 
11 ? 
the v? ? task strategy/low 
enthus
ls1on/
83 
Table 3 
r Performance,
 
ndard Deviatio
ns fo
Cell Means an
d Sta
n Variables? 
Attitude, and 
Perceptio
Means 
(standard devi
ations) 
ision 
ent siasm No Visi
on V
Depend Enthu TS 
TSb No TS 
~ Condition N
o TSb 
13.0 5.3 
Performance 
.7 
Low 15.4 
13
( 4. 8) 
11. 1) (13.4)
 ( 9. 6) 
Quality (
11.8 12.8
 10.4 
High 10.5 
10.9) (13.1
) ( 8. 7) 
(8.5) (
48.7 58.0 
.4 59.7 
Per f or:rnance Low 48
 (13.8) 
 (15.9) (17.2
) (14.4)
Quantity
.9 49.0 
53.0 
High 48.2 
54
8) (18.0) (1
5.5) 
(11.9) (15.
.9 2.7 
2.3 
Anxiety 
2
Low 2.7 
(. 7) 
(. 5) (. 9
) (. 7) 
(Post) 
2.4 2.7 
2.5 
High 2.5 
(. 8) (. 6) 
(. 7) (. 6
) 
and performanc
e 
a p y is the numb
er errors 
lit
erformance qu
a
 and energy w
ere 
 number of pag
es. Anxiety
quantity is th
e
er items were e. All oth
sured using a 
-point scal
mea 5
 a 7-point sca
le. 
measured using
TS= Task stra
tegy 
b k strategy; 
No Ts: No tas
84 
(Table 3, cont.) 
Energy Low 4.0 3.9 3.8 3 . 9 
(Post) ( . 5) ( . 4) ( . 5) ( . 5) 
High 3.9 3.8 3.9 3.9 
( . 5) ( . 4) ( . 5) ( . 5) 
Task Low 3.8 4.0 4.0 4.9 
satisfaction (1. 6) ( 1. 7) ( 1. 3) (1. 4) 
High 3.8 4.3 3.8 4.3 
(1. 5) (1. 6) ( 1. 4) (1. 5) 
Enthusiasm Low 3.4 3.2 3.7 3.7 
For Task ( 1. 3) ( 1. 4) ( 1. 2) (1. 4) 
High 3.3 3.4 3.4 3.7 
(1. 4) (1. 5) ( 1. 4) ( 1. 2) 
Enthusiasm Low 4.2 3.9 4.5 4.5 
For Vision ( 1. 4) ( 1. 3) ( 1. 0) ( 1. 3) 
High 4.4 4.4 4.4 4.7 
(1. 2) ( 1. 2) ( . 9) ( 1. 1) 
Willingness Low 3.4 3.6 3.8 3.9 
to Work (1. 3) ( 1. 2) ( 1. 1) ( 1. 2) 
High 3.6 3.8 3.3 3.9 
( 1. 2) (1. 4) ( 1. 1) ( 1. 4) 
.-_ --- - - ------ -? -;:-- . -
85 
(Table 3, cont.) 
Congruence Low 5.6 5.5 6.0 6.4 
of Beliefs ( . 7) (1. 0) ( . 9) ( . 6) 
and Values High 5.8 5.8 6.3 6 . 2 
( . 7) ( . 8) ( . 7) ( . 7) 
Trust Low 4.4 4.4 4.8 5.3 
in Leader (1. 0) ( . 7) ( 1. 2) ( . 9) 
High 4.6 4 . 8 4.8 5.0 
(1. 1) ( 1. 4) (1. O) (1. 0) 
Liking Low 4.5 4.2 4.5 5.0 
the Leader ( . 9) ( . 7) ( . 9) ( . 9) 
High 4.7 4.8 4.9 5.0 
( . 8) (1. 0) ( . 9) (1. O) 
Charisma Low 3.4 3.8 4.5 4.8 
(Form SY) ( 1. 3) ( . 9) (1. 0) ( . 8) 
High 3.8 4.1 4.9 4.9 
( 1. 1) ( . 8) ( . 9) ( 1. 0) 
Intellectual Low 2.9 3.6 3.8 4.5 
Stimulation ( 1. 4) ( 1. 4) ( 1. 3) (1. 5) 
High 3. 0 4.4 3.9 4.4 
( 1. 2) ( 1. 4) ( 1. 4) ( 1. 2) 
86 
(Table 3' cont.) 
3.0 3.7 4.0 
Inspiration LOW 2.7 
(1. 2) (1. 3) 
( 1. 5) (1.2) 
3.3 4.0 3.9 High 2.9 
(1. 3) ( 1. 3) 
( 1. 3) 
(1. 5) 
4.9 4.5 5.5 
Task LOW 4.5 
(1.8) ( 1. 3) 
(1. 4) 
Clarity ( 1. 3) 
5.1 
High 5.1 5.1 
5.0 
(1. 4) 
( 1. 3) (1.6) 
( 1. 5) 
87 
Table 4a 
 Means and Standard Deviations fo
r Performance I 
Vision:
Attitude and Perception Variables?
 
Means 
(standard deviations) 
Dependent 
No Vision Vision 
Measure 
12.9 10.5 
Performance Quality 
(11. 1) ( 9. 9) 
 52.0 
Performance Quantity 
52.6
(15.9) (15.7) 
2.6 2.5 
Anxiety (Post) 
( . 7) ( . 7) 
3.9 3.9 
Energy (Post) 
(. 5) (. 5) 
4.0 4.3 
Task Satisfaction 
( 1. 6) (1. 4) 
3.3 3.6 
Enthusiasm for Task 
(1. 4) ( 1. 3) 
nce 
Performance quality is the number
 errors and performa
a 
quantity is the number of pages. 
Anxiety and energy were 
l other items were 
measured using a 5-point scale. 
Al
measured using a 7-point scale. 
88 
(Table 4a, cont.) 
Enthusiasm for Vision 4.2 4.5 
(1. 3) ( 1. 1) 
Willingness to Work 3.6 3.7 
(1. 3) ( 1. 2) 
Congruence in Beliefs and Values 5.7 6.2 
( ? 8) (. 7) 
Trust in Leader 4.5 4.9 
(1. 1) (1. 0) 
Liking Leader 4.6 4.8 
~. 9) (1. 0) 
Charisma ( Form BY) 3.7 4 . 8 
( 1. 1) ( 1. 0) 
Intellectual Stimulation 3. 5 . 4 . 2 
(1. 5) ( 1. 4) 
Inspiration ? 2.9 3.9 
(1. 4) ( 1. 2) 
Task Clarity 4.9 5. 0 
(1. 5) ( 1. 4 ) 
89 
Table 4b 
Task Strategy: Means and Standard Deviations for 
Performance, Attitude and Perception Variables? 
Means 
(standard deviations) 
Dependent No Task Task 
Measure strategy Strategy 
Performance Quality 13.0 10.4 
(10.6) (10.3) 
Performance Quantity 48.6 56.2 
(15.0) (15.7) 
Anxiety (Post) 2.6 2.5 
( . 7) ( . 7) 
Energy (Post) 3.9 3.9 
(. 5) ( . 5) 
Task Satisfaction 3.9 4.4 
( 1. 4) ( 1. 6) 
? Performance quality is the number errors and performance 
quantity is the number of pages. Anxiety and energy were 
measured using a 5-point scale. All other items were 
measured using a 7-point scale. 
90 
(Table 4b, cont.) 
Enthusiasm for Task 
3.5 3.5 
( 1. 3) (1. 4) 
 
Enthusiasm for Vision 
4.4 4.4
( 1. 2) ( 1. 2) 
 to Work 3.5 3.8
 
Willingness
(1. 2) ( 1. 3) 
9 6.0 
Congruence in Beliefs and Values 
5.
(. 8) ( ? 8) 
 Leader 4.6
 4.9 
Trust in
(1. 1) ( 1. 0) 
4.6 4.7 
Liking Leader 
(.9) (1.0) 
4.4 
Charisma (Form 
.1 
BY) 
4
( 1. 3) ( 1. 0) 
Stimulation 3
.4 4.2 
Intellectual 
(1. 4) (1. 4) 
3.3 3.5 
Inspiration 
(1. 5) ( 1. 3) 
4.8 5.1 
Task Clarity 
(1. 4) (1. 5) 
91 
Table 4c 
asm: Means and Standard Deviations for Performance, Enthusi
Attitude and Perception Variables
8 
Means 
(standard deviations) 
Dependent Low High
 
Measure Enthusiasm Enth
usiasm 
ormance Quality 12.2 11.3
 
Perf
(10.9) (10.3) 
Performance Quantity 53 . 2 
51.4 
( 16. 1) (15.5) 
Anxiety (Post) 2.6 
2.5 
( . 7) (. 7) 
3.9 
Energy (Post) 3.9
 
( . 5) ( . 5) 
4.1 
Task Satisfaction 
4.2 
( 1. 6) (1. 5) 
? Performance quality is the number errors and perfor
mance 
antity is the number of pages. Anxiety and energy w
ere 
qu
measured using a 5-point scale. All other items we
re 
measured using a 7-point scale. 
---- ?- - --?~ -~.  --??-- -- ?---? -----. __,. _ ..~ ;-
92 
(Table 4c, cont.) 
Enthusiasm for Task 3.5 3.5 
(1. 3) (1. 4) 
Enthusiasm for Vision 4.3 4.5 
( 1. 3) ( 1.1) 
Willingness to Work 3.7 3.6 
(1. 2) ( 1. 3) 
Congruence in Beliefs and Values 5.9 6.0 
( ? 9) (. 7) 
Trust in Leader 4.7 4.8 
(1. 0) ( 1. 1) 
Liking Leader 4 . 5 4.9 
(. 9) (. 9) 
Charisma (Form BY) 4.1 4.4 
(1. 2) ( 1. 1) 
Intellectual Stimulation 3.7 4.0 
( 1. 5) (1. 4) 
Inspiration 3.3 3 .5 
(1.4) ( 1. 3) 
Task Clarity 4.8 5.1 
(1. 5) ( 1. 4) 
93 
lations among the i
ndependent 
able 5 shows the c
orre
T
nd attitude and 
variables, perform
ance variables, a
perception variabl
es. 
Effects of Indepen
dent Variables 
sented in chapter 
4 regarding the 
The propositions p
re
iables were tested 
via 
ndent var
effects of the ind
epe
le was 
ariance. When the
 dependent variab
analysis of v s used 
lity (practice tria
l performance) wa
performance, abi g 
n & cox, 1957). A
lso, when testin
as a covariate (Co
chra
mental measures of
 mood 
 pre-experi
the effects of moo
d, the
and the post-experi
mental measures 
ed as a covariate was us ng 
endent variable, th
us testi
of mood were used 
as the dep
 mood after contro
lling for initial 
the differences in
computations and th
e 
riance 
mood. The analysi
s of va
ented in Table 
 of the significan
t effects are pres
strength
6. 
The Effects of Vis
ion 
o communicated a 
roposition 1 stated
 that a leader wh
P
ality in a positive
 
mance qu
Vision would affec
t perfor
ader who 
(i.e. fewer errors)
 compared to a le
direction 
n be seen in Table
 6, there 
sion. As ca
communicated no vi
ision effect. Exa
mination of the 
Was a significant 
v
in 
t fewer errors were
 made 
means, in Table 4a
, shows tha
94 
Table 5 
Correlations Among Independent and Dependent Variables 
2 ? 3. 4. 
5. 6. 7 ? 
1. 
. 11 ? -.02 
1 - Vision .02 .04 
.02 -.01 
-.06 . 12? . 24 ?? 
2 ? Task strategy .0 .07 2 
3 ? Ent -.05 
.03 .04 -.06 
husiasm 
-.36 .. .45 .. .03 
4 ? Ability Quality 
- . 10? .49 .. 
5 ? Ability Quantity 
-.02 
6 ? Performance Quality 
7 ? Performance Quantity 
8 ? Anxiousness (Post) 
9 ? Energy (Post) 
l0. Task Satisfaction 
11 ? Enthusiasm for Task 
12 ? Enthusiasm for Vision 
13 ? Willingness to work 
14 ? Congruence 
15 ? Trust in Leader 
16 ? Liking the Leader 
17 ? Charisma (Form SY) 
18 ? Intellectual stimulation 
19 ? Inspiration 
20. T ask Clarity 
:Q<. 0 5 , ?? 12<. 01 
95 
(Table 5, cont.) 
8. 9. 10. 11. 12. 13
. 14. 15. 16. 
. 12? .07 . 3 3 ?? . 
19 .. .15 .. 
1. -.Q6 .00 .09 . 12? 
. 11 ? .05 
2 . -.os -.03 .16 .. .03 .00 . 12? 
.04 
.09 . 07 .18 .. 
3 . 9 -.01 --06 -.08 -.02 -.01 .0
 5 .06 .08 .09 .09
 
4. - . 13? . 18 ?? .17 .. . 12? .0
. 10? .02 -.02 
5. .03 -.01 . 13? -.01 .01 
-.02 -
8 .. .08 .15 ..
 
.1 2 . 1
4 ? 
6. - . 3 4 ?? .03 .37 .. . 10? .0
. 11 ? .02 
7. -.05 .09 . 2 4 ?? .06 .07 
.08 -.03 
 1 
8 . - . 11? -.38 .. - . 14 ?? 8 -.06
 -.03 -.04 -.0-.0
9. - . 10? -.17 .. -.17 .. -.17
 .. -.02 -.07 -.05 
. 2 5?? . 14 ?? 
10. .47 .. . 4 7?? 
. 31 ?? . 12? 
.72 .. . so?? . 14 ? . 2 9
 ?? . 13 ? 
11. 
. 44 ?? . 21 ?? . 31?? . 2 3 ??
 
12. 
. 2 o?? . 4 0 ?? . 31 ?? 
13. 
.38 .. . 36 .. 
14. 
.60 .. 
15. 
16. 
17. 
18. 
19. 
2 0 . 
96 
(Table 5, cont.) 
17. 18. 19. 20. 
1. . 45 .. . 2 4 ?? . 3 5?? .04 
2 . . 12? .28 .. .09 . 13 ? 
3 . . 16 ?? . 10? .09 .09 
4. -.04 - . 10? -.01 .26 .. 
5. .08 -.03 .04 .22 .. 
6 . . 06 .00 . 10? . 3 8 ?? 
7 . . 07 .06 . 11 ? .32 .. 
8. -.02 -.04 .04 -.25 .. 
9. - . 16 ?? -.09 - . 13? .09 
10. . 27 .. . 21 ?? . 3 o?? . 3 5 .. 
11. . 3 3 ?? . 29 .. . 3 3 ?? . 15 .. 
12 . . 39 .. . 3 3 ?? . 41 ?? . 15?? 
13. . 41 ?? . 4 0 .. . 38 .. .09 
14. . 41 ?? . 3 0 .. . 2 9 ?? . 11 ? 
15. . 58 ?? . 3 4 ?? . 40 .. .22 .. 
16. . 54 ?? . 32 .. . 3 3 ?? .18 .. 
17. . 55 .. .68 .. .15 .. 
18. .62 .. . 04 
19. . 1 0 ? 
20. 
97 
Table 6 
nalysis of (Co- )Variance Resu
lts? 
A
Dependent 
w 2 
Y.ariable Source df M
S E 
formance Ability 1 6360.
5 7 3. 1 ?? .20 
Per
3.4+ .01 
Quality Vision ( A ) 1 297
.8 
 
Task Strategy (B) 1 350.1
 4. o? .01 
5 . 9 
Enthusiasm (C) 1 80
.
A B 1 154 .
4 1.8 
X 
118.4 1. 4 
A X C 1 
1 205.5 2.4 B X C 
30.0 . 3 
A X B X C 1 
dual 269 87.1 Resi
+ 
Q<.10, ? ]2<.05, ?? ]2<.01 
e to 
? Sample size varies slightly 
across the analyses du
missing data. omega-squared (
w2)is presented only for 
significant E values. 
98 
(Table 6, cont.) 
Perform 9a 3n _ce Ability 1 16148.0 
5 .. .23 
Quantity Vision (A) 1 12.0 . 1 
Task Strategy (B) 1 4932.5 28. 1 ?? .07 
Enthusiasm ( C) 1 493.1 2.8 
A X B 1 13.9 .1 
A X C 1 216.3 1. 2 
B X C 1 62.8 . 4 
A X B X C 1 26.4 . 2 
Residual 270 175.6 
Anxiety Anxiety (Pre) 1 53.3 169.1 .. .37 
(Post) Vision (A) 1 . 4 1. 2 
Task Strategy (B) 1 . 1 . 2 
Enthusiasm ( C) 1 . 2 . 5 
A X B 1 1. 2 3. 9? .01 
A X C 1 1. 6 5. o? .01 
B X C 1 . 2 . 7 
A X B X C 1 . 7 2.1 
Residual 270 . 3 
99 
(Table 6, cont.) 
1 27.4 2 2 s. o?? .46 Energy Energy (Pre) 
.o . 4 
(Post) Vision (A) 1 
. 1 
Task strategy (B) 1 . 0 
. 2 
Enthusiasm ( C) 1 
. 0 
. 0 . 7 
A X B 1 
1 .o . 4 A X C 
.o . 6 
B X C 1 
1 . 0 . 1 A X B X C 
Residual 265 
. 1 
5.4 2.4 
Task Vision (A) 1 
16.2 7. 2?? .02 
Satis- Task Strategy (B) 1 
.9 . 4 
faction Enthusiasm ( C) 1 
1 2.3 1.0 A X B 
1 4.7 2.1 A X C 
1 .1 .o 
B X C 
1 1. 8 . 8 
A X B X C 
2.3 
Residual 272 
100 
(Table 6, cont.) 
Enthusiasm Vision (A) 1 6.7 3. 8. .01 
for Task Task Strategy (B) 1 . 3 . 2 
Enthusiasm (C) 1 . 1 .1 
A X B 1 .7 . 4 
A X C 1 . 5 . 3 
B X C 1 1.8 1.0 
A X B X C 1 . 0 . 0 
Residual 272 1.8 
Enthusiasm Vision (A) 1 5.7 4. o? .01 
for Vision Task Strategy (B) 1 . 0 . 0 
Enthusiasm (C) 1 3.2 2.2 
A X B 1 1. 5 1.1 
A X C 1 1.8 1. 2 
B X C 1 .8 . 6 
A X B X C 1 . 0 . 0 
Residual 273 1.4 
101 
(Table 6, cont.) 
Willingness Vision (A) 1 1. 9 1.2 
To Work Task Strategy (B) 1 6.4 4. 1 ? .01 
Enthusiasm (C) 1 . 2 ? 2 
A X B 1 . 2 .1 
A X C 1 3.9 2.5 
B X C 1 ? 8 . 5 
A X B X C 1 1. 3 . 9 
Residual 273 1. 6 
Congruence Vision (A) 1 20.3 34.3 .. .11 
of Beliefs Task Strategy (B) 1 ? 1 . 2 
and Values Enthusiasm (C) 1 1.0 1. 7 
A X B 1 . 5 . 7 
A X C 1 1.1 1. 9 
B X C 1 . 5 .8 
A X B X C 1 1. 7 2.9 
Residual 272 . 6 
102 
(Table 6, cont.) 
Trust Vision (A) 1 10.8 1 o. o?? .03 
in Leader Task Strategy (B) 1 3.3 3 . 1 
Enthusiasm (C) 1 .8 . 7 
A X B 1 1.0 1.0 
A X C 1 3.3 3.0 
B X C 1 . 0 . 0 
A X B X C 1 1. 4 1. 3 
Residual 272 1.1 
Liking Vision (A) 1 4.9 6. 1 ? .02 
the Leader Task Strategy (B) 1 . 3 . 4 
Enthusiasm ( C) 1 7.2 8. 8 .. .03 
A X B 1 2.9 3.6 
A X C 1 1.0 1.2 
B X C 1 . 0 . 0 
A X B X C 1 3.5 4. 2? .01 
Residual 272 . 8 
103 
(Table 6, cont.) 
Vision (A) 1 70.4 69.2 ..
 . 19 
Charisma 
(Form 8Y) Task Strategy (B) 1 3.5 3.4
 
