ABSTRACT 
 
Title of Thesis: RELATIONSHIP OF SETTING AND INTERNAL 
ATTITUDES ON THE POSITIVE AND NEGATIVE 
INTERACTIONS BETWEEN DIFFERENT RACE COLLEGE 
STUDENTS 
 
Degree candidate: Warren L. Kelley 
 
Degree and year: Master of Arts, 2003 
 
Thesis directed by: Professor Mary Ann Hoffman 
 Department of Counseling and Personnel Services 
 
This thesis investigated how the quality of cross-racial interactions varied 
across situations in a campus environment and also examined the influence of internal 
attitudes toward diversity on the quality of these interactions.  The influence of  race, 
gender and cohort on cross racial interactions was also examined.   The stratified, 
random sample of 1,000 students included 250 students each from four racial/ethnic 
categories - White, African American/Black, Asian Pacific American, and 
Hispanic/Latino.  Respondents completed the Miville-Guzman Universality-Diversity 
Scale-Short, excerpts from the Cultural Attitudes and Climate Questionnaire, and the 
author developed Cross Racial Interaction Scale to measure the quality of cross racial 
interactions in 13 situations on a college campus.  This study found that the quality of 
cross racial interactions did vary by situation and that students clustered in identifiable 
ways related to the similarity of their cross racial experiences.  Evidence was also 
found in support of previous research that Black students continue to perceive their 
cross racial experiences less positive than their White counterparts.  
 
RELATIONSHIP OF SETTING AND INTERNAL ATTITUDES ON THE 
 
POSITIVE AND NEGATIVE INTERACTIONS 
 
BETWEEN DIFFERENT RACE COLLEGE STUDENTS 
 
By 
 
Warren L. Kelley 
 
Thesis submitted to the Faculty of the Graduate School of the 
University of Maryland, College Park in partial fulfillment 
 of the requirements for the degree of 
 Master of Arts 
 2003 
 
Advisory Committee: 
 
Professor Mary Ann Hoffman, Chair 
 Professor Dennis Kivlighan, Jr. 
 Dr. Linda Tipton 
 Professor William Sedlacek 

ii 
 
TABLE OF CONTENTS 
 
List of Tables iii 
 
Chapter 1 - Introduction 1 
 
Chapter 2 - Literature Review 7 
Campus Climate Studies 7 
Cross Racial Interactions in Different Settings 17 
Attitudes Toward Diversity and Cross Racial Interactions 23 
Statement of Problem 27 
 
Chapter 3 - Method 33 
Design Statement 33 
Participants 33 
Measures 34 
Procedures 40 
 
Chapter 4 - Results 43 
Preliminary Analysis 43 
Hypothesis One 50 
Hypothesis Two 58 
Hypothesis Three 60 
Additional Analysis 63 
Research Question One 63 
Research Question Two 63 
Preliminary Validity Check of CRIS 66 
 
Chapter 5 - Discussion 70 
Summary of Findings 78 
Limitations 82 
Implications for Future Research 85 
 
Appendices 
Appendix A  88 
Appendix B  89 
Appendix C  90 
Appendix D  91 
Appendix E  92 
Appendix F  93 
Appendix G  94 
References 95 
iii 
 
LIST OF TABLES 
 
1. Demographic Characteristics of Sample 47 
 
2. Pearson Correlation Coefficients for CRIS Scores 48 
 
3. Clusters Mean CRIS Scores Per Situation 56 
 
4. Demographic Characteristics of the Seven Clusters 57 
 
5. Mean CRIS Scores by Race 59 
 
6. Pearson Correlation Coefficient for CRIS and M-GUDS-S Scores 62 
 
7. CRIS Scores (Versions A and B) by Race 65 
 
8. Pearson Correlation Coefficients for CACQ Items 68 
 

1
 Chapter 1 
 Introduction 
 Fostering healthy interpersonal relationships between people of different 
racial/ethnic groups is important in a country with rapidly increasing ethnic minority 
populations.  Yet there are signs that segregation is increasing along with evidence of 
growing racial and ethnic hostility ((American Commitments National Panel, 1995; 
Mack, Tucker, Archuleta, DeGroot, Hernandez & Cha, 1997; Pascarella, Edison, 
Nora, Hagedorn & Terenzini, 1996; Ponterotto & Casas, 1991).  For example, the 
Southern Poverty Law Center (1999) reports that there are more than 500 hate groups 
operating in the United States. 
 One way to improve race relations is to improve interpersonal interactions 
across racial/ethnic lines.  There is strong evidence that increasing positive cross racial 
interactions is associated with improved racial attitudes (Powers & Ellison, 1995; 
Sigelman & Welch, 1993) 
 Universities provide prime opportunities to increase positive cross racial 
interaction. It is often the first place students have experiences in a diverse 
racial/ethnic community (American Commitments National Panel, 1995).  A solid 
majority of Americans view higher education as very important in preparing people to 
function in a more diverse society (DYG, Inc., 1998).  
 Students on college campuses express a strong desire for an increase in cross 
racial interaction and communication (Duster, 1993; Fisher & Hartmann, 1995).  
2
 Students have stressed the importance of cross racial friendships as a means of 
learning about others, building a sense of understanding between diverse groups and 
reducing prejudicial attitudes (Humphreys, 1998; Nesdale & Todd, 1998). It has been 
shown that friends and peers are a significant source of influence on students? 
openness to diversity (Pascarella et al., 1996) and attitudes about racial issues 
(Newswanger, 1996).  Studies on college campuses have found a positive correlation 
between increasing, positive cross racial/ethnic interaction and academic development 
(e.g., critical thinking) and satisfaction with college (Astin, 1993; Helm, Sedlacek & 
Preito, 1998).  
 Ironically even while there is growing diversity on college campuses (Carter & 
Wilson, 1998) and students desire greater levels of cross racial interaction, they are 
often disappointed by the reality they find (Humphreys, 1998; Kropp, 1992; 
Lederman, 1993; Reisberg, 1999; Smith 1997). 
 Universities, therefore, provide a unique opportunity in society to improve race 
relations by improving cross racial interactions.  The public wants it, students want it, 
and there are positive learning outcomes as a result.  However, actually achieving 
positive cross racial interactions remains a challenge.  It is within this context that the 
current study will focus on understanding better, and potentially improving, these 
interactions. 
Are there situations (e.g, classrooms versus residence hall) on a college 
campus where cross racial interactions are relatively more positive than others?  If so, 
3
 these situations might provide clues to better understanding how to improve cross 
racial interactions across other settings. Campus climate studies represent one line of 
research that provide a better understanding of the nature of cross racial interactions 
on college campuses.   Some climate studies have focus on specific settings like 
residence halls (Johnson-Durgans, 1994; Newswanger, 1996), or compared two 
settings (Korgen, Mahon & Wang, 2003; McClelland & Auster, 1990), but none have 
been found that systematically compared the quality of cross racial interactions across 
the different situations in the college environment.  Doing so would help in 
determining where positive (and negative) cross racial interactions are occurring.  
 Many campus climate studies focus on the perceptions students from different 
racial or ethnic backgrounds have about their experiences on campus.  These studies 
generally report that minority students are experiencing more negative campus 
experiences than White students (Ancis, Sedlacek & Jonathan, 1998). Minority 