Enthusiasm (C) 1 6.3 6. 2? .02 
A X B 1 .7 . 7 
A X C 1 . 6 . 6 
B X C 1 1.1 1.1 
A X B X C 1 . 2 . 2 
Residual 269 1.0 
Intellectual Vision ( A) 1 31. 9 17. 4
 ?? .05 
Stimulation Task Strategy ( B) 1 ? 44. 4 24. 3 .. .07 
Enthusiasm ( C) 1 3.0 1. 6 
A X B 1 2.9 1. 6 
A X C 1 3.3 1.8 
B X C 1 1.0 . 5 
A X B X C 1 4.2 2.3 
Residual 273 1.8 
104 
(Table 6' cont.) 
I nspiration Vision (A) 1 62.8 38. 1 ?? .12 
Task strategy (B) 1 3.3 2.0 
Enthusiasm (C) 1 2.2 1. 3 
A X B 1 . 6 .4 
A X C 1 . 4 . 2 
B X C 1 . 5 . 3 
A X B X C 1 1. 3 . 8 
Residual 273 1. 6 
Task Vision (A) 1 .7 . 3 
clarity Task Strategy (B) 1 9.3 4. 4 ? .01 
Enthusiasm ( C) 1 3.6 1. 7 
A X B 1 2.9 1. 4 
A X C 1 1. 7 . 8 
B X C 1 5.8 2.7 
A X B X C 1 1. 2 .6 
Residual 273 2.1 
105 
 vision condition than in the no vis
ion conditions 
the ' 
supporting proposition 1 . 
ough this effect was only significan
t at the 
Alth
Q<.10 level when controlling for ab
ility, this is solely 
the result of a degrees of freedom e
ffect. The vision-
nificant at the .05 
performance quality correlation was 
sig
level (as shown in Table 5) but the 
vision-ability 
correlation was virtually zero (K=.0
2, ns). This was 
 coefficient 
confirmed by computing a partial co
rrelation
Where the effect of ability was remo
ved from the vision-
performance quality relationship. T
he partial correlation 
the first-order 
Was K=-.11 (Q<.05) which is the sam
e as 
correlation between vision and perfo
rmance quality. 
As stated above, examination of the 
performance 
ree-way 
quality means in Table 3 indicates a
 possible th
raction. However, this effect was n
ot significant in 
inte
the analysis of variance when contro
lling for ability. 
ion on the 
Proposition 2 concerned the effects 
of vis
 and perceptions of the subjects tow
ard the task 
attitudes
and leader. As can be seen in Table 
6, 8 of the 13 
ificantly 
attitude and perception variables w
ere sign
 the vision variable. Examining the
 means 
affected by
shows that vision, compared to no vi
sion, led to 
significantly (a) more enthusiasm fo
r the task, (b) more 
ngruence 
enthusiasm for the vision, (c) more 
perceived co
106 
jects' beliefs and v
alues, (d) 
etween the leader's
 and sub
b
of liking the leade
r, 
(e) more 
more trust in the 
leader, 
(g) 
eption of the leade
r as charismatic, 
(f) greater perc
and (h) more inspir
ation . 
re intellectual stim
ulation, 
mo
 vision on anxiety,
 
There were no sign
ificant effects of
n, willingness to w
ork for the 
energy, task satisf
actio
rall, proposition 2 
was 
leader , or task clarity. O
ve
supported. 
eader's vision had 
an impact on the 
To sum up, the l
e vision 
' performance quali
ty; subjects in th
subjects who 
ignificantly fewer 
errors than those 
condition made s who 
factual information
. Subjects 
heard the leader g
ive 
 had more posit i ve 
were exposed to the
 visionary leader
leader than those w
ho 
ttitudes and percep
tions toward the 
a
jects who heard the 
saw the nonvisionar
y leader. sub
"buy into" the lead
er's ideas, as 
Vision seemed to 
congruence, trust, 
gnificant effects o
n 
indicated by the si
effect  nt son 
li' k i? ng, and charisma. 
The significa
 for the vision 
enthusiasm for the 
task and enthusiasm
s, 
e vision seemed to 
excite the subject
indicates that th
ifference in subjec
ts ' 
ke a d
although it did no
t ma
in the future. 
w? l lingness to work for 
the leader 
J. 
e subjects to think
 about new 
p? y, the vision led t
h
ina11 a good 
d inspired them to 
do 
Ways to perform the
 task an
107 
as shown by the significant e
ffects on intellectual 
job, 
stimulation and on inspiratio
n. 
The Effects of Task Strategy 
tion 3 concerned the effects o
f task strategy 
Proposi
the task 
on performance quality and qu
antity. Subjects in 
 and 
st n had significantly fewer err
ors
rategy conditio
ing 
inserted significantly more p
ages than those not receiv
ategy information (see Table 
6). Thus, 
any task str
Proposition 3 was supported . 
ific propositions were made re
garding 
Although no spec
th  effects of task strategy on 
attitudes and perceptions, 
e
fects were found. The task 
four additional significant e
f
strategy information positively
 affected subjects' 
on, willingness to work for t
he 
feelings of task satisfacti
. 
leader, intellectual stimulat
ion, and task clarity 
.Th_e Interaction of Vision and
 Task Strategy 
 the vision by task strategy Proposition 4 concerned
interaction on attitudes and 
perceptions. This 
ction significantly affected 
subjects' anxiety, 
intera
r 13 
While h avi? ng no s1? gn1?  f i? can 
t ee ff e cts on the oth
dent variables. Figure 4 show
s equal levels of 
depen
h i? sion 
an X?i 
en, wen no v
ety when no task strategy was
 giv
 when neither were given, but 
anxiety was 
Was given, or
were given. 
reduced when both a vision an
d task strategy 
108 
l, there was little support fo
r this proposition. 
In genera
The Effects of Enthusiasm 
 5 concerned the effects of e
nthusiasm on 
Proposition
only 
the attitude and perception m
easures. Table 6 shows 
nthusiasm reached statistical 2 of the 13 effects for e
d 
significance. A high enthusi
asm leader was liked more an
n a low 
Was regarded as being more ch
arismatic tha
ess used by the leader did no
t 
enthusiasm leader . The proc
affect subjects' anxiety, ene
rgy, task satisfaction, 
o work in 
enthusiasm for the task or vi
sion, willingness t
th gruence of beliefs and values,
 trust in the 
e future, con
, or task 
leader, intellectual stimulat
ion, inspiration
clarity. 
on, it can be seen that enthu
siasm had no 
In additi
nificant effects on performan
ce quality or quantity. 
sig
In general, proposition 5 was
 not supported . 
nthusiasm 
.TI::Le Interaction of Vision and
 E
concerned with the effects of 
the 
Proposition 6 was 
Vision by enthusiasm interact
ion on attitudes and 
 
Perceptions. As shown in Tab
le 6, only 1 of the 13
riables showed this interactio
n. 
attitude and perception va
Vision and enthusiasm interac
ted to affect anxiety. 
usiasm 
Figure 5 shows that while sub
jects in the high enth
 l f 
cond 1?  t i?o n had similar levels of anx
i? e t Y rega rd ess o 
thusiasm/no 
Vision condition, the combina
tion of low en
109 
vision resulted in more anxiety than low 
enthusiasm/vision. It may be that the presence of eith
er 
high enthusiasm, vision, or both serve to reduce one's 
anxiety while low enthusiasm plus no vision increase 
anxiety . In general, this proposition was not supported. 
Summary 
Overall, vision significantly affected performance 
quality as well as most attitudes and perceptions toward 
the leader. Task strategy significantly affected 
performance quality and quantity as w~ll as intellectual 
stimulation. Enthusiasm only affected a few attitudes an
d 
perceptions. The interactions between the independent 
variables did not significantly affect performance, 
attitudes, or perceptions. 
110 
Figure 4 
Vision by Task strategy Interaction on Anxiety 
3 
No Vision 
::-----
-- Vision co 
-i. ->. 2 CD 
~ 
C 
< 
1 
Task Strategy 
No Task Strategy 
111 
Figur e 5 
Vi sion by Enthusiasm Interaction on Anxiety 
3 No Vision 
-- Vision 
-?0  ll. 2 
1 
High 
LOW 
EnthuBl&Bffl 
112 
Chapter 7 
Results: Supplementary and Mediator Analyses 
To further explore the relationships of the 
experiment, a supplementary analyses on the two charisma 
scales was conducted. Also, the effects of the goal 
variables were examined via analysis of variance, and the 
mediator analyses of the goal-related variables were 
conducted . 
Charisma Scale Components 
Bass' leadership scales are frequently used in 
empirical leadership research (Avolio, Waldman, & 
Einstein, 1988; Hater & Bass, 1988). Bass (1985; Bass & 
Avolio, 1989) has found that these scales show good 
reliability and validity. However, the scales have been 
criticized for confounding leader behavior with a leader's 
effects on follower attitudes and performance, which is a 
common problem in many leadership measures (Uleman, 1991; 
Yukl, 1989). For these two reasons, a further 
investigation of the two charisma scales (Form SY and Form 
5) was undertaken. Using the charisma items from both 
scales, a principal components analysis with varimax 
rotation resulted in two meaningful components. It should 
113 
many of the factor loadings load highly on 
be noted that 
both factors . 
The components are not process and content 
compon ent s, but seem to reflect (a) the subordinate's (or 
rater's) perception of leader behavior and (b} followers' 
att i? tudes toward the leader. Table 7 shows the results of 
the principal components analysis. Thus, if one 
conceptualizes that leader actions influence subordinates' 
att?i tudes about the leader, then the criticism that Bass' 
charisma scale confounds these two steps is warranted. 
However , one must also consider the fact that charismatic 
leadershi' p defined broadly does include bo t h 1 ead  er 
behav1? ors (e.g. commun. icati. ng a vision) as well as the 
subordinate's attitudes regarding the leader (e.g. whether 
the subo rdi? nate likes the leader). clear 1 Y, fut  ure 
researchers will need to consider the purpose of their 
study when deciding whether to use Bass' measure or not. 
The effects of the independent variables on each 
comp anent was tested via analysis of variance by forming 
two new scales, a leader behavior scale and an attitude 
toward the leader scale. There were s1. gn1' f1' can t ma.i n 
effects of vision and of enthusiasm on the leader behavior 
scale and on the attitude toward the leader scale. These 
effect s were similar to the effects on th e c h ari?s ma (Form 
BY) scale shown in Table 6 . The two components that 
114 
Table 7 
 
al Components Analysis
 of Charisma Scales??b
Princip
Attitude 
Leader About 
Behavior Leader 
Item (Form) 
.46 . 31 
The CEO served as a r
ole model 
for me. (BY) 
bilized a collective s
ense .21 
The CEO mo
of mission. (BY) 
.72 . 36 
The CEO instilled pri
de in being 
associated with him. 
(BY) 
eds .73 .3
2 
The CEO engaged in wo
rds and de
age of 
Which enhanced his im
competence. (BY) 
peared on the form for
 the 
? Items are worded as 
they ap
actor, the term Presid
ent 
ale 
male actor. For the f
em
? t  d o f 
instead of CEO was used, and f
em?i ni? ne, ins ea
mascu1i?ne , pronouns were used. 
b aken from Bass (1985) 
? Form 5R items 
Form BY items are t
 from 
hted by B. Bass and B
. Avolio. The survey
a~e copyrig
Wh'  items are taken is lice
nsed by Consulting 
ich the
eproduction or 
Psychologists Press (
CPP), Inc. and any r
pproved  b Y CPP. 
Use 0 f the survey and/or item
s must be a
115 
(Table 7, cont.) 
 CEO made me aware of strongly held 
.81 .20 
The
values, ideals, and aspirations 
Which are shared in common. (SY) 
.78 .25 
The CEO demonstrated a strong 
conviction in his beliefs 
and values. (BY) 
The CEO projected a strong, dynamic, 
.65 .45 
and magnetic presence. (8Y) 
52 
I am ready to trust the CEO to 
.54 ?
overcome any obstacles. (8Y) 
6 .JO 
I have complete confidence in the CEO
 .7
( SR) . 
9 
In my mind, the CEO is a symbol of 
.4
success and accomplishment. (SR) 
.66 
The CEO makes me feel good when 
.47 
I'm around him. (SY) 
.47 .63 
The CEO makes me proud to be 
associated with him. (8Y) 
.49 
The CEO is someone in whom I have 
complete faith. (8Y) 
.79 
The CEO has a special gift for seeing
 .19 
What is really worthwhile for me 
to consider. (SR) 
.23 
The CEO has my respect. (5R) 
116 
(Table 7, cont.) 
.17 .83 
The CEO shows enthusia
sm for what 
I need to do. (SR) 
.33 .73 
The CEO has a sense o
f mission which 
municated to me. (5R) he com
.61 
r .45 
The CEO increases my 
optimism fo
the future. ( 5R) .70 
trust in his ability 
.50 
The CEO has my 
to overcome any obsta
cle. (5R) 
117 
id not have different effects th
an the Form BY 
emerged d
cted using the 
scale. Thus, no further analys
is were condu
separate components. 
Goal Variable Results 
Q_escriptive Statistics and Cor
relations 
es are 
The means and standard deviatio
ns for goal scal
ong the 
shown in Tables 9 a 9b, and 9c. 
correlations am
I I 
own in Table 10. Correlations 
between 
goal scales are sh
ormance 
goal scales and the independent
 variables and perf
Varl? ables are presented in Table 11.
 
lm.alysis of Co variance Results
 
 effects of 
Using analysis of co-variance, 
the
ined 
i ndependent variables on each g
oal measure were exam
a nd are shown in Table 12. 
s for 
Vision significantly affected p
ersonal goal
commitment for quality, and self
-efficacy 
quality, goal 
or quality and quantity. The m
eans in Table 9a show that 
f
vision, subjects set more 
When the leader communicated a 
er errors), were more 
difficult goals (i.e. tried for
 few
committed to those quality goal
s, and had higher self-
eff'  when the leader communicated n
o vision. 
lcacy than
stressed quality, it is logical 
that 
Because the vision 
? l ff ed. 
goals in Ya 
ect
and self-efficacy for quality w
ere ma
118 
Table 8 
ations for Goal Scales
a 
Cell Means and Standa
rd Devi
Means 
 standard deviations) (and
 V~i=?s=i?=o=n=-=-- Visio
n 
Dependent Enthusiasm -~N~o:....._
condition No TSb 
No TS 
5.1 
Quality  14.3 12.6 
8.9 
Low
Goa1 .7) ( 8. o) ( 5. 6
) (5.4) 
(8
9 10 . 1 
High 9.9 9.5 
8.
. 0) ( 4. 8) (7.8)
 (7.4) 
( 6
51. 9 
Quantity Low 41.1 49.8 
44.8 
Goal .2) (19.0) ( 14. 
7) (15 . 7) 
(14
44.9 44.4 
High 45.3 47.2 
.1) (13.7) ( 11.
 7) 
(16.1) (13
41.4 
Anticipated 
49.6 
Low 52.4 51. 8 
.7) (12.5) 
Satisfaction (14.9) (14.3) 
(10
50.1 49.2 
for- Quality High 47.9 46.6 
1. 4) (49.2) 
(10.8) (10.0) 
(1
a ors and quantity goal
 
err
Quality goal is the n
umber of 
. n for 
is th d satisfa
ctio
e number of pages. A
nticipate
11 t 
ange f rom o 99; 
qualit Y and quant
.i ty has a possible r
le; 
mmitment was measured
 using a 7-point sca
goa1 co
qualit y self-efficacy are m
ean z-scores . 
Y and quantit
b sk strategy 
No Ts ~ No task strategy
; TS= Ta
119 
(Table 8, cont.) 
59.7 
Anticipated Low 65.6 58.1 
60.1 
(11 ( 9. 5) (9. .0 8) )S   
(11. 4) 
atisfaction 
58.2 606 .1 6.  
61. 6 
for Quantity High  8 
(9.7) ( 9. 6) (7.9) (10.8) 
 