students feel more socio-cultural alienation (Mack et al., 1997), feel the climate is less 
friendly (McClelland & Auster, 1990), and are significantly less satisfied (Fisher & 
Hartmann, 1995; Student Affairs Research Services, 1994).  
 Another line of research that can be helpful in understanding the relationship 
between different campus situations and the quality of cross racial interactions 
revolves around Allport?s (1954) contact hypothesis.   Subsequent work has refined 
the original contact hypothesis (Amir, 1969; Brewer & Brown, 1998; Cook, 1985; 
Pettigrew, 1998).  The central premise of the contact hypothesis is that bringing 
4
 different groups (e.g., racial/ethnic) together  under certain prerequisite conditions will 
promote the development of more harmonious intergroup relations and reduce 
intergroup tensions (Brewer & Brown, 1998; Gaertner, Rust, Dovidio, Bachman & 
Anastasio, 1994). Since the characteristics of situations on college campuses can vary 
widely across classrooms, dorm rooms, work settings and social activities the contact 
hypothesis provides a theoretical basis to suggest that the quality of cross racial 
interactions may also vary across these different situations. 
 A qualitative campus climate study conducted at a large, Eastern, public 
university, using the contact hypothesis as a basis found that positive, as well as 
negative cross racial interactions of students centered around four major 
environmental settings: the academic setting; the employment setting; the residential 
setting; and the social setting (Cotton, Kelley & Sedlacek, 2000).  Within these 
settings, specific situations were repeatedly mentioned as important locations of cross 
racial interaction, for example, classrooms, recreation activities and parties.  Cotton et 
al. supports the hypothesis that the quality of cross racial interactions will vary by 
different situations in the campus environment, however, this qualitative study cannot 
be generalized.  Nor does it systematically measure the positive or negative nature of 
these interactions across other situations.  
 Is it possible that in trying to understand cross racial interactions, rather than 
focusing on the situation, it is more important to understand the internal attitude of a 
person toward people of other races?   There is a clear link between one?s attitudes and 
5
 many forms of social behavior (Baron & Byrne, 1994).   For example, prejudice is an 
attitude (usually negative) toward a specific group and has been shown to influence 
social interactions with people from that group (Harris et al., 1992).  This suggests that 
a student?s attitude toward diversity may influence the quality of their interaction  with 
students from other races.   
 No study has been found that systematically investigated how both situations 
(e.g., classrooms, residence halls) and internal attitudes toward diversity on a college 
campus may influence the quality of the cross racial interactions among students.   
Counseling psychologists could benefit from such a study.  Understanding the context 
in which more healthy relationships occur between students of different races and 
ethnicities on college campuses has relevance both to the field of psychology 
generally and to counseling psychology specifically.  Negative intergroup 
relationships have been related to increased anxiety in individuals (Brewer & Brown, 
1998; Stephan & Stephan, 1985) as well as the arousal of fear, disgust, contempt, 
anger and jealousy (Smith, 1993).   Negative intergroup anxiety leads to avoidance of 
people in the other group (Stephan & Stephan).  Cultural mistrust can be fed by 
negative cross racial experiences and inhibit the relationship between Black students 
and White counselors (Atkinson, Morten & Sue, 1998). On the other hand a positive, 
interracial climate is associated with increases in individual students? effective coping 
strategies, higher self-esteem and better academic performance in elementary school 
children (Marcus-Newhall & Heindl, 1998).   
6
 The purpose of the present correlational study will be to investigate both how 
the quality of cross-racial interactions vary across situations in a campus environment 
and also to examine the influence of the person?s internal attitude on the quality of 
these interactions.  The influence of  race, gender and cohort on cross racial 
interactions will also be examined.  The study will use a random sample of students 
from the undergraduate population of a large and racially diverse mid-Atlantic 
university. 
 