4.8 5.2 
Goal LOW 4.6 4.6 
( ( . 9) 
( . 
1 9. )  Commitment ( . 9) 1) 
4.8 4.8 
for Quality High 4.8 4.5 
(1. 1) ( . 8) 
( . 9) (1. 1) 
4.9 5.24  Goal LOW 4.7 .6 
(1. 0) (1. 0) 
Commitment ( . 9) ( 1. 2) 
4.8 4.8 
for Quantity High 4.9 4.8 
(1. 1) ( 1. 0) 
( . 9) (1. 1) 
.03 .42 
Quality LOW -.28 -.14 
(. 7 3) (.76) 
Self-Efficacy (.75) 
( . 91) 
-.06 .18 
.03 
High -.01 
( . 7 2) (. 8 3) 
( . 82) 
(.75) 
.08 .34 
Q .u 0a 4n  tity LOW -.33 
( . 6 3) (. 78) 
S (.e  8l 6f )-  Efficacy (.72) 
.11 -.04 .02 
High -.11 
(. 69) (.56) 
(.61) (. 65) 
120 
Table 9a 
Vision: Means and Stan
dard Deviations 
for Goal Variables? 
Means 
(standard deviations) 
Dependent 
 Vision Vision 
M_easure 
No
11.7 8.4 
Quality Goal 
(7.4) (6.9) 
45.7 46.2 
Quant i ty Goal 
(15.9) ( 14. 0) 
49.8 47.8 
Anticipated Satisfacti
on 
(12.9) (13.6) 
for Quality 
61.l 60.5 
Ant l? ci?p ated Satisfaction 
(10.4) (10.0) 
for Quantity 
of errors and quantity 
goal 
? Quality goal is the n
umber 
ction for 
is the number of pages
. Anticipated satisfa
ua1 ? uantity has a possible 
range from 11 to 99; 
q ity and q
int scale; 
goal commitment was me
asured using a 7-po
an z-scores. 
qua1 i? ty and quantity self-e
fficacy are me
121 
(Table 9a, cont.) 
Goal Commitment 
4.6 4.9 
or Quality (
1. 0) ( ? 9) 
f
mitment 4.7 
4.9 
Goal Com
(1. 0) (1. 0) 
for Quantity 
-.13 .15 
Quality Self-efficacy 
(. 8) ( ? 8) 
0 .11 
Quantity Self-efficacy 
-.1
(. 7) (. 7) 
122 
Table 9b 
Task Strategy: Means and S
tandard Deviations 
for Goal Variables? 
Means 
(standard deviations) 
sk Task 
Dependent 
No Ta
Strategy Strategy 
M_easure 
10.6 9.4 
Quality Goal 
( 7. 5) ( 7. 0) 
44.9 48.0 
Quantity Goal 
( 14. 6) (15.1) 
47.5 
Anticipated Satisfaction 
50.1 
 (
12.2) (14.2) 
for Quality
62.1 59.5 
Ant1? c1? pated Satisfaction 
(9.9) (10.3) 
for Quantity 
f errors and quantity goal 
ality goal is the number o? Qu
e number of pages. Antici
pated satisfaction for 
is th
ty has a possible range fro
m 11 to 99; 
quality and quanti
as measured using a 7-poin
t scale; 
9 oa1 commitment w
z-scores. 
quality and quantity self-
efficacy are mean 
123 
(Table 9b, cont.) 
Goal Commitment 4.7 4.8 
for Quality (1. 0) ( 1. 0) 
Goal Commitment 4.8 4.8 
for Quantity (1. 0) ( 1. 1) 
Quality Self-efficacy -.03 .06 
( ? 8) ( ? 8) 
Quantity Self-efficacy -.06 ? 08 
( ? 7) ( ? 7) 
124 
Table 9c 
Mean s a nd Standard Deviat i on
s 
Enthusiasm: 
for Goal Variables? 
Means 
(standard deviations) 
High 
Dependent Low
 
asure Enthusiasm
 Enthusiasm 
Me
9.6 
Quality Goal 
10 . 4 
(7.9) (6.6) 
ity Goal 46.5
 45.4 
Quant
(16.3) (13.5) 
49.1 48 . 5 
Anticipated Satisfaction 
(13.8) (12.8) 
for Quality 
Ant? ? isfaction 6
1. l 60.5 
icipated Sat
(10.7) ( 9. 6) 
for Quantity 
ber of errors and quantity g
oal 
? Quality goal is the num
is the number of pages. An
ticipated satisfaction for 
uantity has a possible range
 from 11 to 99; 
quality and q
ment was measured using a 7-
point scale; 
goal commit
 z-scores. 
quality and quantity self-ef
ficacy are mean
125 
(Table 9c, cont.) 
4.8 4.7 
Goal Conmli tment  
(1. 0) ( 1. 0)
for Quality 
4.9 4.8 
Goal Commitment 
(1. 0) (1.
 0) 
for Quantity 
.oo .04 
y 
Quality Self-effic
ac
(. 8) (. 8
) 
.03 .oo 
Quantity Self-effi
cacy 
) 
( ? 8) (.7
126 
Table 10 
 
rrelations Among
 Goal Variables
Co
4. 5. 6. 
7 . 8. 
1. 2 . 3. 
_33?? -.24?? -.64
?? -.07 
1  - . 52
.. . 00 
? Quality Goal 
? 03
2 ? Quant i ty Goal 
35?? .31?? -.56
 ?? -.04 
.26?? -_
3 ? Anticipated S
atis-
faction for Qu
ality 
-.02 -.03 -.0
5 - . 61?? 
4 nticipated Satis
-
? A
tion for Quanti
ty 
fac a1?? .4o ... 15 .. .
5 ? Goal Commitme
nt 
for Quality 
6 Goal Commitment
 
? 
for Quantity 
7 ? Self-efficacy
 
for Quality 
8 ? Self-efficacy
 
for Quantity 
~<.os, ?? Q<.0
1 
127 
Table 11 
between Goal Variable
s and Independent 
Correlations 
and Performance Varia
bles? 
y_ 
4 .43 ... OJ 
or Quality .22 ... 08 .0
7 .28 .. -.0
Goal f
-.04 -.07 .43
 .. -.06 . 58 .. 
Goal for Quantity . 02 
.14? 
 
10? -.02 -.29 ... 
06 -.34 .. -.05
Anticipated satis-
-.08 - .
faction for Quality 1 .. 
3 -.13? -.02 -.05
 -.24 .. -.09 -.4
Anticipated satis-
-.0
faction for Quantity
 
.22 ... OJ 
Goa1 Commitment .14 ..
. 03 -.04 
for Quality 
.06 . 12? . 05 
- . 03 
Goal Commitment .08 
. 01 
for Quantity 
 07 .51?? .15 .. 
? 03 .36 ..
.
Self-efficacy .18 ...
 06 
for Quality 
.38 .. -.06 
efficacy .12? .15 ..
 -.04 -.08 
Self-
for Quantity 
12<.05, ?? Q < .Ol 
 - v? . = Task strategy; E = Enthu
siasm; A-QL = 
? V - lSJ.on; TS 
ance 
ality; A-QT= Ability Qua
ntity; P-QL = Perform
Ability Qu
Qua1? J.ty; P-QT = Performanc
e Quantity. 
128 
Table 12 
ce Results for G
oal Scales? 
rian
Analysis of Co-
va
Dependent 2 
df MS E 
w
Variable Source
 
281.7 2 5. o?? .07 
ity 1
 
Quality Abil
60.9 14. 3 ..
 .04 
(A) 1 1
Goal Vision 
1.3 
ask Strategy (B) 
1 15.1 
T
 
iasm (C) 1
 10.9 1.0
Enthus
1 1. 3 . 1
 
A X B 
8 ?? 
1 110.9 
9. . 03 
A X C 
1 37.7 3.3
 
B X C 
16.4 1.5 1 
A X B X C 
268 11. 3 
Residual 
+  ?? 12< .0l 
!2< .10,? ? Q< . 0
5,
nalyses due to 
e varies slight
ly across the a
? Sample siz  only for 
Omega-squared (w
2 )is presented
missing data. 
alues. 
significant E v
129 
(Table 12, cont.) 
12776.9 62. 5?? .18 
Quantity Ability 
5.6 . 1 
Goal vision 
(A) 1 
(B) 1 363.0 
8 . 2?? .02 
Task strategy 
41. 6 .9 
Enthusiasm (C) 1 
1 . 2 .o 
A X B 
220.6 5. o? .01 1 
A X C 
1 96 . 0 2.2 
B X C 
1 5.0 .1 
A X B X C 
265 44.4 
Residual 
14129.3 26.0 .. . 08 
Anticipated Ability 
1 189.8 1. 2 (A) 
Satisfaction Vision 
1 354.3 2.2 
for Quality Task strategy 
(B) 
. 2 
(C) 1 33.3 Enthusiasm 
1 74.7 
. 5 
A X B 
792.9 5. o? .01 1 
A X C 
. 9 
1 150.5 
B X C 
1 317.3 
2.0 
A X B X C 
266 158.7 
Residual 
130 
(Table 12, cont.) 
Anticipated Ability 11637.0 1 7. 3?? .05 
satisfaction Vision (A) 1 23.1 ? 2 
for Quantity Task Strategy (B) 1 561. 9 5. 9? .02 
Enthusiasm (C) 1 . 4 . 0 
A X B 1 399.4 4. 2? .01 
A X C 1 296.6 3.1 
B X C 1 43.4 . 5 
A X B X C 1 13.8 ? 5 
Residual 267 94.9 
Goal Ability 1 5.9 6. 4 ? .02 
commitment Vision (A) 1 4.2 4. 6? . 01 
for Quality Task strategy (B) 1 . 2 . 2 
Enthusiasm ( C} 1 .5 . 5 
A X B 1 . 9 1.0 
A X C 1 . 8 . 8 
B X C 1 1.8 1. 9 
A X B X C 1 . 1 .1 
Residual 265 .9 
131 
(Table 12, cont . ) 
Goal A 1 
.9 . 4 bility 
2 2.0 Co .m 1 mitment vision (A) 1 
Task strate (B) 1 .o . 0 for Quantity gy 
. 4 
Enthusi . a 4s  m (C} 1 
1 1. 5 1. 5 A X B 
1 3.4 3.3 A X C 
1 . 1 . 1 B X C 
A X B X C 1 
. 4 . 4 
Residual 264 
1. 0 
21.2 3 81 . 9?? .12 Quality Ability 
4.8 s. s?? .03 
Self-effi 1c  acy vision (A) 
1 . 3 . 6 Task strategy (B) 
? 7 
Enthusiasm (C} 1 
. 4 
1 .o .o 
A X B 
1 . 1 . 2 
A X C 
1 1. 8 3.3 
B X C 
1 . 4 . 6 
A X B X C 
.5 
Residual 2 51 
132 
(Table 12, cont
.) 
 13.4 3 2 . o?? .10 
Quantity Ability
 1
3.1 7. 4
 ?? .02 
1 
Self-efficacy Visio
n (A) 
 (B) 1 2.0
 4. g? .01 
Task Strategy
3 .8 
nthusiasm (C)
 1 . 
E
1 . 1 . 
3 
AX B 
6. 1 ? . 02 
1 2.6 
AX C 
1 .7 1
. 7 
B X C 
1 . 3 .
 7 
AX B X C 
. 4 
Residual 
255 
133 
Figure 6 
vision by Enthusiasm Interaction on Goal Variables 
(a) 150 
No Vl1lon 
?g 411 
? VIilon !;. 
'ii 
0 
c:, 
l:: 
i? ::,  
0 40 
High 
Enthu.lam 
(b) 1111 
-; 150 
ii VIiion 
::, 
0 
2 
& No VIiion 
i 411 
i 
'ii 
(I) 
"B?"   a. 
::? 
~ 40 
High 
Enthu.lam 
134 
(Figure 6, cont.) 
.s 
(c) 
.2 
! 
0 
?0  ., vi.ion I 
.!! 
~ 
8 
E 0 
J?.  
;g No vi.ion 
~ - . 1 
'ii 
6 
? .I 
-.S,L----__..---------+-----
High 
(d) 20 
111 
?...  V11lon 
~ 
!:!;!. 
ii 10 
& No vi.ion 
~ 
& 
e; 
High 
LoW 
135 
Figure 7 
Vision by Task strategy Interaction on 
Anticipated satisfaction for Quantity 
66 
->, ;: .C.,  
: V, i sion 
0..  
.2 60 
C 
.
--
2 No Vision 
0.. , 
.-.!.!,  rn 
"O 55 -G..>,  
.e-
0 
;: 
C 
< 
50 
Task Strategy 
No Task Strategy 
136 
Task strategy significantly affected personal goals for 
quantity, anticipated satisfaction for quantity, and self-
efficacy for quantity. The enthusiasm variable did not 
significantly affect any independent variable. 
Unlike the attitude and perception variables where 
ere were few interaction effects, there was a consistent 
th
vision by enthusiasm interaction for personal goal for 
quality and quantity, anticipated satisfaction for 
quality, and self-efficacy for quantity. Figures 6(a) -
G(d) show that when enthusiasm is 10w, vision has an 
effect compared to no vision. When enthusiasm is high, 
sion has less of an effect. High enthusiasm may have 
vi ? 
distracted the subjects' attention away from listening to 
e vision, while it helped keep students' attention 
th
during the no vision condition. 
Vision also interacted with task strategy to affect 
7
anticipated satisfaction, shown in Figure ? When a 
Vision was presented, anticipated satisfaction was the 
sa me regardless of whether a task strategy was gi? ven or 
not. (Note that anticipated satisfaction is negatively 
related to goal level.) when no vision was communicated, 
th ? ? t d 
e task strategy resulted in 1ower anticipa e 
satisfaction (i.e. higher standards) than compared to no 
task strategy. 
137 
imilar 
 variables di
d not show s
nt
The independ
e
g that the e
ffects 
performance, 
indicatin
 on 
interactions translated in
to 
 self-efficac
y were not 
on goals and se 
ormance. Th
erefore, the
s . . perf
imilar effec
ts on 
ctically sign
ificant. 
ra
nteractions 
may not be p
i  the vision, ally
tent, specifi
c
In sum, lead
ership con
lf-efficacy f
or 
e subjects' 
goals and se
influenced th ls and 
 variable inf
luenced goa
 strategy
quality. Th
e task
 had an effec
t 
y
uantity. Ent
husiasm onl
or q
Self-efficac
y f
self-efficacy
 in 
ipated satisf
action, and 
on goals, an
tic
th vision . 
interaction 
wi
sults 
.Mg_diation Re iators between
 
to be med
Goals have b
een shown 
, ? 1990). Due
 to 
e (Locke & La
tham
feedback. and
 performanc
en goals and 
performance 
nt relationsh
ip betwe
the consiste  (Locke & Lat
ham, 
 in previous 
studies
that has been 
found
 independent 
 relationship
 between the
199 0) and the  co-variance 
oals in the a
nalysis of
Variables an
d g
mediators of 
the 
 
s , goals were
 examined as
re ults ecause of the
 
formance rela
tionship. B
leadership -
per d 
een personal 
goals a n
tion betw
high interco
rrela f qua 1 or 1
? t Y and 
 ~<.01, 
nt? ? act
ion (K=-.52,
a icip ated sat
isf
t al., 1991),
 the 
ntity) (Mento
 e
~~-.57, ~<.0
1, for qua
 not examined
 as 
cales were
ticipated sat
isfaction s
an k of effect o
f 
 lac
on, due to th
e
lllediators. In 
additi
138 
variables on the goal commitment for quantity 
independent ? 
scales, this scale was dropped from the analysis. 
Baron and Kenny's (1986) criteria were used to test 
for mediation: (a) the independent variable must be 
i icantly related to the proposed mediator; (b) the 
sign'f' 
i ependent variable must significantly affect the 
nd
dependent variable; and (c) the effect on the independent 
variable on the dependent variable must be less when the 
mediator is controlled. when the independent variable has 
no effect on the dependent variable, perfect (or complete) 
mediation exists, and when the effect is reduced, partial 
mediation exists. using multiple regression techniques, 
all equations were run controlling for ability. Due to 
e lack of relationship between the enthusiasm variable 
th
a any of the goal scales as well as between enthusiasm 
nd 
a the performance variables (as shown in Tables 5 and 
nd 
lO), the first and second of Baron and Kenny's criteria 
were not met, and therefore enthusiasm is not considered 
in the mediator analysis. only the vision and task 
strategy variables were tested for mediation effects. 
Personal Goal as a Mediato_x. personal goals were 
O 1 
test e d as possible mediators with regar d t qua 1 ? t Y and 
quantity for the vision and task strategy variables, for a 
nd 
total of four tests. According to Baron a Kenny's 
Proced ? d for each of th f ure, three equations were use e our 
139 
tests of mediation. Each set of equations represented one 
test of mediation. Table 13 shows the results of these 
tests for personal goals as mediators. 
Quality goals were found to completely mediate the 
v?i sio. n-performance quality link (Set 1, Table 13). F
or 
set 2, quality goals could not mediate because task 
rategy was not significantly related to quality goal. 
st
Quantity goals could not mediate the vision-
performance quantity link (Set 3) because vision was not 
related to the quantity goal variable nor to performance 
Finally, quantity goals partially mediated the 
quantity. 
task strategy-performance quality relationship (Set 4). 
Self-Efficacy as a Mediatox- As with personal goals, 
self-efficacy was tested as a possible mediator between 
the vi? si?o n and task strategy variables an d per f ormance 
quality and quantity. The same procedures were used for 
these four tests of mediation. However, there were 
approximately 20 missing cases of data (due to these 
Jects not responding on the questionnaire) for the 
sub' 
self-efficacy strength variable, which is one of 
var?i ables, along with self-efficacy magn?i tu d e, forming the 
Self-efficacy scale. Although the vision-performance 
quality relationship was significant when testing for the 
main effect (shown in Table 6 and in Table 14, Set 1)' the 
140 
Table 13 
ersonal Goal"
 
iation Test f
or P
Med F for 
Dependent 
ble Variable E
ntered 
Y....ar ia rmance Quali
ty 
ality Goal - P
erfo
Set l: Visio
n - Qu  
. OB . 08 
22.9 ..
Quality Ability
 
12.5 .. 
.12 .04 
Goal Vision 
20 .20 67
. 1 ?? 
.
Performance Abili
ty 
 .01 3
.4 + 
.21
Quality Vision .09 3 4. 7?? .30 
Quality Goal 
. 1 ?? 
.20 .20 67
Performance Abili
ty 
. 10 3 7. 4
 ?? 
.30 
Quality Quality
 Goal 
 .00 o.o
 
.30
Vision 
+ 
<.lo, ? Q<.05
, ?? 12<.0l 
.12 same effects 
may vary 
values for th
e 
io 
? The R
2 and E-rat
due to missin
g 
regression e
quations 
s1?  the lghtly acros
s
 of the E-ra
tio 
se was the s
ignificance
data. In no
 ca
ht variation
. 
affected by 
this slig
141 
Table 13, con
t.) 
(
rmance Qualit
y 
o
tegy - Qualit
y Goal - Perf
Set 2: Task S
tra . 8 .. 
. 08 . 08 
37
Quality Ability
  
. 08 . 08 
o.o
Goal Task strat
egy 
.20 .20 67.
 1 ?? 
Performance Abili
ty 
. 21 .01 
ality Task Strat
egy 
Qu
31 .10 39. 
1 ?? 
.
Quality Goal 
.20 67. 1 ?? .20 
Performance 
Ability 
.30 .10 
y Quali
ty Goal 
Qualit .31 .01 
Task Strategy
 
- Performance Qua
ntity 
y Goal 
Set 3: Vision - Quantit
.18 . 18 
60. 4 .. 
Quantity Ability o.o 
.18 .00 
Goa1 Vision 
. 28 102.5 ..
 
.28 
Performance Abilit
y 
.28 .00 
o.o 
Quantity Vision . 15 70. 3 .. .43 
Quantity Goa
l 
.28 . 28 
102 . 5 .. 
Performance Ability
 70.5 .. 
. 43 .15 
y uantity Goal Quantit Q  .00 o.o .43
Vision 
142 
(Table 13, cont.) 
egy - Quantity Goal
 -
Set 4: Task Strat
Performance Quantit
y 
.18 . 18 60. 4 ..
 
Quantity Ability 
.20 .02 
Goal Task Strategy 
.28 . 28 102.5 ..
 
Performance Ability 
2 ?? 
.34 . 06 2 4.
 