7
 Chapter 2 
 Literature Review 
 The following chapter section reviews relevant literature related to 
understanding cross racial interactions in a college environment.  First, campus 
climate studies will be reviewed, focusing on those studies that recognize the different 
settings that exist in a college environment and those studies that explore the different 
experiences racial groups have on campuses.   Cotton et al. (2000), a campus 
situational study, will be examined in relative depth.  Next, cross racial interactions in 
different settings will be investigated, with a number of studies revolving around the 
concept of the contact hypothesis (Allport, 1954), and others grounded in 
environmental research.  Last, the literature involved in attitudes toward diversity and 
how they might play into cross racial interactions is reviewed. 
Campus Climate Studies
 Campus climate studies provide a broad understanding about the state of 
college  environments.  While these studies encompass a rich diversity of sub 
communities (e.g., race, gender, sexual orientation, religion), the current literature 
review will focus on those pertaining to race.  Many of these studies investigate how 
students of different races are perceiving and experiencing the campus environment 
(Ancis, Sedlacek & Mohr, 1998; Mack et al., 1997; McClleland & Auster, 1990), 
while some focus on specific settings such as residence life (Johnson-Durgans, 1994; 
Newswanger, 1996).  They all provide information about students? interpersonal 
8
 experiences with other students. Climate studies support the idea that there are racial 
differences in perceptions about campus climates.  This literature also lends support to 
the notion that different situations can influence the quality, as well as quantity, of 
cross racial interactions (McClleland & Auster, 1990; Nesdale & Todd, 1998). 
 Differences by environment. Several climate studies investigate the quantity 
of cross racial interaction in different settings within the college environment 
(Hurtado, Dey & Trevino, 1994; McClelland & Auster, 1990).    McClelland and 
Auster (1990) studied the racial climate at a small, Eastern private liberal arts college; 
largely residential and predominantly White.  Data included both qualitative focus 
groups and surveys, with 186 White and 20 Black randomly selected students (Black 
students were over sampled but the total population was extremely small).  They 
compared the level of cross racial interaction occurring at officially sanctioned school 
functions (e.g., football games) and purely social functions (e.g., parties) and found 
that there was less interaction at official school functions than at social functions 
reported by both Black and White students 
 McClelland and Auster (1990) wondered what were the characteristics of the 
official school functions that tended to produce greater segregation, than social 
functions and conjectured about the role that the institution has in hindering or 
promoting intergroup contact. When asked about their perception of the racial climate 
in the social and school function settings, however, 70-75% of African American 
students and 90-93% of White students said the climate was friendly.  This study is 
9
 useful since it compares cross racial interactions in two settings within the campus 
environment. 
 Hurtado, Dey and Trevino (1994) studied the interaction among different 
race/ethnic students and found that the minority students interacted with White 
students more than White students interacted with minority students. For example, 
they found that minority students dined more frequently with someone of a different 
ethnic or racial background than White students. They also found that 42% of Asian 
American students reported interracial or ethnic dating, compared to 24% of Mexican 
Americans, 13% of African Americans and 4% of White students.  These findings are 
counter to the popular perception that minority students self segregate more than 
White students.  This is important since it questions a widely held myth that one key 
to increasing cross racial interaction is to reduce the self-imposed self segregation of 
minority students (Hurtado et al.). This study is also one of the first attempts to 
systematically assess the quantity of cross racial interaction in a variety of situations 
(e.g., dining, dating). However, it does not attempt to study the quality of these cross 
racial interactions.  The contact hypothesis (Allport, 1954)  informs us that simple 
contact is not sufficient to guarantee positive experiences, and therefore, it is 
important to also understand the positive or negative nature of the cross racial contact 
that does occur. 
 Korgen, Mahon, and Wang (2003) compared the level of racial tension, 
interracial friendships and dating at two comparable Northeastern universities, but 
10
 with different racial demographics.  The first institution was relatively homogeneous 
with students of color comprising only 6% of the population while the second was 
more heterogeneous with students of color representing 29% overall and 38% of the 
students living on campus.  The study included 164 students from the first institution 
and 152 from the second. Korgen et al. found that the second institution reported a 
higher level of cross racial friendships.  This might be expected because of the higher 
minority levels.  However, they discovered counterintuitive results when isolating the 
residential environment from the commuter environment.  Residential students at the 
second institution reported significantly higher racial tension, and were far less likely 