Quantity Task strategy 
.46 . 12 61.8 .
. 
Quantity Goal 
8 . 28 102.5 .. .2
Performance Ability 
.43 . 15 72. 
8 .. 
Quantity Quantity Goal 
.46 . 03 14
. 8 ?? 
Task Strategy 
143 
esulted the vi
sion-
e missing case
s r
effects of the
s
ficant 
tionship becom
ing non-signi
la
Performance q
uality re  
efficacy as a 
mediator. To
en testing qua
lity self-
Wh ng cases were 
main effect, t
he missi
replicate this
 
cy magnitude score 
 each subject'
s self-effica
replaced with
 magnitude wer
e 
 self-efficacy
 strength and
because
ean K=.71, ~<.0
1). (Note: an 
orrelated (m
significantly 
c
sample mean se
lf-
he 
ative procedur
e of using t
altern
the missing sc
ores 
 lac
e 
eff"icacy strength
 score to rep
sult 
to this proced
ure). The re
 identical 
Yields results analysis of va
riance 
e main effect 
found in the 
Was that th  
he mediator an
alysis and the
s replicated i
n t
Wa he 
other relations
hips i n t
significance l
evels of the 
gde The "re
placed" self-. 
mediator.  
? nchan
ana 
1ysis were u  
ll tests of me
diation, shown
les were used 
in a
efficacy sca
in Table 14 .. 
mediated the 
y self-efficac
y completely 
Qualit  
nd performance
 quality (Set
a
elationship be
tween vision 
r te the 
f-efficacy cou
ld not media
l, Table 14). 
Quality sel
tionship (Set 
2) due 
ity rela
rformance qua
l
task strategy-p
e
y and 
ionship betwee
n task strateg
lat
to the lack of
 re
gua1? acy. ity self-effic
144 
Table 14 
r Self-efficacy? 
Mediation Test f
o
F for 
Dependent 
~ R2 
Y...ariable Variable Ent
ered 
lity Self-efficacy
 
Set l: Vision -
Qua
formance Quality Per
.13 .13 
3 7. 8 ?? 
Quality Ability 
.16 .03 
-efficacy Vision
 
Self
. 20 69. 3 ?? .20 
Performance Ability 
1 .01 3.5 +
 
.2
Quality Vision 
. 12 48.9 .. 
cy .33 
Quality Self-eff
ica
.20 . 20 6
9.3 .. 
Performance Ability 
 . 13 53. 2
 .. 
cy .33
Quality Quality Self-
effica
.33 .00 0.0 
Vision 
+ ? 12< .0l 
!2< .lo, ? .Q< .05, 
?
 the same effects
 may vary 
-ratio values for2 ? The R and E ing 
 equations due to
 miss
Slightly across 
the regression
data. 
145 
) 
(Table 14, cont
.
Self-efficacy -- Quality 
Set 2: Task St
rategy 
ormance Qualit
y 
Perf
.8 .. 
. 13 .13 37
Quality Ability 
.13 .00 o.o 
elf-efficacy T
ask Strategy 
S
.20 . 20 6
9. 3 .. 
Performance Ability 
 .01 3.5 +
 
.21
Quality Task Strate
gy 
. 8 .. 
fficacy .3
4 . 13 53
Quality Self-e
.20 ? 20 6
9.3 .. 
Performance Ability 3 53. 2 ?? 
cacy .33 
. 1
Quality Quality Se
lf-effi
. 1 ? 
.34 .01 4
Task Strategy 
- Quantity self
-efficacy -
Set 3: Vision 
mance Quantity
 
Perfor
.14 . 14 43
.8 .. 
Quantity Ability 
.16 .02 
lf-efficacy Vi
sion 
Se
6.1 .. 
.24 . 24 8
Performance Ability 
.24 .00 o.o 
Quantity Vision .09 36.8 .. 
self-efficacy 
. 33 
Quantity 
.24 .24 86. 1 ?? 
Performance Ability . 08 32.4 .. 
icacy .3
2 
Quantity Quantity 
self-eff
.33 .01 4. 1
 ? 
Vision 
146 
e 14, cont.)
 
(Tabl
elf-efficacy
 -
k Strategy -
Quantity S
Set 4: Tas
Performance 
Quantity 
.14 43.8 ..
 
. 14 
Quantity Abil
ity 
.17 .03 
strategy 
Self-effica
cy Task 
6. 1 ?? 
.24 .24 8
Performance
 Ability 06 23.6 .. .30 . 
trategy 
Quantity Task
 S
. 07 3 0. 4
 ?. 
 
 Self-effic
acy .37
Quantity
. 1 ?? 
.24 .24 8
6
lity  .. 
Performance
 Abi
.32 . 08 
32.4
y 
y Quantity S
elf-efficac
Quantit .37 . 05 
21.7 .. 
Task Strate
gy 
147 
Self-efficacy for quantity did not mediate the 
ision-performance quantity relationship (Set 3), although v? . 
sion affected self-efficacy for quantity, which in turn 
vi ? 
affected performance quantity. This model can be called a 
inking" model, in the sense that self-efficacy for 
II 1 , 
quantity links vision to performance quantity. Finally, 
self-efficacy quantity partially mediated the task 
rategy-performance quantity relationship (Set 4). 
st
Goal Commitment as a MediatOJ:? Goal commitment for 
quality was tested as a mediator between vision and 
performance quality. Table 15, set 1 shows that goal 
mitment for quality completely mediated the vision-
com? 
performance quality relationship. 
Because goals and self-efficacy were also found to 
iate between vision and performance quality, a test of 
med' 
joint mediation was conducted to determine the effects of 
goal commitment above those of goals and self-efficacy. 
Table 15, set shows that when goals and self-efficacy 
2 
are added into the regression equation before goal 
comm ?i tment, the effect of goal commitment i? s no t 
Goals and self-efficacy still mediate the 
significant. When goals and 
Visi' on-performance quality relationshl'P ? 
self-efficacy are added into the regression equation after 
goal commitment (Table l5, set 3), all three variables 
medi? ate the vision-performance qualitY re 1 ati?o nship. 
148 
ough goal commitment does serve as a m
ediator, goals 
Alth 
se f-efficacy are more fundamental med
iators than goal 
and 1 
commitment. 
?..onclu si?o n 
 
Proposition 7, which concerned goals a
s a mediator,
rally supported. proposition 8, which 
concerned 
Was gene
as a mediator, was also generally suppo
rted. 
Self-effi' cacy 
ound 
exploratory, consistent relationships w
ere f
Although 
goal-related variables and leadership c
ontent. 
between 
ance quality relationship was completel
y 
rform
The vis
i' on-pe
mediated by quality goals, quality sel
f-efficacy, and goal 
erformance 
comm ?itment for quality. 
The task strategy-p 
Y relationship was partially mediated by q
uantity 
quantit 
goals a nd quantity self-efficacy. 
 
The same pattern of findings were obse
rved in the
of variance results (Table 12). As sho
wn in 
analysis 
 2 ' vision was mainly related to
 the quali. ty 
Table 1
(for personal goal, goal commitment, an
d self-
Variables O 
th 
effl' cacy) while task strategy was onlY
 re 1 a t e d t e 
variables (for personal goal, anticipat
ed 
quantity 
sat?l sfaction, and self-efficacy). Wh
en the leader 
information about ideal performance sta
ndards, 
Provided 
such as with a vision of quality
, subjects set more 
ality 
diff i' cult goals, were more 
committed to their qu
quality 
goals , had higher self-eff
icacy regarding the 
149 
Table 15 
Mediation Test for Goal Commitment? 
F for 
Dependent 
Variable Variable Entered 
set 1: vision - Goal commitment for Quality -
Performance Quality 
GC-Quali tyb Ability .02 
.02 
Vision .04 .02 
67. 4 ?? 
Performance Ability .20 
. 20 
Quality Vision .21 .01 
3.1 + 
Ge-Quality .23 .02 
67. 4 ?? 
Performance Ability .20 .20 
Quality Ge-Quality .22 .02 
Vision .23 .01 2 .1 
+ ?<. 1 o, ? g<.05, ?. g<.01 
? The R2 and K-ratio values for the same effects may vary 
slightly across the regression equations due to missing 
data. 
b GC-Quality = Goal Commitment for Quality 
150 
 15, cont.) (Table
elf-efficacy
 
d S
Mediation o
f Goals an
Set 2: Join
t 
t for Quali
ty 
commitmen
entered bef
ore Goal 
.22 .22 
rformance Ability
 
Pe .23 .01 1.
 9 
Quality Vision 
Quality Goa
l 
? 3 7 ? 14 
27. 3 .. 
uality Self-
efficacy 
Q
? 3 7 ? 00 0
.0 
GC-Quality 
.22 .22 
Performance
 Ability 
Quality Quality 
Goal 28.3 .. 4 
elf-efficacy
 .37 .1
Quality S
.37 .00 
0.0 
GC-Quality .0 .oo .oo 0
Vision 
151 
(Table 15, cont.) 
t 3: Joint Mediation .of Goals and Self-effi
cacy 
Se
entered after Goal Commitment for Quality 
 70.6 .. 
Performance Ability .22 .2
2
Quality Vision .23 .01 1
.9 
GC-Quality .24 .01 4.
 5? 
Quality Goal 
Quality Self-efficacy . 3 7 . 13 24.
5 .. 
Performance Ability .22 . 22 70
.6 .. 
Quality GC-Quality .24 .02 5. 5? 
Quality Goal 
Quality Self-efficacy .37 .13 25.2 .. 
Vision .00 .oo 0.0 
152 
 quality. W
hen 
 .  had higher 
performance
Of their work, a
nd
gy, in this
 case, 
k strate
th tas
e leader pr
ovided a 
ubjects set 
 to perform 
the task, s
ow
suggestions
 on h
wer satisfac
tion 
ticipated lo
higher oals for qu
antity, an
g
ing their pe
rformance 
ndards) reg
ard
e . , higher 
sta
 the quantit
y 
(i. arding
efficacy reg
ntity, had h
igher self-
qua
rmance quan
tity. 
 higher perf
o
of Work, an
d had
 mediated th
e vision-
nd self-effi
cacy
a
In sum, goa
ls -
p and the ta
sk strategy
lity relation
shi
a
Performance
 qu high 
onship. In
 order for 
i
ormance qua
ntity relat
f  must resul
t i  
Per t'ons
n
to exi'st , the lea
der's ac 1 
 
Performance  having high
 
als and
the f 1 rs setting d
ifficult go
0 lowe
self-eff i' cacy. 
153 
Chapter 8 
Integration of Quantitative Results 
To provide an integration of the results p
resented in 
hapters 6 and 7, three path models are pre
sented. The 
c
firS , initial model and the second, revise
d model 
t
ra t e the re
sults from chapter 6. The third, final 
integ 
incorporates the exploratory results from c
hapter 7 
model ? 
e revised model. These models, especially 
the 
into th 
t the essence of this dissertation's 
f'i nal model, represen
 to summarize the previous 
find i' ngs, and are inte
nded
chapters i?n a parsimonious manne
r. 
For purposes of all analyses in this chapt
er, 
c ion terms among the independent variables
 were 
intera t' 
dropp ed from consideration
 because of their lack of 
ance, 
cons?i
pact on the perform
 stent and meaningful im
erception variables (see chapter 6). In 
attitude , 
and p
ition, due to the high correlation between 
enthusiasm 
add ' 72
task and enthusiasm for the vision (~= - , 
Q<.01, 
for th e 
in Table 4), these variables were combined 
into a 
shown . th
on enthusiasm. Also, e 
Singl e scale, task and vi
si
ingness to work for the leader variable and
 the two 
Will ' 
mooa variables were dropped bec
ause theY were not 
ated to the independent variables or to 
meaning fully r
el
the Performance variables. 
1 5 4 
A variation of path ana l ysis was used to develo
p an 
exploratory path model (Kenny, 1979) ? In the m
odel, shown 
igure 8, the independent variables (leader acti
ons), 
in F
coed as a zero-one dummy variable. Ability (s
hown 
Were d 
as Cl for quality and c2 for quantity in the p
ath models) 
se as a covariate in al l structural equations was u d 
(Kenny, 1979). The equations were computed (a)
 with each 
iable 
attitude and perception variabl e as the depend
ent var
performance variables and manipulated leader and the 
 the independent var i ables and (b) w?i th the eac
h 
actions as
Performance variable as the dependen
t ~ari? a bl e an d 
manipulated leader actions as the independent v
ariables. 
path coefficients are the standardized regressi
on 
The 
icie. nts from the structura
l regression equations. 
COeff' ?
 purposes of pr
esentation, only paths with statistical 
For
icance- at the .lO level or below were retained.
 
signif' 
Initial Path Model 
th
 conceptualize the causal relationships in e To
path model, it was assumed that in the leadersh
ip 
initial 
Process , the leader engages in certain ac
ti.o ns, wh.i ch 
nd 
affect the actions (performance), attitudeS, a
However, during the 
Percept?i ons of the followers. th
expe ri.m ent, the leader's actions wer
e followed by e 
SUbjecst , performance which was followed by th e 
155 
measurement of attitude and perception variables. Th
is 
timing of events allows us to test whether this causal 
sequence is viable but does not allow us to test for other 
causal sequences, such as whether attitudes and 
perceptions cause performance. Thus, a three-stage model 
was tested. The model is shown in figure 8 and the 
standardized path coefficients are shown in Table 16. 
Figure 8 shows that vision and task strategy directly 
affect performance quality and quantity. Vision directly 
affects seven attitudes and perceptions. Task strategy 
directly affects the subjects' feeling of intellectual 
stimulation. Both vision and task strategy indirectly, 
through performance, affects five attitudes and 
perceptions. Enthusiasm did not directly affect 
performance but did directly affect liking the leader and 
perceived leader charisma. 
Performance quality was directly related to task 
satisfaction, task and vision enthusiasm, and task 
clarity. Performance quantity was directly related to 
task clarity and inspiration. 
The independent variables mainly affected attitudes 
and perceptions toward the leader but not task attitudes I 
while the performance variables mainly seemed to affect 
the task attitudes but not attitudes and perceptions 
toward the leader. It is clear that two distinct sets of 
156 
Figure 8 
ial Path Mod
el" 
Init
d Perceptions
 
Attitudes an
ons Per
formance 
leader Acti
sk Satisfaction
 
Ta
'c2 
Task & Vision 
Vision Enth
usiasm 
~ rmance Quality Perfo Task Clarity 
' c t 'ct 'c2 
Congruence of
 Beliefs 
 gy / & Value
s 
Task Strate
~ uantity ader Performance Q Trust in Le
'ct..._ C2 
er 
Liking the Lea
d
Process 
Charisma 
n 
Intellectual Sti
mulatio
'c1 
Inspiration 
t 
fficients ar
e no
o:r presentation
, path coe
" F' PU:rposes o f ance o f .10 fi c
  with statis
tical signi
Shown a nd onl y path
s
 qual i t y is 
y covariate 
f or
O:r b low Th
e abil it
e are shown. 
 ntity a s c 2 . Show f or q ua
n as Cl and 
157 
Table 16 
Standardized Path Coefficients for Figure 8 
Independent Dependent Variables? 
Variables P-QL P-QT SAT T&V CLAR CONG 
0 
Performance Quality .34 ?
0 .12+ .29? .00 
0 
Performance Quantity .15? .09 .19? .00 
.05 . 11+ .00 ? 3 3 ?? Vision 
Task Strategy .11? . ?21?? .08 -.03 .06 .09 
Enthusiasm .05 -. 07 - .04 .04 .08 .03 
Clb ? 4 8.. ? 2 3 ?? ? 0 6 ? 0 3 . 21 ?? . 04 
C2b .07 ? 22 ?? - ? 10 
.23 .36 .20 .04 .29 .13 
? P-QL = Performance Quality, P-QT = Performance Quantity, 
SAT= Task Satisfaction, T&V = Task & Vision Enthusiasm, 
CLAR= Task Clarity, CONG= Congruence of Beliefs & Values 
b Cl= ability co-variate for quality; C2 = ability co-
variate for quantity 
158 
Table 16, c
ont.) 
(
riablesb 
nt Dep
endent Va
Independe STIM INSP 
LIKE CHAR 
~ 
TRST 
8 
.09 .00 
.00 .0
.10 
Performance
 Quality 
.02 .05 
. 13? 
ormance Quantity .1
0 .03 
Perf . 24 ?? . 3 4 ??  45 .. 
 . 17 .
. . 13 .
Vision .. . 03 
. 02 .09 
.26 
Task Strategy 
.06 
 ?? .07 .07 
. 18 .. . 1
4
Enthusiasm 
. 06 
 -.13 + -
.07 
.05 -.02
Cl .04 
-.08 -.04 
.00 -.01 
.07 
c2 
.15 .15 
R2 .07 .07
 .24 
11s error 
Leader, CHA
R = 
king the 
b T  in Leader,
 LIKE= Li
T ~ Trust ulation, INSP RS
IM = Intelle
ctual Stim
Y)' ST
Charisma (F
orm B
::::: Inspirati
on 
159 
attitude and perception variables emerged, a set of task 
attitudes (task satisfaction, task and vision enthusiasm, 
and task clarity) and a set of attitudes and perceptions 
toward the leader (congruence of beliefs and values, truSt 
in leader, liking the leader, charisma, intellectual 
stimulation, and inspiration). The possibility that these 
two sets of variables could emerge as meaningful 
components was examined by conducting a principal 
components analysis on the attitude and perception 
measures and then re-analyzing the path model using the 
resulting components. This is discussed below. 
Revised Path Model 
The results of the principal components analysis are 
presented in Table 17. This confirmed the presence of a 
task attitudes component and an attitudes and perceptions 
toward the leader component. These components were 
computed as scales, task attitudes and 
attitudes/perceptions toward the leader, and were entered 
as dependent variables in the path model, as shown in 
Figure 9. The standardized path coefficients are shown in 
Table 18. 
Vision and task strategy variables affect task 
attitudes indirectly through the performance variables. 
Also, the vision and task strategy variables as well as 
160 
Table 17 
de and 
Components Ana
lysis of Attitu
Principal 
erception Varia
bles 
P
Leader-Task-
ted Relat
ed 
Rela
Variables variables 
t? Variabl
e 
-At itude/Perce
ption .14 
sk Satisfaction
 
Ta .31 .69 
v ? d Task Enthusiasm lsion an -.01 .64 
Task Clarity 1 
liefs and Value
s .0
Congruence of 
Be
.25 
Trust in the L
eader .71 
. 04 
Liki' ng the Leader
 