to have dated interracially than their commuter counterparts.  The authors conjecture 
that this might be due to a so-called tipping effect (Glazer, 1995, as cited in Korgen et 
al., 2003), that as the number of minority members grow in an area and majority 
members shrink, segregation can actually increase without intentional interventions to 
create a positive environment.  This study is useful in the current review because it 
reflects how cross racial interactions can vary between the commuter and residential 
environments on a college campus. 
 Newswanger (1996), studied whether there were significant differences in 
White racial identity attitudes using the White Racial Identity Attitudes Scale (Helms 
& Carter, 1990) between White students who had lived with African American 
roommates in residence halls and those who did not.  Participants included 187 White 
students only from an Ivy League institution in four categories: residential students 
11
 assigned with Black roommates; randomly selected first year students; randomly 
selected seniors; and residential advisors.  No difference was found in White racial 
identity attitudes between White students living with Black roommates and other 
White students. However, White roommates reported the highest number of both
 positive and negative changes in thinking about race.   White students cited the lack of 
preparation for encounters with their Black roommates, preexisting negative thoughts 
and feelings about racial differences, and mixed actual experiences with interactions 
as reasons for negative changes in thinking about race.  Newswanger observed that 
structured environments and positive interventions need to be created in order to 
improve cross racial interactions in residence hall settings. A limitation of this study is 
that Newswanger did not examine environments outside of the residence halls. 
 The Campus Assessment Working Group (2003) conducted 21 qualitative 
focus groups with both racially homogeneous and heterogeneous groups and involving 
157 undergraduate students at a large, mid-Atlantic university.  They found that 
students expressed experiences of positive interaction with students of other races in 
classroom and residential settings.  The authors observed that these settings may 
provide the informal, ongoing opportunity for contact that can foster positive 
relationships.  They also found that students reported that large scale, campus wide 
events (e.g., major art, entertainment and even sporting events) which were not 
focused on any particular race or culture, but were attended by all races were 
experienced as positively contributing to a sense of interpersonal connection across 
12
 racial lines.  This was true even when students didn?t actually interact directly with 
someone of another race. 
 Differences by race. From a number of studies a picture emerges of different 
perceptions between White students and minority students, particularly African 
American students.  Minority students feel more socio-cultural alienation (Mack et al., 
1997); African American students feel the climate is less friendly (Ancis, Sedlacek, 
and Mohr, 1998; McClleland & Auster, 1990); and significantly less satisfying than 
White students (Fisher & Hartmann, 1995; Student Affairs Research Services, 1994).  
 In a study of perceptions and experiences about the campus cultural climate, 
Ancis, Sedlacek, and Mohr (1998) found significant differences between racial and 
ethnic groups on a number of dimensions related to the climate.  Respondents 
comprised 578 African American, Asian American, Latino, and White undergraduates 
(60% return rate) from a large mid-Atlantic university.  African American students 
consistently reported more negative experiences compared to Asian American, Latino, 
and White students.  Specifically, African-American students  experienced greater 
racial/ethnic hostility; greater pressure to conform to stereotypes; less equitable 
treatment by faculty, staff, and teaching assistants; and more faculty racism than their 
Asian-American, Latino, and White counterparts.  White students consistently 
reported less racial tension, few expectations to conform to stereotypic behavior, an 
experience of being treated fairly, a climate characterized by respect for diversity and 
the most overall satisfaction. 
13
  Mack et al. (1997) surveyed 150 seniors at five small private, predominately 
White and highly selective  colleges in California.  White students had a positive view 
of cross racial interactions and did not perceive their race as a barrier or burden.  Black 
students tended to describe the climate as more hostile, uneasy, and uninviting, with 
Asian and Latino students holding views between these two contrasting groups.  For 
example, Black students were more likely than Asian or White students to agree that 
the campus was racially hostile and that students were uneasy in the presence of their 
group.  On the other hand White students were less likely than Asian, Black, and 
Latino students to perceive a lack of campus activities for their racial group. White 
and Latino students were most comfortable interacting with other racial groups, 
whereas Asian students were the least comfortable.  This study does not examine how 
these experiences, for all races, may vary by different environments on campus (e.g., 
classrooms, residence halls). 
 Johnson-Durgans (1994) surveyed students in the residence halls of a 
predominantly White university in the Midwest in order to examine differences in the 
perceptions of racial climates between African American and White students. 
Respondents totaled 3,347 (56.8% return rate) comprised of 10.3% African American, 
82.2% White and 7.5% other students of color.  The questionnaire included a series of 
items related to cross racial interaction with peers, such as, ?It does not matter in my 
hall what race you are, everyone gets along well? and ?The residents in my hall are 
friendly.?  Significant differences were found between the perceptions of African 
14
 American students and White students related to hall peers, as well as the resident hall 
environment, hall government, and hall staff.  The authors reported non-parametric 
results which only showed magnitude of these differences but not direction. 
 Fisher and Hartmann (1995) studied the racial climate of a predominately 
white, large Midwest university.  Their study included 120 White and 120 Black 
students (79% and 78% response rates respectively) who answered 23 open-ended 
questions.  They found that 44% of the Black respondents had personally been the 
target of other students? racial prejudice, six times higher than  reported by their White 
counterparts (7.5%).  Similarly, 54% of Black students and 46% of White students 
reported witnessing racial prejudice by others (about 90% targeting Blacks and 10% 
targeting Whites).  A majority of both groups felt that making friends with others was 
affected by race (79% of Black and 56% of White students).   This study did not 
examine whether these negative experiences were more or less prevalent in different 
situations on campus (e.g., in classrooms, in residence halls). 
 These studies related to racial differences report relatively consistent findings 
that White students are having more positive experiences than Black students, and  in 
some cases other students of color, in college environments.  It is reasonable to 
conjecture that in terms of peer-to-peer relationships, students of color will report a 
differing quality of cross-racial interactions than White students.  None of these 
studies have systematically explored how the quality of cross racial interactions might 
vary by race across situations on campus. 
15
  Cotton campus situational study. Cotton, Kelley and Sedlacek (2000) 
conducted a survey to gain student input about the situations in which they had 
positive and negative experiences with same race and different race students. The 
study included 75 in-depth, semi-structured interviews of undergraduate students, 
stratified by race and randomly selected at a large, public mid-Atlantic university.  The 
primary purposes of this qualitative study were to (1) better understand the situations 
in which students have positive and negative cross racial interactions, and (2) attempt 
to uncover common characteristics of those situations to guide strategies to enhance 
positive contact.  The study used the critical incidents technique (Flanagan, 1954), a 
strategy of interviewing to gather the critical characteristics of a situation or behavior.  
An inductive approach was used to analyze the responses and to develop a 
classification structure (Cotton et al.).  Multiple raters were used to enable assessment 
of inter-rater reliability.   
 Four themes emerged from the interviews: (a) Academic settings - Students 
described academic situations as an opportunity to engage in positive dialogue across 
racial and ethnic lines.  Although large classes presented challenges for interaction, 
students generally seemed to accept a safe classroom setting that encouraged 
discussion.  Small-group classroom discussions were especially conducive to cross 
racial interaction.  (b) Campus employment -  On and off campus employment 
experiences were frequently mentioned as opportunities for positive cross racial 
interaction.   Work settings could foster accomplishment and shared responsibilities 
16
 for achieving common goals.  (c)  Residential life -  In numerous cases, students 
indicated that living in the residence halls provided ongoing opportunities for 
interaction with students of different races in a non-threatening environment.  Planned 
programs as well as through informal contact and development of friendships were 
cited as ways in which these interactions might occur. (d) Social interaction -  Social 
activities provided frequent opportunities for positive (and negative) interactions 
among racial groups.  These activities include belonging to a student group, going to 
parties, participating in sports, taking part in a fraternity/sorority event, attending 
campus-wide events (e.g., basketball games), and going to off-campus activities.  
 Cotton et al. (2000) observed that the settings described by students varied 
from structured situations with significant institutional oversight (i.e., classrooms, 
workplaces) to unstructured situations in which students are free to choose with whom 
and how they will associate.  The structured settings provide more opportunities for 
campus intervention in shaping cross racial interactions.  The unstructured 
environments are less amenable to institutional intervention. 
 Cotton et al. (2000) concluded that faculty should be engaged to further 
promote the classroom opportunities for cross racial interaction; the work environment 
should be recognized for its potential value to enhance these interactions; and the 
social environment should be explored to find more safe, non-anxiety producing ways 
for students to interact across racial lines. 
 The study by Cotton et al (2000) is very helpful in depicting a schema for 
17
 categorizing students? cross racial interactions within the university environment.  
This schema includes broad settings (such as academic, work, residential or social) 