.21 
Leader's Charis
ma .70 
.13 
ectual Stimula
tion 
Intell .73 .25 
Inspiration 
161 
Figure 9 
Revised Path Model" 
Performance Attitude/Perception Factors Leader Act Iona 
Vision 
~ Performance Quality ---1~- Task Attitudes 
/ ""--c:1 ""--c:2 Task Strategy 
~ Attitudes/Perceptions Performance Quantity 
Toward Leader 
...._C1 ...._C2 
Process 
" For purposes of presentation, path coefficients are not 
shown and only paths with statistical significance of .10 
or below are shown. The ability covariate for quality is 
s hown as Cl and for quantity as C2. 
162 
Table 18 
Figure 9 
th Coeffic
ients for 
Standardiz
ed Pa
Variables? 
Dependent 
nd
I ependen
t TA A&P-L 
? P-OL
 P-OT 
Var
~ . 33 ?? .03 
lity 
Performanc
e Qua
.19 .. .06 
tity 
Performanc
e Quan  .. 
2 . 08 
.37
-.0
Vision . 14 ? 
.27 .. . 05 
gy 
Task Strat
e
 . 14
 ? 
 . 04
05 -.07
Enthusiasm
 .
03 
. 2 3 ?? . 13 ?
 -.
Clb 
. 5s?? . 1
4 ? .oo 
.07 
 .20 
6 .25
.23 .3
rmance Qu
antity; 
P-QT = Perfo
? P-QL -
 
ormance Q
uality; P t? - Perf  & ercep 
ions 
= Attitudes
TA~ Task 
P-L 
Attitudes;
 A&
Towa rd Leader ? C2 = abili
ty co-
l'ty
b  ariate for
 qua l , 
 abi'li' ty co-vCl~
"ariat for quanti
ty. 
e 
163 
en usiasm directly affected attitudes and perc
eptions 
th
But, enthusiasm di'  d not h
ave any 
toward the leader. ? 
e ects on performance or ask 
direci or i'ndi'rect ff t 
attitudes. This model represents an integratio
n of 
results from chapter 6. 
Final Path Model 
The exploratory mediation results were integrat
ed 
revised path model by analyzing a fi?n al path with the 
The goal and self-efficacy variables were inser
ted 
model. 
between 1 eader actions (or independe
nt vari? a bl es) and 
Perform ance because they w
ere found to med?i ate between 
Although goal comm?i tm en t f or qua li' ty  was
 
these vari'ables. 
und t be a
 mediator, goals and self-efficacy for 
fo 0 
quality were more fundamental mediators than go
al 
commitment. Thus, goal commitment f or qua 1i? ty 
wa s om'i tte d 
from t he path model. 
This resulted in a four-stage model, with leade
r 
actio ns affecting the goals and self-effi
?c acy, wh i' ch i? n 
n affect performance. performance, in turn, af
fects 
tur 
itudes and perceptions. This model is shown in
 Figure 
att? 
lo ' and the standardized pat
h coefficients are shown in 
Table 19. 
This final model represents the essence of the 
rt integrates the all 
find?i ngs o f this dissertation. 
164 
Figure 10 
Final Path Model" 
Attitude/Perception 
eader Action? Goal? & Self-efficacy Perform
ance Factor? 
L
Vision LQ uality Goal 
~ 'c1 ~ Performance ---+- Task Attitudes 
\ '\ Quality Self- Quality 'c1 'c2 
efficacy 'c1 
'c1 'c2 
Task St,ategy ~ 
Quantity Goal Attitudes/ 
'c1 Performance Perceptions 
Quantity Self- Quantity Toward Leader 
efficacy 'c1'c2 
Process 'c1 'c2 
r purposes of presentation, path coefficients are 
not 
" Fo
shown and only paths with statistical significance
 o f .10 
or below are shown. The ability covariate for qual i t y i s 
s hown a s Cl and for quantity a s C2 . 
165 
Table 19 
Standardized Path coefficients for Figure 10 
Independent Dependent variables? 
Variables G-QL SE-QL G-QT SE-QT 
Quality Goal 
Quality Self-efficacy 
Quantity Goal 
Quantity Self-efficacy 
Performance Quality 
Performance Quantity 
Vi?ion . 21 ?? ? 17?? . 02 
Task Strategy .07 . 06 .16? . 13 ? 
Enthusiasm .05 .04 -.05 -.04 
Clb ? 3 0 .. .43 .. -.08 - . 1s? 
C2b ? 3 g??? .07 
.13 .20 .21 .17 
? G-QL = Quality Goal; SE-QL = Quality Self-efficacy; G-QT 
= Quantity Goal; SE-QT= Quantity Self-efficacy. 
b Cl= ability co-variate for quality; C2 = ability co-
variate for quantity 
166 
(Table 19, cont.) 
Independent Dependent Variablesc 
Variables P-QL P-QT TA A&P-L 
Quality Goal . 14 ? .06 .12 + .06 
Quality Self-efficacy .34 .. .00 . 02 . 07 
Quantity Goal -.11 .04 .26 .. 
Quantity Self-efficacy -.04 .04 .08 - . 02 
Performance Quality .27.. .01 
Performance Quantity .11? -.05 
Vision -.01 -.02 .05 ? 3 2 ?? 
Task Strategy .11? .18 ... 05 .07 
Enthusiasm .04 -.03 .02 . 14 ? 
Cl .31 .. -.17 .. -.08 -.05 
C2 .04 .4o?? . o9 -.04 
.40 .51 .28 .24 
c P-QL = Performance Quality; P-QT = Performance Quantity; 
TA= Task Attitudes; A&P-L = Attitudes & Perceptions 
Toward Leader 
167 
results into one diagram. Consistent with the mediation 
results, vision indirectly affects performance quality 
through quality goals and quality self-efficacy. Task 
strategy indirectly affects the performance quantity 
through quantity goals. Task strategy also directly 
affects quantity self-efficacy as well as both performance 
variables. With respect to the rest of the model, it is 
similar to the revised model. Vision, task strategy, and 
enthusiasm each directly affect attitudes and perceptions 
toward the leader. Task attitudes are only affected 
indirectly by vision and task strategy, through 
performance, goals, and self-efficacy. 
In sum, with respect to goals, self-efficacy, 
performance, and task attitudes, leadership content is 
important. That is, performance was only affected by 
content and not by process. More specifically, the 
vision, which was a vision of quality, affected quality 
variables (goals, self-efficacy, and performance). Task 
strategy, which was concerned mainly with quantity and to 
a lesser degree with quality, affected quantity variables 
(goals, self-efficacy, performance). Vision and task 
strategy affected task attitudes indirectly. With respect 
to attitudes and perceptions toward the leader, all three 
leader actions were influential. 
168 
mary: Conceptual
 Model 
Sum
, a conceptual m
odel 
the results
To further simp
lify 
11. This figure
 
igure 
 the findings is
 presented in F
of
ps found in this
 
e of the relatio
nshi
shows the essenc
e content of a 
n at a more gen
eral level. Th
dissertatio
ffects followers
' 
matic leader's 
communication a
charis t followers' 
f-efficacy, whic
h in turn affec
goals and sel ' task 
ance in turn aff
ects followers
rm
Performance . P
erfo
process affect 
es. Both the c
ontent and the 
attitud e leader. 
udes and percep
tions toward th
followers' attit nt and 
els, it is clear 
that both conte
From these mod re 
but the patterns
 of effects a
Process have ef
fects, 
adership aff ects
 
In other words, 
charismatic le
different. ? ts a?l? f 
t 
des, but i  
 f eren 
both ttituPerf? ormance and 
a
different ways . 
components work 
in 
169 
Figure 11 
Conceptual Model 
ttitudes 
ntent- G nd Self-efficacy - P
erformance -Task A
Co oafs a
Attitudes/ 
Perceptions 
Process Toward Leader 
170 
Chapter 9 
Qualitative Res
ults 
e propositions of th
ualitative data 
supported many 
Q
cts were thinkin
g 
o what the subj
e
nd 
a Provided insi
ght int
du ? ring the experim
ent. 
' Questions Duri
ng Experiment 
Subjects
ded any 
ession, the exp
erimenter recor
After each s uring the 
the subjects ask
ed the leader d
questions that 
 certain conditi
ons 
hether
eriment in orde
r to examine w
exp
itions. Overall
, 
cond
 to more questio
ns than other 
led  this total, 17 s. Of
SUb' der 31 questi
on
Jects asked the
 lea
score the 
ring procedures 
(e.g. how do I 
concerned sco he 
b ' ?), 3 
concerned t
i n
d rer? what counts
 as an erro
r complete each item
?), 5 
., do 
questionnaires 
(e.g
ent i? 
d 
experim tse 
lf ( e.g. owe 
concer  
the 
ned the process
 of
d the type of w
ork 
cerne 
ork on binder 2 
or 3 now?), 6 con
W ?  BKC' s 
oes BKC use recy
cled paper. are
done by BK C (e.g, d . 
ost questions we
re 
Work hus,
 m
Processes autom
ated?). T
with 
m t ? questions about 
how to proceed 
infor a ion-seeking 
the experiment. 
ere asked in the 
no vision 
Fewer questions
 w
nct ? t? vision condition
s (10 and 21 
co  ions than the i
ossibly indicati
ng that the 
questions, resp
ectively), p
171 
jects to think about th
e task, 
Vision led the sub
lation. Slightly fewer
 
increasing their intel
lectual stimu
 the no task strategy v
ersus task 
questions were asked i
n
uestions, respectively
), 
rategy conditions (13 
and 18 q
st
s (12 and 
in the low versus high
 enthusiasm condition
and 
al 
19 stions, respectively). 
An approximately equ
que
le 
uestions were asked of 
the male versus fema
number of q
tively). An overall X2 tes
t 
leader (14 and 17, res
pec
indicated a 
(combining across male 
versus female leader) 
ween experimental cond
ition 
ificant relationship b
et
sign
= 9.25, ~<.05). 
and the number of ques
tions asked (X2 
Subjects' Comments 
f the experiment, subje
cts were asked to 
At the end o
ences . 
Write any comments abo
ut their experi
.fu!spicious Subjects 
t and leaders were inte
nded to be as 
The experimen
 
s possible. rt would 
have been difficult to
realistic a
ht that the 
di' 
he subjects thoug
rectly measure whether
 t
stionnaire items 
leaders were real or c
redible. Direct que
ect, even if asked at 
the end of the 
to this eff
bjects that the leaders
 
experiment, could have
 cued the su
kely that this 
Were actually confeder
ates. It is quite li
 subsequent sessions if
 these 
Would have contaminate
d
is to other subjects. 
subjects would have co
mmunicated th
172 
ness was only assessed via su
bjects' written 
So, suspicious
In general, what 
comments to some general ques
tions (e.g., 
[leader]?). From these writte
n 
did you think of the 
comments, 22 (of the 282) sub
jects indicated 
suspiciousness about the lead
er. Seven of these 22 
comments were in the no visio
n/no task strategy/low 
sion 
enthusiasm cell. Also, 15 we
re in the no vi
e that a leader's expression condition. This may indicat
of a vision was congruent wit
h the subjects' ideas of 
normal" leader behavior where
as a leader's communication 
"
of factual information was in
congruent with subjects' 
eless, it was encouraging 
ideas of leadership. Neverth
ly expressed some 
that only 8% of the subjects 
spontaneous
degree of suspiciousness. Th
ere were no differences 
tor regarding suspiciousness 
between the male and female a
c
(12 and 10 comments, respecti
vely). 
L.?.ader Influence 
When asked whether the leader
 influenced their 
mainly on the vision and 
Performance, responses depend
ed 
n the 
task strategy variables and, 
to a lesser extent, o
enthusiasm variable. sixty-tw
o subjects wrote that the 
on their performance. The 
leader had a positive impact 
g a vision 
leader had a more positive im
pact when expressin
no vision (58 and 4 comments, 
respectively) 
as compared to 
to no 
a nd 
ompared 
When communicating a task str
ategy as c
173 
nd 22 comments, respec
tively). Some 
task strategy (40 a
s that indicate this ar
e the 
representative commen
t
following: 
do my best! 
She kept expressing qu
ality ... I wanted to 
ntion and do a good jo
b. (Vision, 
She made me pay atte
task strategy, high en
thusiasm) 
, so he made it more 
Yes, he seemed to care
 about it
 was thinking "speed &
 
st I
interesting. Yes, at 
fir
e 
racy." Then I thought 
just "accuracy" and th
accu
(Vision, task strategy,
 low 
speed followed. 
enthusiasm) 
igation to 
She made me feel as tho
ugh it were my obl
ectively and efficientl
y. 
perform this task eff
Vision, no task strateg
y, high enthusiasm) 
(
t the task was not as 
The CEO made me realiz
e tha
g helpful hints about 
difficult as I thought
 by givin
me to 
to carry on the experim
ent. The CEO helped 
how 
 
 by offering a new way
 to complete the task.
do better
n, 
ppeared to be a stickl
er for quality. (Visio
He a
siasm) 
task strategy, low ent
hu
174 
y encouraged us to excel 
as well as 
Yes, she reall
e tasks. (Vision, 
[gave] us tips on how to 
complete th
k strategy, high enthusia
sm) 
tas
 important it was, and it
 made me take 
He told us how
e reminded me that ' 
the experiment more serio
usly. H
re not acceptable. 
mistakes, no matter how 
few, a
done carefully. (Vision, 
no task 
Work should be 
strategy, low enthusiasm)
 
er did not 
Similarly, 38 subjects w
rote that the lead
n a 
ence their performance o
r did influence it but i
influ
r. Most of these comment
s were in the no 
negative manne
ts, 
Vision as compared to vis
ion (19 and 9 commen
were in the low enthusias
m as 
spectively) condition an
d 
re
iasm condition (28 and 10
 comments, 
compared to high enthus
ts indicating 
respectively). Some rep
resentative commen
these findings as are fol
lows: 
did not influence me abou
t the task at all, 
The CEO 
y. No, during the task 
either positively or neg
ativel
 in the room or 
I didn't think about the 
CEO being
hing he said before the e
xperiment. (No 
about anyt
husiasm) 
vision, no task strategy
, low ent
175 
e was very boring and a wa
ste of time. 
I thought sh
 other way. (No 
She actually turned my ent
husiasm the
sk strategy, low enthusiasm
) 
vision, ta
was challenging myself and
 could [have) cared 
No, I 
trategy, low 
less about him. (No vision
, no task s
enthusiasm) 
't important. 
She made it seem that it r
eally wasn
project was self-motivated
. 
Any interest in doing the 
(No vision, no task strateg
y, low enthusiasm) 
tistical findings that 
These comments support the
 sta
he 
e vision had the greatest 
impact on performance. T
th
sed attention on quality a
nd away from 
Vision focu
 to 
tity, showed why the task w
as important, and served
quan
 to work toward it. To a jects
motivate and inspire the s
ub
tegy information helped th
e 
lesser extent, the task st
ra
a low 
bjects figure out how to d
o the task better, while 
su
thusiasm leader was not pa
rticularly liked by the 
en
subjects. 
Attitudes Toward the Task 
mments about the 
Subjects were asked to wr
ite their co
rote 
sk or experiment in gener
al. Thirty-one subjects w
ta
th t the task was fun, challen
ging, and 
at they though
 condition (23 
interesting. This was rel
ated to the vision
176 
comments in vision and 8 in no vision) and the task 
strategy condition (23 comments in task strategy and 8 in 
no task strategy). These comments were unrelated to 
enthusiasm (14 comments in low enthusiasm and 17 in high 
enthusiasm). Some representative comments are the 
following: 
I thought it was interesting and really wanted to be 
able to finish one of the binders or at least get 
about half way through with O errors! (Vision, task 
strategy, high enthusiasm) 
The experiment motivated you to do good [sic] and 
improve as you went along. It fired you up as you 
got busy. (Vision, task strategy, high enthusiasm) 
Good experiment - I enjoyed the challenge. (Vision, 
task strategy, high enthusiasm) 
I felt it was a good experiment for me as well as 
helping out the company. (No vision, task strategy, 
high enthusiasm) 
I liked the experiment. It was kind of fun. (Vision, 
task strategy, low enthusiasm) 
177 
ing. Actually, is a chance f
or me to 
Interest
d 
exercise a part of my brain t
hat hasn't been use
, task strategy, low enthusias
m) 
lately. (Vision
periment was quite good. The 
experiment showed me, 
Ex
may be there is a lot 
no matter how simplistic a ta
sk 
put into it prior to 
of thought process that must 
be 
execution. (Vision, task strat
egy, high enthusiasm) 
worthwhile experience. (Visio
n, task 
It was a 
strategy, high enthusiasm) 
ments 
Similarly, most of the 47 sub
jects who wrote com
ting that the task/experiment
 was difficult, boring, 
indica
o vision condition (29 commen
ts 
and tedious were in the n
in no vision and 17 in vision
). There were small 
tegy variable (27 comments in
 
differences for the task stra
no task strategy and 20 in ta
sk strategy) and the 
asm and 20 
enthusiasm variable (27 comm
ents in low enthusi
h enthusiasm). some represen
tative comments are the 
in hig
following: 
I did the experiment for the 
points and that's all. 
n, task strategy, low enthusia
sm) 
(No visio
It was boring. (No vision, ta
sk strategy, low 
enthusiasm) 
178 
I really found the experiment boring - I don't know 
or understand why they are doing this. (No vision, no 
task strategy, low enthusiasm) 
From the f irst binder we did, I had the impression 
that the task was boring and tedious. (No vision, 
task strategy, low enthusiasm) 
No, the task was boring, and he d i d nothing to help. 
(No vision, no task strategy, high enthus i asm) 
It was boring and quite stupid! (No vision, no task 
strategy, high enthusiasm) 
A very boring and tedious e xperiment . (No vision, no 
task strategy, high enthusiasm) 
Overall, these comments indicate that without the 
vision, the experiment made no sense. It was simply "busy 
work" to the subjects. They did not understand the 
purpose of the experiment, why the task was important, or 
how their work would help the company. The subjects' 
perceptions of the task are especially interesting 
considering that, regardless of condition, each subject 
performed the same task under the same conditions. The 
leader's vision (or lack of it) and to a lesser extent, 
179 
the leader's enthusiasm level affected how the subjects 
perceived the task. The characteristics of the job itself 
were held constant. 
Experimenter's Observations 
Serving as the experimenter, my own observations 
generally confirm the subjects' comments about the 
experiment. There were differences in the mood or 
atmosphere of the experiment. In particular, during no 
vision/no task strategy/low enthusiasm sessions there was 
tension in the room. While the subjects were working, the 
room was very quiet. Subjects had facial expressions of 
boredom and frustration. While waiting for other subjects 
to finish scoring or to complete a questionnaire, subjects 
tended to close their eyes, put their heads down on their 
desks, or work on class assignments. Overall, the 
atmosphere was negative and tense. 
In contrast, during the vision/task strategy 
sessions, regardless of enthusiasm level, the atmosphere 
was one of working quickly to get the job done. The room 
was filled with sounds of papers shuffling and binders 
opening and closing. While waiting, subjects organized 
their work areas or checked over their work a second time 
to try to figure out why they made a mistake. Overall, 
the atmosphere was a positive one . 
180 
Although the subjects listened to the low enthusiasm 
leader, they tended to have blank stares on their faces 
while listening or to look away from the leader while 
listening. During the high enthusiasm sessions, subjects 
tended to watch the leader intently. The high enthusiasm 
leader seemed to "grab" the subjects' attention, 
regardless of what he/she was saying. 
Actors' Comments 
Throughout the experiment, both actors were asked to 
keep a journal of their impressions of the experiment. 
They each provided a written summary of their experiences 
during the experiment. Although they both had the talent 
and experience to portray the desired behaviors (and that 
they did so was confirmed via the manipulation checks), 
they found the experiment an emotionally difficult 
experience for three reasons. 
First, it was difficult for them to display the same 
behaviors and to communicate the same scripts over a 
period of eight weeks. Although they had the scripts 
memorized, they found it a "rote exercise" to say the same 
thing many times. The male actor felt that at first, the 
vision material was inspirational and interesting, but 
after expressing the vision eight times he found the 
material "empty" and non-motivational. 
181 
Second, the requirements of the experiment went 
against their natural tendencies. The male actor had 
considerable experience doing "one- man" shows where he did 
a lot of improvising. He felt that the requirements of 
the low enthusiasm condition and the lack of interaction 
with the subjects went against his natural tendencies, 
causing him to feel uncomfortable. He also had a personal 
style of using humor, which was suppressed during the 
experiment. The female actor was mostly experienced with 
stage plays and enjoyed the experiences of developing a 
rapport or "getting in sync" with the other actors during 
performances. Working alone was not enjoyable to her. 
She felt she was boring during the no vision sessions 
whereas she felt she was being helpful during the vision 
and task strategy conditions. Although she was polite, 
she also felt that she was being extremely rude to the 
subjects during the low enthusiasm conditions because she 
was not showing interest in the experiment. 
Third, the decomposition of the process and content 
variables was another source of difficulty . It makes 
sense to most people, and the above subject comments 
confirm this, that leaders are interesting and expressive 
people who enjoy talking about their company. Thus, the 
vision/low enthusiasm condition and the no vision 
conditions went against the common perceptions how a 
182 
leader acts. Performing sessions that did not fit with 
common perceptions of how a leader should act caused the 
actors to feel less confident of their performances; they 
felt that these performances were "phony" and that "the 
artifice was showing through." They both found it much 
easier, from a personal and emotional viewpoint, to be 
highly enthusiastic, to give an interesting vision, and to 
be helpful to the subjects by ?telling them the task 
strategies. It was a difficult experience for the actors 
to be purposefully uninterested during the low enthusiasm 
sessions, to purposefully present the irrelevant and dull 
no vision material, and to knowingly withhold the helpful 
task strategy information. 
Still, both leaders intently expressed that the 
experiment was challenging to them. They each learned a 
lot about their personal styles as well as about 
communicating ideas, which they felt would improve their 
acting skills. They enjoyed the experience of working 
with the undergraduate subjects and found the idea of 
helping conduct an experiment to be interesting. The 
implications of these experiences for training leaders are 
discusses in chapter 10. 
183 
Summary 
The qualitative results are valuable three in ways. 
First, they support the quantitative results that the 
vision variable had the most impact on performance, 
attitudes, and perceptions. Second, they suggest some new 
research directions concerning the effect of leadership on 
job attitudes. Third, they have implications for training 
leaders to be effective. 
184 
Chapter 10 
Discussion 
To answer the questions posed by 
this dissertation, 
affects followers' 
the content of a leader's vision 
tudes and perceptions 
performance, task attitudes, and 
atti
 
toward leader both directly and i
ndirectly. The process
or style of a leader's communicat
ion directly affects 
followers' attitudes and perceptio
ns toward the leader, 
fect followers' performance or tas
k 
but does not af
attitudes. 
er, I will discuss the limitations
 o f 
In this chapt
the research, the implications fo
r theory, methodology, 
. 
and practice, and future research
 directions
Limitations 
There at least are four limitation
s to this 
on. , the time span was only 2 hours, 
as 
dissertati First
compared to the years or decades 
which characterize the 
lthough the effect 
tenure of organizational leaders.
 A
 small, it is 
sizes (w
2
) for the independent variables w
ere
encouraging that significant effe
cts were found despite 
ort time span. Over a longer time
 period, the 
the sh
 would probably be stronger due to
 its being 
Vision effects
 (e.g. 
~epeated many times and in many d
ifferent ways
185 
through role modeling) and being reinforced by the reward 
system. The effects of process may play a larger role in 
the long run, such as by encouraging people to remain with 
the organization rather than leave. This could affect 
performance indirectly through the retaining of skilled 
employees. Further, outcomes such as mood could be 
enhanced by repeated doses of leader enthusiasm. 
Second, the leader in this study led only small, 
independent groups of subjects. Different effects may 
occur when leading a large group in that the subordinates 
could motivate each other as a result of the leader's 
actions. Specifically, a leader's impact on a group of 
followers may result in social contagion effects. The 
followers may advance the leader's and other followers' 
ideas for implementing the vision, resulting in a 
synergistic .effect. These group-level effects may result 
in stronger effects, as compared to individual-level 
effects, of the leader on follower performance, attitudes, 
and perceptions. On the other side of the same coin, real 
leaders can motivate one-on-one (e.g. Bass' individualized 
consideration) which could produce stronger effects than 
group-level actions. 
Third, the simulated task performed by the subjects 
was also a function of their ability. Ability accounted 
for a larger proportion of variance in performance than 
186 
did the leader's actions. However, the leader's actions 
were significant when individual ability was controlled, 
indicating that the leader had an effect over and above 
ability. The experimental setting and relatively short 
time period (40 minutes of actual work on the task) may 
have encouraged people to work close to their capacity. 
It is likely when the leader has the opportunity to 
communicate and implement a vision over a period of years, 
the leader will have a stronger effect on followers' 
performance. 
Fourth, results found here will need to be replicated 
with employees in actual jobs. Most jobs include 
attributes not examined here, such as technological 
demands and constraints, and interactions with co-workers 
and supervisors. Actual jobs may also differ on 
traditional job attributes, such as skill variety, task 
identity, task significance, autonomy, and feedback 
(Hackman & Oldham, 1980). Instead of assuming that a 
field study will yield identical results to this 
dissertation, this assumption should be tested (Locke, 
1986). 
187 
Implications 
Theoretical Implication
s 
arismatic 
This study clarified th
e nature of ch
 the components found in
 the 
adership, by decomposin
g
le
e components specified 
by 
Howell and Frost study 
and th
iscussed 
eory. Specific theoret
ical implications are d
th
sion, vision implementa
tion, and enthusiastic 
for vi
communication style. 
as well as the 
Vision. Vision affects 
performance 
and perceptions toward t
he leader. 
followers' attitudes 
cts performance quality 
and 
Vision also indirectly 
affe
se are two 
uantity through goals an
d self-efficacy. The
q
which leader actions ~f
fect follower 
mechanisms by 
r that the vision affect
ed 
performance. It is clea
sed in this 
quality-related variabl
es because the vision u
dissertation stressed q
uality. 
s of vision could simil
arly affect other 
Other type
hanisms. For example, a
 
outcomes through the sam
e mec
cost 
er could communicate a 
vision of being a low-
lead
 In this case, employe
es may then set cost-
Producer.
igh self-efficacy for ac
hieving 
reduction goals and hav
e h
 in actual cost 
them, which would, in tu
rn, result
reduction. 
be noted, on the negativ
e side, that 
It should 
A vision could 
Visions do not have to 
be coherent. 
188 
nceivably promote complex and confli
cting values. An 
co
Value 
example is Sears' vision statement f
rom chapter 3, "
 This vision specifies both value, 
or 
at a decent price."
quality, and decent price, or low-co
st, which can be in 
deoff 
conflict. In this case, employees' 
may choose to tra
one part of the vision for the other
. 
b-
The vision manipulation was made up 
of several su
cluding the vision itself, the visio
n 
components, in
statement, arousing motives via com
petition (comparison of 
), 
performance with competing printing 
companies' employees
self-efficacy. It is possible that 
each sub-
and building 
component has a different effect. H
owever, fuller 
o 
decomposition could result in sub-co
mponents that are to
ak to cause any significant effects. 
Future 
small or we
research could test this possibility
. 
Vision Implementation: Task Strategy
. One way a 
leader can implement the vision at a
 micro-level is via 
roach the 
information dissemination concerning
 how to app
dissertation found that the task str
ategy 
task. This 
information facilitated performance.
 Because the task 
cts could 
strategy provided information on how
 the subje
 task better, it affected mainly per
formance 
Perform the
quantity and, to a lesser extent, pe
rformance quality. 
 
Quantity goals mediated the task stra
tegy-performanc~
e another 
quantity relationship. Goals were f
ound to b
189 
ism through which the leader, via
 the task strategy 
mechan
information, influences followers
. 
As with the vision, it is clear t
hat other types of 
e 
task strategies would yield effec
ts congruent with th
e of strategy. For example, a str
ategy 
content of each typ
to achieve cost-reduction may inc
lude giving employees 
as asking a 
suggestions in order to reduce co
sts, such 
nt on large orders. Again, future
 
supplier for a discou
research could examine the effect
 of other types of task 
r direct and 
strategi es, but it is quite likel
y that simila
indirect effects will occur. / 
ders need 
It is also important to point out
 that lea
to take many types of actions to 
implement the vision. 
 only one type of implementation, This experiment examined
disseminating information in the 
form of a task strategy. 
Clearly, leaders must manage a we
alth of information 
). Leaders are constantly gatheri
ng, 
(Locke et al., 1991
integrating, and disseminating in
formation (Fleishman et 
al., 1992; Locke et al., 1991). 
This dissertation 
 way 
supported information disseminati
on as one effective
to implement the vision. 
Besides disseminating information
, leaders use five 
he vision (Locke et al., 
additional methods to implement t
19 , including (a) agenda developmen
t; (b) structuring; 
91)
d) 
(c) selecting, acculturating, and
 training; (
190 
motivating; (e) team building; and (f) pro
moting change, 
innovation, and risk-taking . Although fut
ure research is 
ded to test for the effects of these actio
ns, it is 
nee
quite likely that these actions will have 
direct and 
indirect effects on employees' performance
 similar to the 
s found here. A leader who uses many suit
able and 
one
coordinated implementation methods is more
 likely to 
successfully implement the vision than one
 who uses few 
lementation methods. Clearly, the vision w
ill probably 
imp
not be successfully implemented if the lea
der only 
disseminates information (or exclusively p
ursues only one 
ntation method), no matter how relevant or impleme
important the information. 
Enthusiastic Communication style. The pro
cess used 
 not affect 
by a charismatic leader to communicate did
Performance. However, it did affect attitu
des and 
Perceptions toward the leader, although mo
re weakly than 
vision and task strategy. 
The presence of an enthusiasm by vision in
teractions 
usiasm 
on goals and self-efficacy indicates that 
high enth
have detrimental effects in combination wi
th a vision, 
may 
but facilitative effects in combination wi
th no vision. 
istracted 
When hearing the vision, subjects may have
 been d
by the leader's high enthusiasm and not ha
ve paid close 
attention to the vision. But, high enthus
iasm may have 
191 
tated subjects' attention to the 
no vision script. 
facili
f-efficacy 
These interactions only affected 
goals and sel
affect performance. Therefore, th
ey may not 
and did not 
have important practical implicat
ions for increasing 
performance. 
ults of the manipulation checks s
uggest that 
The res
an enthusiastic style may be attri
buted to leaders who 
communicate a vision or task stra
tegy regardless of the 
 possible explanation for 
actual enthusiasm behaviors. One
this is that the actors actually 
displayed higher 
enthusiasm during the vision and 
task strategy 
resent. However, 
manipulations than when these wer
e not p
they were trained to exhibit the 
same enthusiasm behavior 
ndition . A second explanation is regardless of content co
that people see high enthusiasm a
nd vision communication 
 
as concepts that go together and 
would therefore be likely
to report that both occurred. I
ndividuals probably 
n in the same or highly 
organize high enthusiasm and visio
e mere act of 
related knowledge structures or s
chemas. Th
on or task strategy "shows" that 
the 
communicating a visi
leader is enthusiastic toward and
 interested in the 
h enthusiasm 
organization. Subjects may have 
perceived hig
aring the leader communicate the 
vision even though 
When he
asm 
the leader displayed the same lev
el of enthusi
192 
 of vision condition. These 
explanations will 
regardless
need to be examined in future
 research. 
Mechanisms of Charismatic Lea
dership. The 
 
relationship between content,
 but not process, and
 by goals and self-efficacy. 
The 
performance was mediated
vision and the task strategy 
caused followers to 
nd 
spontaneously set challenging
 goals for performance a
icacy. This supports House's
 
increased their self-eff
(1977) proposition that charis
matic leaders affect 
followers' goals and self-con
fidence. The present 
indings represent a first ste
p toward understanding how 
f
llower performance. In 
charismatic leaders affect fo
ings supported previous goal 
setting 
addition, these find
nd self-efficacy research tha
t goals serve as mediators 
a
lso supported Locke's (1991) 
(Locke & Latham, 1990). It a
t goals 
concept of the motivational '
'hub," which argues tha
re closer to actions than oth
er causal 
and self-efficacy a
variables in the realm of mo
tivation. 
 of the main, or direct, effe
cts of 
Although strength
he independent variables on t
he dependent variables were 
t
ngth of the effects (as 
sma11 (Cohen, 1977), the stre
e R2 values in the regression 
equations 
indicated by th
testing for mediation) were c
onsiderably larger for the 
les. This 
indirect effects through the 
mediating variab
n order for a 
indicates that the mediators 
are essential i
193 
to affect followers' performance, leader's charisma 
attitudes, and perceptions. 
 literature should 
Other mechanisms suggested by the
e examined in future research , su
ch as commitment to the 
b
leader (House, 1977). 
heoretical Implications. Th i s dis
sertation 
Other T
provided support for the ideas th
at the perception of 
harismatic 
charisma is due to actions exhibi
ted by c
 consistent with recent approache
s that 
leaders. This is
focus on the specific acquired sk
ills and actions (e.g . 
er 
developing and communicating a vi
sion) than rath
st on a charismatic style or perso
nality 
focusing ju
(Collins & Porras, 1991; Conger , 
1989; Uleman, 1991; 
usiasm were 
Willner, 1984). Both vision and 
enth
a leader being regarded as charism
atic. 
associated with 
In addition, this dissertation an
swered the question 
posed by Conger and Kanungo (1988
) about whether 
ip is either a dichotomous or sca
lar 
charismatic leadersh
Phenomenon. If charisma is a dic
hotomous phenomenon, all 
ld be 
characteristics of charismatic le
adership wou
 be perceived as charismatic. An required in order to
interaction of vision and enthusia
sm on the charisma scale 
stead, only 
Would be expected if this were th
e case. In
r additive, effects for vision and
 enthusiasm on 
main, o
icates 
the perception of charisma were f
ound. This ind
194 
e presence of vision or enthusiasm
 separately 
that th
te to the perception of the leader
 as charismatic . 
contribu
Thus, charisma is more properly c
onceptualized as a scalar 
h vision has 
phenomenon (e.g., a continuum) . 
Althoug
an enthusiasm, each contributed to
 the 
stronger effects th
perception of charisma. When nei
ther vision nor high 
 charisma (measured with 
enthusiasm are present, perceived
cale) is low. Charisma is rated a
s 
Bass' Form BY s
slightly higher when only high en
thusiasm is present, 
h 
followed by vision only. It is ra
ted highest when bot
vision and high enthusiasm are pr
esent. 
ilar to those 
The effects on task attitudes are 
sim
found in previous research on the 
motivation sequence 
(Locke, 1991). Most previous rese
arch has examined the 
task satisfaction, 
relationship between performance 
and 
concluding that performance succe
ss is the main 
determinant of satisfaction. Thi
s dissertation found that 
sfaction, task 
high performance led to a high tas
k sati
nd enthusiasm for the task/vision.
 
clarity, a
t i s also interesting to not e the 
relationshi ps that 
I
performance 
were not present in the fin a l mod
el . That i s , 
titudes a nd percep t ions toward the
 
did not affect at
leade r. Some t h e or i sts have s pecu
lated tha t the p resence 
. f rence of attitudes a nd 
of lea ders hip a nd t h e i n e 
r is attributed t o the leader d 
Pe r ceptions toward t h e lea e 
195 
ing to the follower's own or group succ
ess (Calder, 
accord
1977; Pfeffer, 1977). That is, if an 
individual has high 
performance, the individual will then a
ttribute this high 
gardless of 
performance as being due to the leader,
 re
he leader had any effect. In fact, Pf
effer 
whether t
(1977) speculates that leaders have no 
actual effects on 
true, then no 
followers' performance. If this theory
 were 
of leader actions on performance, attitu
des, or 
effect 
perceptions would have been found but f
ollower performance 
 
would have been significantly related t
o attitudes and
ions toward the leader. Instead, the l
eader's 
percept
actions had predictable and meaningful 
effects on 
performance, attitudes, and perception
s; follower 
performance did not significantly affec
t attitudes and 
support was found 
perceptions toward the leader. Thus, n
o 
for this type of attributional leadersh
ip theory. 
This dissertation also has implications
 for more 
et al. 
general leadership models. Specificall
y, the Locke 
ss model of leadership states that effe
ctive 
(1991) proce
leaders possess certain traits (one of 
which is charisma) 
ibute to 
and knowledge, skills, and abilities w
hich contr
ility to develop a vision. The vision 
must then 
their ab
be implemented in order for the leader 
to be effective. 
Although this dissertation did not spe
cifically test all 
 three 
these processual linkages, support was 
found for
196 
tionships specified by Locke et a
l.'s model: (a) 
rela
trategy 
effective leaders communicate a v
ision, (b) task s
e way that effective leaders imple
ment vision, and 
is on
(c) the vision-leader effectivenes
s relationship is 
efficacy. 
mediated by followers' goals and s
elf-
, the qualitative results suggest 
an 
Finally
interesting new research direction
 - the effects of 
cts in the 
charismatic leadership on job atti
tudes. Subje
r task strategy conditions found t
he task 
vision o
nteresting and engaging, but those
 in the no vision or no 
i
 task was boring 
task strategy conditions reported 
tha~ the
senseless. The vision attempts te
ll each employee how 
and 
its into the purpose and values of
 the 
their specific job f
organization. The leader may have
 influ~nced the 
 (Hackman 
subjects' experienced meaningfulne
ss of the task
& Oldham, 1980). 
n 
In sum, this dissertation provides
 evidence that eve
eadership is a complex phenomenon 
with 
though charismatic l
and 
many different effects, it can be 
systematically 
objectively studied. 
Methodological Implications 
ssertation has three implications 
for the 
This di
methodology used to study charism
atic leadership. First, 
udy 
it increases our confidence in our
 ability to st
pulating 
leadership in a controlled setting
. Mani
197 
dership components in an experime
ntal setting 
lea
olled, 
permitted extraneous sources of v
ariance to be contr
f causation in the leadership pro
cess to be 
the direction o
determined, and the leadership el
ements to be separated. 
letely rule out other 
More research will be needed to c
omp
on (e.g. the effects of attitudes
 