and specific situations within these settings (e.g., classrooms, formal residence hall 
programs, student groups, etc..).  The study supports the idea that students are 
experiencing unique, cross racial interactions specific to these settings and situations.  
 Are some broad settings or specific situations reported by Cotton et al. (2000) 
more conducive to positive cross racial interactions than others?  To what extent are 
the findings above accurate reflections of the entire student body?  While very 
informative, the findings by Cotton et al. cannot be generalized without a random 
sample, nor can situations (e.g, classrooms, formal residence hall programs, student 
groups) be compared to each other relative to cross racial interactions without a 
quantitative measure of these experiences.  
Cross racial interactions in different settings
 There is ample research to support the concept that the environment in which 
cross racial interactions take place can affect the quality of those interactions.  The 
contact hypothesis and environmental studies are two lines of research that inform this 
issue.  
 Contact hypothesis. Allport?s (1954) contact hypothesis originated a body of 
research that elaborates on the experiences of intergroup relations.  The contact 
hypothesis provides evidence that bringing people of different racial/ethnic 
backgrounds into contact with each other under certain conditions may lead to 
18
 improved interpersonal relations as described below. The contact hypothesis can 
provide a useful perspective in terms of investigating how cross racial interactions 
may vary depending on the characteristics of the different situations on a college 
campus.  
 As originally defined by Allport, and refined by later researchers, the four 
conditions (Brewer & Brown, 1998) that must be present in order for contact with  
members of different racial/ethnic groups to enhance intergroup relations include:  a) 
equal status of all group members within the contact situation; b) cooperative 
interdependence among group members; c) norms that support  positive relations; and 
d) interactions that disconfirm stereotypes and encourage the transmission of 
individuating information about group members (Allport, 1954; Cook, 1985; Gaertner, 
Dovidio & Bachman, 1996; Marcus-Newhall & Heindl, 1998; Pettigrew, 1998).  
Brewer and Brown observe that the contact hypothesis is one of most long-lived and 
successful ideas in the history of social psychology. 
 While many studies of the contact hypothesis were conducted in experimental 
settings (Cook, 1985; Cook & Pelfrey, 1985; Gaertner et al., 1989; Wilder, 1984; 
Wright, Aron, McLaughlin-Volpe and Ropp,1997), a number of studies were also 
conducted in field settings (Green, Adams & Turner, 1988; Marcus-Newhall & 
Heindl, 1998; Powers & Ellison, 1995).  Gaertner et al. (1994), for example, found in 
a study of 1,357 students attending a multi-cultural high school a correlation between 
a reduction in intergroup bias/prejudice and the presence of contact hypothesis criteria.  
19
 In general the field based studies such as Gaertner et al. confirm the validity of the 
contact hypothesis in actual settings. 
 Nesdale and Todd (1998) tested the contact hypothesis in a college residential 
environment.   The authors chose this setting because they saw it as meeting the 
requirements called for by the contact hypothesis.  Students have equal status and 
ample opportunity to meet each other and disconfirm negative stereotypes.  Hall 
settings typically emphasize collegiality and egalitarian social norms, and 
programming activities and hall staff encourage cooperation and collaboration.   The 
study included 246 students (127 international; 119 Australian) living in three 
residence halls in an Australian university.  In one hall the international students 
(mostly Chinese) were the majority, in the second hall the international and Australian 
students were about equal, and in the last hall the Australian students were the 
majority.  The authors found that the minority group in each hall had a significantly 
greater amount of cross-cultural contact with the majority than vice versa.  The 
authors also found that this level of contact generalized for students to many of their 
other settings on campus - classroom assignments, leisure activities, travel and general 
university activities, i.e., if a student reported high cross-cultural contact in the 
residence hall, she/he also reported high cross-cultural contact in other settings in the 
university.  This study is important because it examined the amount of cross cultural 
contact in a number of settings in a university environment as well as evidence that 
there seems to be a cross-over effect from one setting (e.g., residence hall) to another 
20
 (e.g., leisure activities).  While the study measured the quantity of contact, it did not 
attempt to measure the quality of these cross cultural contacts.   
 Pettigrew (1994) reinforced the importance of the contact hypothesis criteria to 
influence behaviors in a particular situation.  He asserts that one of the strongest 
findings from social psychological research is the power of situational norms to shape 
intergroup interaction. Further, that people are fully capable of immediate behavioral 
change as they move from one set of situational norms to another. Pettigrew cites 
several examples of field studies of this phenomenon. Minard (1952, as cited in 
Pettigrew, 1994) found that most Black and White coal minors followed a pattern of 
racial integration below the ground and racial segregation above the ground.  Reitzes 