possible types of causati
toward the leader on task attitud
es). 
Second, this dissertation further
 developed Howell 
and Frost's (1989) "actor" method
ology. The actors were 
monstrate 
able to successfully learn and co
nsistently de
. Displaying charismatic leaders
hip 
charismatic leadership
was a difficult experience for th
e actors, but this was 
d not to the 
mainly due to the demands of the 
experiment an
demands of learning to be charism
atic. The actor 
ay to study leadership 
methodology seems to be a viable 
w
processes. Future users of this 
methodology should be 
certain to take appropriate steps
 in selecting and 
their behaviors. 
training actors as well as verify
ing 
Finally, this study shows that ba
sic research on 
leadership is possible and sugges
ts the ultimate 
 span, an 
feasibility of building, over a l
ong time
of leadership. This contrasts wi
th being 
inductive theory 
nce 
overly concerned with building a 
grand "theory" in adva
7). By 
of sufficient knowledge (Campbell
, 197
rismatic 
systematically studying the effec
ts of cha
198 
leaders, progress can be made toward understanding this 
phenomenon. 
Practical Implications 
The fact that charismatic leadership can be trained 
has implications for the current trend toward work teams 
and decentralization. Organizations are relying more and 
more on personal influence rather than an explicit 
hierarchy of authority to accomplish the work. 
Organizations may want to take advantage of the present 
findings by applying them to their leader training and 
selection methods. 
First, if performance is the main criterion, the 
vision is the most important factor for which to train 
and/or select. This is consistent with Bennis and Nanus' 
(1985) observation that a charismatic style or personality 
trait (i.e. process) was not related to leadership 
effectiveness. Rather, as suggested by Bennis and Nanus 
and by the results found in this dissertation, an 
effective leader should provide direction to followers, 
through the vision and vision implementation. 
Second, the fact that actors can be trained to 
exhibit charismatic leadership behaviors may mean that it 
is possible for almost anyone to learn how to be 
charismatic. This dissertation provided evidence that the 
effects on performance is the result of acquired skills 
199 
rned or developed, rather th
an a "style" or 
that can be lea
ality trait with which one i
s born. This is 
person
t 
ngruent with Conger and Kan
ungo's (1988) viewpoint tha
co
 be trained. 
charismatic leadership qual
ities can
e results of this dissertati
on advise practitioners 
Th
with improving performance d 
and trainers who are concer
ne
's communications. 
to emphasize the content of
 a leader
mployees must be skilled at 
the visioning process. 
E
nt some of their learning 
Collins and Porras (1991) re
cou
 their 
experiences from advising le
aders who are developing
sion. They proyide a framew
ork 
vision and strategic vi
n useful 
that has guided their traini
ng methods and has bee
nd vision. 
for helping trainees unders
ta
In addition, it is desirable
 to train and develop 
plement the vision. The 
employees in skills needed 
to im
n 
Locke et al. ? ( 1991) model s
pecifies six implementatio
uld be trained to use. Alth
ough 
methods which employees co
some of these methods are a
lready familiar to 
ations, such as selecting an
d training employees, 
organiz
that leaders are trained in e 
organizations need to ensur
ese areas rather than leavin
g these functions up to a 
th
rce 
specialized department (for
 example, the human resou
department) or committee. 
ird, if the organization wis
hes to affect attitudes 
Th
ployees should 
and perceptions toward the 
leader, then em
200 
receive communication skill training. Although such 
training will not affect an individual's personality, it 
will improve their communication process. Speech and 
articulation skills were not examined in this 
dissertation, but such training should include these 
skills as well as listening skills and interpersonal 
sensitivity skills (Conger & Kanungo, 1988). Because 
leaders usually interact with followers on a daily basis, 
several types of process skills may be required. And, as 
stated above, this may have other positive results, such 
as reduced absenteeism or reduced leader-follower 
conflict. Such training may not only affect followers' 
attitudes and perceptions toward the leader but may also 
produce increased self-efficacy on the part of the leader. 
Fourth, when selecting leaders, this dissertation 
provides some advice about the types of characteristics 
organizations should look for in an effective leader. 
When selecting for leadership positions, organizations 
should hire individuals who have the ability to develop, 
communicate, and implement a vision. To do this, 
knowledge of the industry and organization is needed. The 
organization may wish to select employees for leadership 
positions based on their relevant knowledge and expertise, 
and then train the individual to apply this knowledge and 
expertise to the organization by developing a meaningful 
201 
vision. It is crucial that the leader learn how to 
translate this expertise so that it focuses attention and 
provides a unique image of the future. The importance of 
this can be seen in the no vision manipulation. In this 
condition, the leader was knowledgeable and was perceived 
as an expert. However, the leader failed to translate 
this expertise into an image of the future; the leader did 
not provide any guidance that was relevant to the 
subjects' work on the task. 
Although both process and content components are 
important, it may be easier to hire an individual with an 
enthusiastic communication style and well-developed 
communication skills and then assist that person in 
developing a vision. Although the communication style is 
not as crucial as vision, it is nevertheless an important 
factor to consider when hiring for a leader position. The 
organization may want to consider the type of position 
when determining the communication skills that the leader 
will need. For example, a vice-president may need public 
speaking skills more than a group manager, who may need 
interpersonal and listening skills. 
Fifth, the leader must affect goals and self-efficacy 
in order to affect performance. Although a leader could 
let followers spontaneously set goals, as was done in this 
dissertation, a leader may want to take more direct action 
202 
to ensure that goals are set and that followers have high 
self-efficacy. To affect goals, a leader could either (a) 
require that personal goals are set without specifying the 
content of the goal or (b) assign challenging goals to 
employees. To affect self-efficacy, the leader could 
engage in verbal persuasion, train employees in relevant 
skills, ensure that employees' have initial job successes, 
and serve as a role model for employees (Bandura, 1986). 
This would ensure that the leader would affect employees' 
performance. In fact, setting goals and building self-
confidence is another way that leaders implement the 
vision (Locke et al., 1991). Organizations frequently set 
goals at several levels of the organization, but the 
leader should ensure that these goals are set and that 
they are congruent with the vision. 
Future Research Directions 
The above discussions point out several areas where 
future research is needed. Further examination of the 
relationships between charismatic leader actions, goals, 
performance, attitudes, and perceptions would be useful. 
It would also be useful to see if the present findings 
generalize to different types of visions as well as 
different (and different combinations of) implementation 
methods. 
203 
Additional variables warrant inclusion in future 
studies. The enthusiasm variable may play a more 
important role when absenteeism, turnover, leader-follower 
conflict, and citizenship behaviors are included. That 
is, positive, as compared to negative, attitudes and 
perceptions toward a leader may result in lower 
absenteeism, lower turnover, less conflict between the 
leader and followers, and an increase in citizenship 
behaviors or discretionary behavior that is not explicitly 
rewarded by the organization (Organ, 1988). In fact, 
Podsakoff, MacKenzie, Moorman, and Fetter (1990) found 
that transformational leadership behaviors indirectly 
affected citizenship behaviors displayed by followers 
through followers' trust in the leader. 
Also, the effects of content and process over 
different, and especially longer, time periods, need to be 
examined. Different effects of content and process may 
emerge over the course of several years. The vision may 
have even stronger effects when communicated and 
implemented in many different ways. And, having a highly 
enthusiastic style may be needed to maintain followers' 
commitment to the vision over long periods of time. 
Some specific future research projects are discussed 
below. First, research on the visioning process is 
needed. Field research is needed to address such issues 
204 
s initially come up with an 
idea for the 
as how leader
a meaningful and 
vision, how they develop thi
s vision into 
ea that can be communicated 
to followers, how 
coherent id
to" the vision and deal with
 
they get others to ''buy in
 
those who do not buy into th
e vision, and how leaders
t the vision. Also, the qua
litative 
choose to implemen
terature suggests that visio
ns can outlast the leader 
li
n die with a leader. Also, ca
who developed the vision or 
s can evolve slowly over tim
e or can remain the same 
vision
1). 
over time (Collins & Porras,
 1991; Locke et al., 199
the reasons why these 
Research is needed to examin
e 
effects may occur. 
be 
Second, other types of causa
l models need to 
odel includes the nonrecursiv
e 
examined. One interesting m
effects of followers' perfor
mance, attitudes, and 
ptions on subsequent leader 
actions. For example, 
perce
 follower performance could 
increase both the leader's 
high
o communicate and implement y t
confidence in his/her abilit
ce in the 
the vision as well as the fo
llowers' confiden
 
eader and commitment to the 
vision. On the other hand,
l
 result in the leader and 
low follower performance may
on methods 
followers questioning the vi
sion or implementati
ollower commitment to the vi
sion. A 
as well as lower f
e 
time-series model could be u
sed to examine thes
relationships. 
205 
Third, future studies should attempt to train 
individuals to exhibit charismatic leadership behaviors. 
Although training programs can be used to teach almost any 
skill, it is possible that only those who are naturally 
expressive, such as actors, can benefit from training. rt 
may be that while anyone can benefit from vision training, 
only some may benefit from communication style training. 
Pre-training and post-training observer ratings could 
measure an individual's level of perceived charisma when 
giving a speech and when interacting with followers and/or 
peers. Changes in follower performance could be measured. 
Fourth, a related issue of interest is investigating 
the effects of vision and vision implementation at lower-
levels of the organization. Leadership theories and 
research usually assume a top-down approach, however, this 
dissertation provides evidence that it may be possible to 
train leaders at all levels to exhibit charismatic 
leadership behaviors. Because today's organizations are 
becoming flatter and more decentralized, employees will 
have to rely more on informal and personal types of power 
rather than the formal authority granted by the 
organization. Vision and vision implementation could have 
important implications for work teams and, in general, 
lower-level employees. 
206 
Fifth, it would be interesting to examine the extent 
to which charismatic leadership is related to other types 
of leadership, including traditional approaches such as 
the behaviors of structuring and consideration (Yukl, 
1989) as well as more recent approaches, such as self-
leadership (Manz & Sims, 1989; Manz & Sims, 1991). Being 
considered a charismatic leader does not preclude the 
leader demonstrating structuring (task-oriented) behaviors 
or considerate (people-oriented) behaviors. In fact, 
these behaviors may be ways that a charismatic leader 
implements the vision. The self-leadership approach 
proposes that a "SuperLeader" shares power with followers, 
enabling them to become leaders themselves. Although some 
researchers view visionary leadership and SuperLeadership 
as distinct, mutually exclusive concepts (e.g. Manz & 
Sims, 1991), others view these concepts as compatible 
(Avolio & Bass, 1988). Although clarification of these 
concepts is needed at a theoretical level, the 
relationships between these various types of leadership 
should be examined empirically. 
Sixth, the effects of charismatic leader actions on 
job attitudes could be examined, as suggested by the 
qualitative results. A future experimental study could 
manipulate vision (versus no vision) and measure the 
corresponding effects on followers' experienced task 
207 
meaningfulness. A correlational methodology could also be
 
employed to examine this effect. 
Conclusion 
overall, this dissertation suggests that although 
charisma is a complex phenomenon, is it not impossible to 
study it in a systematic manner. This dissertation 
advanced our knowledge of charismatic leadership. It is 
hoped that these findings will spark further research on 
this phenomenon. 
208 
References 
Avolio, B.J. & Bass, B.M. (1988). Transformational 
. 
leadership, charisma, and beyond. In J.G. Hunt, B.
R
Baliga, H.P. Daehler, & C.A. Schrieshiem (Eds.), 
Emerging Leadership Vistas . Lexington, MA: Lexingto
n 
Books. 
Avolio, B.J. & Gibbons, T.C. (1988). Developing 
transformational leaders: A life-span approach. In 
J.A. Conger, R.N. Kanungo, & Associates, Charismatic
 
Leadership. San Francisco: Jessey-Bass. 
volio, B.J., & Waldman, D.A. & Einstein, W.O. (1988). A
Transformational leadership in a management 
simulation game. Group & Organization Studies, 1.1., 
59-80. 
Bandura, A. (1986). Social foundations of thought a
nd 
action: A social cognitive theory. Englewood Cliffs
, 
NJ: Prentice-Hall. 
Baron, R.M., & Kenny, D.A. (1986). The moderator-med
iator 
variable distinction in social psychological 
research: Conceptual, strategic, and statistical 
considerations. Journal of Personality and Social 
Psychology, 51, 1173-1182. 
Bass, B.M. & Avolio, B.J. (1989). Manual for the 
Multifactor Leadership Questionnaire Profile. Palo 
Alto, CA: Consulting Psychologists Press . 
209 
Bass, B.M. (19 8 1) . Stogdill 's handbook of leadership (2nd 
edition) . New York, NY: The Free Press. 
Bass, B.M. (1985). Leadersh ip a nd performance beyond 
expectations. New York , NY : The Free Press. 
Bass, B.M. (1990). Handbook of l eadership (3rd edition). 
New York, NY: The Free Press. 
Beck, R.N. (198 7 ). Visions , values, and strategies: 
Changing a ttitudes and c u lture. Academy of Management 
Executive, i, 33-41. 
Bennis, W. & Na nus, B. (198 5). Leaders: The strategies for 
taking charge. Ne w York : Harper & Row. 
Berlew, D.E. (1974) . Leadership and organizational 
excitement . California Ma nagement Review, 17(2), 21-
30. 
Brad f ord, D.L. & Cohen, A.R. (1 984) . Managing for 
excellence: The guide to developing high performance 
in contemporary organizations. New York: John Wiley & 
Sons. 
Bra dley, R.T. ( 198 7 ) . Charisma and social structure: A 
s tudy o f low and power, wholeness and transformation. 
New York: Paragon. 
Burns , J.M. (19 78). Le adership. New York, NY: Harper and 
Row. 
210 
Calder, B.J. (1977). An attribution theory of leadership. 
In B.M. Staw & G.R. Salancik (Eds.), New direction in 
organizational behavior. Chicago: St. Clair. 
Campbell, D.J. (1991) . Goals levels, complex tasks, and 
strategy development: A review and analysis. Human 
Performance, till, 1-31. 
Campbell, J.P. (1977). The cutting edge of leadership: An 
overview. In J.G. Hunt & L.L. Larson (eds.), 
Leadership: The cutting edge. USA: Southern Illinois 
University Press. 
Cochran, W.G. & Cox, G.M. (1957). Experimental Designs, 
2nd edition. New York: Wiley. 
Cohen, J. (1977). Statistical power analysis for the 
behavioral sciences, Rev. ed. New York: Academic 
Press. 
Collins, J.C. & Porras, J.I. (1991). Organizational vision 
and visionary organizations. California Management 
Review, 34(1), 30-52. 
Conger, J.A. (1989). The charismatic leader. San 
Francisco: Jossey-Bass. 
Conger, J.A . (1991). Inspiring others: The language of 
leadership. The Executive, 2, 31-45. 
Conger, J.A. & Kanungo, R.N. (1987). Toward a behavioral 
theory of charismatic leadership in organizational 
211 
settings . Academy of Management Review, 12(4). 637-
647. 
Conger, J.A., Kanungo , R.N. , and Associates. (1988). 
Charismatic lead e rship . San Francisco: Jossey-Bass. 
Dow, T.E. Jr. (1969) . The t h eor~ of charisma. Sociological 
Quarterly, 10, 306-318 . 
Earley, P.C. ( 1 98 5 ). Influence of information, choice, and 
task comp l exity upon goa l acceptance, performance, 
and personal goals. Jour n a l of Applied Psychology, 
70, 481-491 . 
Eden, D. (1990). Pygmalion in management. Lexington, MA: 
Lexington Books. 
Edinger, J.A. & Patters on, M.L . (1983). Nonverbal 
involvement and social c ontrol. Psychological 
Bulletin, 2._l , 3 0 - 56 . 
Fleishman, E.A . , Mumford, M.D., Zaccaro, S.J., Levin, 
K.Y., Korotkin, A. L ., & He in, M. B. (1991). Taxonomic 
e f forts in the description of leadership behavior: A 
synthesis and f unctional i nterpretation. Leadership 
Quarterly, ~ , 2 4 5 -288. 
Friedman, H.s., Prince, L.M., Riggio, R.E. & DiMatteo, 
M.R. 
(1980). Und e rstanding and assessing nonverbal 
expressiveness: The affective communication test. 
Journal of Personality and social Psychology. 39(2), 
333 - 3 51. 
212 
Harrison, R. (1987). Harnessing personal energy: How 
companies can inspire employees. Organizational 
Dynamics, Autumn, 4-21. 
Hackman, J.R., & Oldham, G.R. (1980). Work redesign. 
Reading, MA: Addison-Wesley. 
Hater, J.J. & Bass, B.M. (1988). Superiors' evaluations 
and subordinates' perceptions of transformational and 
transactional leadership. Journal of Applied 
Psychology, 73(4), 695-702. 
Hollenbeck, J.R., Klein, H.J., O'Leary, A.M., & Wright, 
P.M. (1988). An investigation of the construct 
validity of a self-report measure of goal commitment. 
Presented at the annual meeting of the Academy of 
Management. 
House, R.J. (1977). A 1976 theory of charismatic 
leadership. In J.G. Hunt and L.L. Larson (Eds.). 
Leadership: The cutting edge. Carbondale, IL: 
Southern Illinois University Press. 
House, R.J., Spangler, W.D., & Woycke, J . (1991). 
Personality and charisma in the U.S. Presidency: A 
Psychological Theory of Leader Effectiveness. 
University of Pennsylvania: Unpublished Manuscript. 
House, R.J., Woycke, J., & Fodor, E.M. (1988). Charismatic 
and Noncharismatic Leaders: Differences in Behavior 
and Effectiveness. In J.A. Conger & R.N . Kanungo 
213 
(Eds.), Charismatic leadership. San Fra
ncisco: 
Jessey-Bass. 
: An interview 
Howard, R. (1990). Values make the com
pany
 
with Robert Haas. Harvard Business Rev
iew, Sept-Oct,
133-144. 
Howell, J.M. & Frost, P.J. {1989). A lab
oratory study of 
charismatic leadership. Organizational 
Behavior and 
Human Decision Processes, ? .11_, 243-269. 
mpions of 
Howell, J.M. & Higgins, C.A. (1990). Cha
technological innovation. Administrativ
e Science 
Quarterly, 1...2, 317-341. 
cy and 
Jauch, L.R. Glueck, W.F. (1988). Bus
iness poli& 
strategic management. New York: McGraw-
Hill. 
Hulin, C.L. (in press). Job .satisfactio
n as 
Judge, T.A. & 
a reflection of disposition : A multiple
 source causal 
 Decision 
analysis. Organizational Behavior and H
uman
Processes. 
k: 
Kenny, D.A. (1979). correlation and ca
usality. New Yor
Wiley. 
w York: The 
Kotter, J.P. (1990). A force for chang
e. Ne
Free Press. 
rship 
Kouzes, J.M. 
e
& Posner, B.Z. (1987). The lead
challenge. San Francisco, CA: Jessey-Ba
ss Publishers. 
Locke, E.A. (1986). Generalizing from l
aboratory to field 
settings. Lexington, MA: Lexington Boo
ks. 
214 
Locke, E.A. (1991) ? The motivation sequence, the 
motivation hub, and the motivation core. 
Organizational Behavior and Human Decision Processes, 
50, 288-299. 
Locke, E.A., Kirkpatrick, S.A., Wheeler, J., Schneider, 
J., Niles, K., Goldstein, H., Welsh, K., & Chah, D. 
(1991). The essence of leadership. New York: 
Lexington Books. 
Locke, E.A., Frederick, E., Lee, C., & Bobko, P. (1984). 
Effect of self-efficacy, goals, and task strategies 
on task performance. Journal of Applied Psychology, 
69, 241-251. 
Locke, E.A. & Latham, G.P. (1990). A theory of goal 
setting & task performance. Englewood Cliffs, NJ: 
Prentice Hall. 
Manz, c.c. & Sims, H.P. Jr. (1989). Superleadership. New 
York: Berkley Books. 
Manz, c.c. & Sims, H.P. Jr. (1991). Superleadership: 
Beyond the myth of heroic leadership. Organizational 
Dynamics, Spring, 18-35. 
McClelland, D.C. (1985). Human motivation. Glenview: IL: 
Scott Foresman . 
Mento, A.J., Locke, E.A., & Klein, H.J. (in press). The 
relationship of goal level to valence and 
instrumentality. Journal of Applied Psychology. 
215 
organ, D.W. (1988). Organizational citizenship Behavior. 
Lexington, MA: Lexington Books. 
Podsakoff, P.M., MacKenzie, S.B., Moorman, R.H., & Fetter, 
R. (1990) . Transformational leader behaviors and 
their effects on followers' trust in leader, 
satisfaction, and organizational citizenship 
behaviors. Leadership Quarterly, i, 107-142. 
Peters, T. (1987). Thriving on chaos. New York: Harper & 
Row. 
Pfeffer, J. (1977). The ambiguity of leadership. Academy 
of Management Review,~' 104-112. 
Rosener, J.B. (1990). Ways women lead. Harvard Business 
Review, November-December, 119-125. 
Rosenthal, R. & Rosnow, R.L. (1969). Artifacts in 
behavioral research. New York: Academic Press. 
Sashkin, M. (1986). True vision in leadership. Training 
and Development Journal, May, 58-61. 
Sashkin, M. (1988). The visionary leader. In J.A. Conger & 
R.N. Kanungo (Eds.), Charismatic Leadership. San 
Francisco: Jessey-Bass. 
Sims, H.P . (1977). The leader as manager of reinforcement 
contingencies: An empirical example and a model. In 
J.G. Hunt and L.L. Larson (Eds.), Leadership: The 
cutting edge. Carbondale, IL: Southern Illinois 
University Press. 
216 
Smith, B.J. (1982). An initial test of a theory o f 
charismatic leadership based on responses of 
subordinates. University of Toronto: Unpublished 
Doctoral Dissertation. 
Tucker, R.C. (1968). The theory of charismatic 
leadership. Daedulus, 97, 731-756. 
Tichy, N.M. & Devanna, M. A. (1986). The transformational 
leader. New York, NY: Wiley and Sons. 
Uleman, J.S. (1991). Leadership ratings: Toward focusing 
more on specific behaviors . Leadership Quarterly,~' 
175-188. 
Weber, M. (1947, originally published in 1924). The 
theory of social and economic organization. A.M. 
Henderson & T. Parsons, trans . ; T. Parsons, Ed. New 
York: Free Press. 
Webster's New Collegiate Dictionary. (1974). Springfield, 
MA: G&C Merriam Co. 
Willner, A.R. (1984). The spellbinders: Charismatic 
political leadership. New Haven: Yale University 
Press. 
Yukl, G. A. (1989). Leadership in organizations, 2nd 
edition. Englewood Cliffs, NJ: Prentice Hall. 
Yukl, G.A. & Van Fleet, D.D. (1982). Cross-situational, 
multimethod research on military leader 
217 
effectiveness . Organizational Behavior and Human 
Performance, 30, 87-108 . 
218 
Append i x A 
Vision Script 
ke a pledge that our customers will At BKC, we ma
receive high quality printing and bindin
g . Let me explain 
 and 
my vision for BKC. This vision describ
es my long-term
eal goals for BKC it i s the direc
tion we will be 
id
heading well into the next cent ury. 
From the first day of business, I have pr
ided myself on 
the fact that BKC strives to g i ve the 
customer a quality 
ly days of the company when I had only product. In the ear
a handful of employees, I wou l d often
 help assemble the 
pleted an 
binders on large orders. One time w
e had com
e customer called and told us that they entire order when th
had rewritten a set of pages. They need
ed us to insert the 
t of pressure with 
new pages right away. We were under a l
o
other orders, so I explained to my em
ployees that BKC is 
ssemble quality products and that we 
will do 
here to a
whatever the customer wants. The custom
er was amazed that 
we made the changes so quickly and accu
rately. 
just a few months ago, a large order of 3
0,000 
In fact, 
binders was due to be shipped out one
 Friday afternoon. 
 that 
While packing the binders in boxes, a su
pervisor found
many of the binders had errors. Well, 
the supervisor got 
together with the binder employees and
 they d e cided that 
219 
they would work into the night and even come in ov
er the 
weekend to fix the mistakes. They made sure th
at the 
binders "meant quality." 
When BKC began to grow quickly, I realized that 
r 
needed a way to tell each employee that we are in bu
siness 
to produce top quality products. So, I develope
d the 
following company vision of where we are heading: 
BKC Printing Company is in the business of 
providing both national and international binding 
and printing services. We pledge to our 
customers that every binding and printing job 
will be perfect. In order to have high quality 
products, we are constantly striving to develop 
new techniques and methods of delivering perfect 
binding and printing. We are on the road to 
providing the highest quality products through 
continual striving for improvement. 
To sum up in a vision statement, "BKC is bound to
 
quality!" 
Our competitors are located throughout the Unite
d 
States. You know, a couple of months ago, one of our main 
competitors, Decautur Printing - they are located nea
r Fort 
Wayne, Indiana -- told me that they don't think we ar
e good 
enough to compete with them. I know that we are. So
, your 
performance during this experiment will be compared 
to the 
220 
performance of employees at Decautur Printing and at another 
competitor - Skyway Publishing in Colorado Springs. 
I know that this task is pretty difficult, but I really 
think that you can do well. It has been my experience that 
business students like yourself pick up the task pretty 
quickly and are able to turn out the high quality binders 
that BKC is known for. We're counting on you and I think 
you' 11 do a super job. Just remember, "BKC is bound to 
quality!" 
221 
Appendix B 
No Vision Script 
About 80% of our jobs are printed on high-grade machi
ne 
coated paper while the remainder are printed on high
-grade, 
long-grain paper that is similar to what most people
 use in 
copy machines. Because our business revolves around
 paper, 
the pages, and cardboard, for the binders, I would l
ike 
for 
to tell you a little bit about the process of making
 paper. 
The process begins with trees. After bark is remove
d, 
logs are fed into a "chipper" which cuts the logs in
to wood 
chips. These chips are then ground up using water 
and an 
e 
abrasive stone. Sometimes hot steam is used to sof
ten th
chips. Depending on the type of paper desired, a var
iety of 
other machines are used in conjunction with several ty
pes of 
chemicals in order to soften the chips. At this stag
e, the 
ground up chips are called mechanical pulp. 
ext, the pulp undergoes mild chemical treatment,
 
N
usually consisting of a sodium sulfite solution b
uffered 
with sodium carbonate . Sometimes a "no-sulfur" sodium
 ash 
is used instead. These chemicals are called the "
cooking 
liquor." The pulp is cooked in the liquor unde
r high 
temperatures and pressures. The cooking process elim
inates 
the non-cellulose fibers from the wood components. T
he used 
liquor can serve as an ingredient for a variety of
 other 
222 
purposes, including tanning agents, linoleum adhesives, and 
a road paving material. 
Next, the pulp is bleached to produce white fibers. 
The type of bleaching operation depends on several factors: 
the type of wood used to make the pulp, the pulping process, 
the degree of whiteness desired, and the purpose for which 
the paper will be used. For some pulps, a strong bleaching 
agent is required such as hydrogen peroxide. A newer 
technique uses a combination of chlorine and oxygen to 
bleach the pulp. Instead of bleaching the pulp all at once, 
there are usually several stages of bleaching. 
After the bleaching is complete, the pulp fibers are 
washed to remove chemicals and impurities. They are then 
give a mechanical treatment called refining, which makes the 
fibers stronger. Rosin and alum are added to increase water 
resistance so that the paper is suitable for pen-and-ink 
writing. At this stage, pigments and dyes are added if 
colored paper is desired. 
At this point, the fibers, which contain mostly water, 
travel through a machine at a very fast rate. The machine 
drains the water using suctioning devices as the fibers go 
through. The result is a wet web of paper that is carried 
on a conveyer belt to a pressing machine that smooths and 
dries the paper. 
By now, the paper is over 20 feet wide and on large 
223 
rolls. If desired, coating materials are added which produc
e 
a smooth or special surface. It is ready to be cut to siz
e 
and finished for shipment. The rolls are trimmed, sorted
, 
counted, and packaged. The paper is then transported to th
e 
customer, in this case, BKC Printing. 
224 
Appendix C 
Task Strategy Script 
(Note: For the task strategy condition, this 
script 
followed either the vision or no vision script]
 
Before we start, our training department has dev
eloped 
work methods so that there are fewer errors. You
 might want 
to try some of these, or if you have any of your
 own ideas, 
you might want to use them -- it's up to you. 
lace 
You may want to use the empty desks around you. 
P
the instructions somewhere you can refer to the
m quickly. 
Place the section dividers and pages on your r
ight or on 
r desk in front of you, the binder you are worki
ng on 
anothe
in front of you, and place the other materials 
to your on 
another desk- to your left. 
nother helpful thing to do is to write out the o
rder 
A
that the pages will go in - on a piece of scrap 
paper or on 
the 
the back of a questionnaire, please do not w
rite on 
instructions or the answer key - and then insert 
these pages 
into the binder once you know the order of the
 pages for 
that section. If you write out the order f
or several 
elf enough 
sections at a time, be sure that you allow your
s
time to gather and insert the pages. 
Also, read the instructions carefully. Many of the 
225 
have similar titles, so be sure to take yo
ur time. 
pages 
t me give you a hint - do not to spread the 
pages out 
Le
 the floor or desk - this takes too much 
time. Because 
on
ages are in alphabetical order, we have 
found that 
the p
people do better if they leave the pages sta
cked up, pulling 
the pages out as needed. Again, these are 
only suggestions 
- its up to you to decide whether to use th
em or not. 
Does anyone have any questions? 
226 
Appendix D 
Attitude and Perception Items 
Task Satisfaction 
In general, I was satisfied when doing the
 task. 
I felt satisfied when assembling the binde
rs. 
I am generally satisfied when doing the ta
sk. 
I did not feel satisfied during the task. 
Congruence of Beliefs and Values 
. 
Quality is the most important aspect of th
is job
important than 
Inserting as many pages as possible is mor
e 
making sure there are no errors. 
ders. 
The CEO is too fanatical about having high 
quality bin
lity. 
The CEO has good reasons for being concerne
d about qua
Trust in the Leader 
I have complete trust in the CEO. 
I do not believe what the CEO says. 
The CEO is trustworthy. 
I find it difficult to trust the CEO. 
Liking the Leader 
This CEO is a really likeable person. 
The CEO seems like a terrific person . 
I would have trouble getting along with th
e CEO. 
I really don't like the CEO. 
227 
Charisma - Form 8 8 
The CEO served as a role model for me. 
The CEO mobilized a collective sense of mission . 
The CEO instilled pride in being associated with him. 
The CEO engaged in words and deeds which enhanced his 
image of competence. 
The CEO made me aware of strongly held values, ideals, and 
aspirations which are shared in common. 
The CEO demonstrated a strong conviction in his beliefs and 
values. 
The CEO projected a strong, dynamic, a~d magnetic presence. 
I am ready to trust the CEO to overcome any obstacles. 
I have complete confidence in the CEO. 
In my mind, the CEO is a symbol of success and 
accomplishment. 
Charisma - Form 58 
The CEO makes me feel good when I'm around him. 
The CEO makes me proud to be associated with him. 
The CEO is someone in whom I have complete faith. 
? Form SY items are taken from Bass (1985). Form 5R items 
are copyrighted by B. Bass and B. Avolio. The survey f rom 
which the items are taken is licensed by Consulting 
Psychologists Press (CPP), Inc. and any reproduction or 
use of the survey and/or items must be appr oved by CPP . 
228 
The CEO has a special gift for seeing w
hat is really 
worthwhile for me to consider . 
The CEO has my respect. 
The CEO shows enthusiasm for what I nee
d to do. 
The CEO has a sense of mission which he 
communicated to me. 
CEO increases my optimism for the futur
e. 
The 
The CEO has my trust in his abilit
y to overcome any 
obstacle. 
Enthusiasm for Task 
ldn't wait to get started on this task. I cou
I was not excited about doing this task
. 
I thought this task was really fun. 
I was enthusiastic about doing this tas
k. 
Enthusiasm for Vision 
about assembling high quality binders. I was enthusiastic 
I was not excited about making sure eac
h binder section 
had no errors. 
I found that trying to make no errors w
as a lot of fun. 
y exciting. 
The idea of trying for few errors was r
eall
Willingness To Work in Future 
ork for the CEO on another experiment i
n 
I would like to w
the future even if it meant getting pai
d minimum wage. 
n another 
I would volunteer my time to participat
e i
experiment for the CEO. 
I would NOT be willing to work for this
 CEO in a real job 
229 
in the future. 
e pay were 10% below average, I wo
uld like to work 
Even if th
for the CEO in a future job. 
Intellectual Stimulation 
 forced me to rethink some of my 
own 
The CEO's ideas have
ideas about performing this task . 
ng the 
I felt I learned something from t
he CEO about doi
task. 
The CEO provided me with new idea
s about assembling the 
binder sections. 
performing the task. 
The CEO didn't give me any ideas a
bout 
t assembling 
I didn't really learn anything from
 the CEO abou
the binders. 
Inspiration 
e CEO motivated me to do better tha
n I originally expected 
Th
I would do. 
The CEO aroused in me the effort 
to work better. 
ecause of the CEO, I did better tha
n I expected I could do. 
B
 CEO stimulated my efforts to exc
el. 
The
The CEO heightened my motivation 
to succeed. 
ure. 
The CEO increased my optimism for
 the fut
Task Clarity 
I knew exactly what to do on this
 task. 
I was not very sure how to insert
 the pages. 
I knew how I was supposed to inse
rt the pages. 
230 
I was not clear about how the task was to be done. 
231 
Appendix E 
Goal and Self-efficacy Items 
Self-efficacy Magnitude - Quality 
Below at the left are listed the number of errors th
at you 
might make when inserting the pages in the next 15 mi
nutes. 
Indicate with a Y for "YES" or a N for "NO" wheth
er you 
think you can make that many err ors or less. 
ERRORS Y OR N 
20 or less 
18 or less 
16 or less 
14 or less 
12 or less 
10 or less 
8 or less 
6 or less 
4 or less 
2 or less 
0 
Self-efficacy Magnitude - Quantity 
Below at the left are listed the number of pages th
at you 
might insert into the binder in the next 15 mi
nutes. 
Indicate with a Y for "YES" or a N for "NO" wheth
er you 
think you can complete that many pages or more. 
HOW CONFIDENT ARE YOU? 
232 
PAGES (0 TO 10) 
0 
5 or more 
10 or more 
15 or more 
20 or more 
25 or more 
30 or more 
35 or more 
40 or more 
45 or more 
50 or more 
Self-efficacy Strength - Quality 
Indicate your degree of confi
dence in being able to make 
at number of errors or less o
n a scale of o (no confidence 
th
tes. 
at all) to 10 (complete conf
idence) in the next 15 minu
 your confidence in making t
hat number of errors or 
This is
less. 
HOW CONFIDENT ARE YOU? 
ERRORS (0 TO 10) 
20 or less 
18 or less 
16 or less 
14 or less 
12 or less 
233 
10 or less 
8 or less 
6 or less 
4 or less 
2 or less 
0 
Self-efficacy Strength - Quanti
ty 
Indicate your degree of confide
nce in being able to insert 
that number of pages or more on 
a scale of o (no confidence 
all) to 10 (complete confidence)
 in the next 15 minutes. 
at 
s is your confidence in complet
ing that number of pages 
Thi
or more. 
HOW CONFIDENT ARE YOU? 
PAGES (0 TO 10) 
0 
5 or more 
10 or more 
15 or more 
20 or more 
25 or more 
30 or more 
35 or more 
40 or more 
45 or more 
50 or more 
234 
ipated Satisfaction - Qualit
y 
Antic
fied 
Indicate the degree to which 
you anticipate being satis
ow. 
with each level of performanc
e with respect to errors bel
DE ANTICIPATED SATISFACT
ION 
IF YOU MA
IS MANY ERRORS (1 TO 
9) 
TH
20 or less 
18 or less 
16 or less 
14 or less 
12 or less 
10 or less 
8 or less 
6 or less 
4 or less 
2 or less 
0 
ipated Satisfaction - Quanti
ty 
Antic
te being satisfied 
Indicate the degree to which 
you anticipa
 of 
with each level of performa
nce with respect to number
pages inserted below. 
235 
IF YOU INSERTED ANTICIPATED SATISFACTION 
THIS MANY PAGES (1 TO 9) 
0 
5 or more 
10 or more 
15 or more 
20 or more 
25 or more 
30 or more 
35 or more 
40 or more 
45 or more 
50 or more 
Personal Goal - Quality 
What is the most number of errors that you would find 
acceptable to make during the next 15 minutes, that is your 
minimum goal for quality? 
errors 
Goal Commitment - Quality 
I am very enthusiastic about this quality goal. 
I am strongly committed to pursuing this quality goal. 
rt wouldn't take much for me to abandon this quality goal. 
236 
rt is quite likely that this quality goal may need to be 
revised, depending on how things go. 
I think this quality goal is a good goal to shoot for. 
Personal Goal - Quantity 
What is the least number of pages that you would find 
acceptable to complete during the next 15 minutes, that is 
your minimum goal for quantity? 
pages inserted 
Goal Commitment - Quantity 
I think this quantity goal is a good goal to shoot for. 
I am strongly committed to pursuing this quantity goal. 
Quite frankly, I don't care if I achieve this quantity goal 
or not. 
Since it is not always possible to tell how tough a task 
is until you've done it for awhile, it's hard to take this 
quantity goal seriously. 
I am very enthusiastic about this quantity goal.