(1953, as cited in Pettigrew, 1994) studied a large steel mill outside Chicago.  Inside 
the mill, Black and White workers were fully integrated in a multiracial union. Outside 
the mill these workers lived in segregated residential areas, and many of the same 
White workers who believed strongly in an interracial union belonged to a 
neighborhood organization that kept Black citizens from living in the area. Pettigrew 
cites the need for reinforcing cross racial bonds and friendships as one way to 
influence situational norms in college environments.  Pettigrew?s review reinforces the 
perspective that situations in a college environment can vary in their effect on cross 
racial interactions. 
 Powers and Ellison (1995) studied whether the contact hypothesis simply 
reflected selectivity bias, i.e., people who seek out situations in which the contact 
21
 hypothesis seems to confirm positive cross racial interactions are actually people who 
have positive internal attitudes toward people of other races.   Powers and Ellison used 
archival data from the National Survey of Black Americans conducted by the 
University of Michigan in 1980, resulting in a sample of 2,107 adults.  They found no 
evidence that the validation of the contact hypothesis was due to an artifact of 
unobserved self-selection processes. This study is of particular interest because it 
focuses on the question of whether the internal attitudes of people toward diversity 
influences their cross racial interactions.  While Powers and Ellison?s research, in this 
case, disconfirmed this phenomenon, evidence will be examined later in the literature 
review that suggests such influence is possible.   
 Wright et al. (1997), found that people do not have to have direct friendships 
with people of other races to experience an improvement in attitudes toward other 
races. The authors experimentally tested this hypothesis with 178 undergraduate 
students at a West coast university.  They found that if a person simply knows 
someone from within their own in-group who has a close relationship with someone 
from an out-group, the person?s attitudes can become more positive toward the out-
 group even though he/she has no direct interaction.  Thus it is possible for a student to 
feel vicarious positive (or negative) cross racial experiences in campus settings.  This 
is useful information, because in a number of situations on campus - e.g., classroom, 
basketball game, party - a student may not have a direct cross racial interaction but 
may observe such interaction among others.  This observation alone may be sufficient 
22
 to affect how the student would rate their experience of cross racial interaction in 
those settings. 
 On the other hand Wilder (1984) found that just one pleasant, direct contact 
with an out-group member had virtually no impact on stereotypes about the out-group 
even if the out-group member seemed typical of the out-group.  Participants were 62 
female undergraduates, 30 from one college and 32 from another college, both 
independent institutions in the Northeast, which harbored prejudiced attitudes and 
stereotypical perceptions toward one another.  The experimental design manipulated 
stereotype-disconfirming interactions between students representing each college, but 
found that this one-time interaction did not reduce stereotypes held about students in 
the other college.  This would suggest that the recurring nature of some campus 
situations (e.g., same classmates all semester, roommate all semester, co-worker) 
would more likely foster improved cross racial interaction than chance, fleeting 
encounters (e.g., conversation at a party, attending one event). 
 In summary, the contact hypothesis provides a theoretical basis, supported by 
empirical study, to suggest that the characteristics of a given situation can affect the 
nature of cross racial interactions.  Since the situations on a college campus can vary 
tremendously, from classroom settings to residential environments to parties, it seems 
plausible that the quality of cross racial interactions can also vary across these 
situations.  However, there have been no systematic attempts to measure whether and 
to what degree the quality of such interactions might vary.  
23
  Environmental assessment. Environmental assessment research has confirmed 
that different settings can have distinctive characteristics (Moos, 1973). The 
characteristics of these settings are governed in part by the behaviors and 
characteristics of the people in these settings  (Moos, 1973).  It has long been 
recognized that people can behave toward people of other races/ethnicities in different 
ways depending on the situation and context (Smith, 1993).  These findings point to 
the discrete, unique nature of different settings in an environment.  At the same time 
there is evidence that the experiences an individual has in one situation can influence 
their mood and behavior in another situation (Moos, 1996). This suggests that the 
quality of cross racial interactions may co-vary for certain situations, e.g., the 
experience in the classroom may influence the experience in similar situations such as 
labs sections or study groups. 
Attitudes toward diversity and cross racial interactions
 The review so far has focused on the potential ways that situations can 
influence the positive or negative nature of cross racial interactions.  There is a body 
of research that suggests that our internal attitudes toward people who are different 
may influence our interactions (behaviors) with those people as well. 
 Link between attitude and behavior. In general attitudes can predict many 
forms of social behavior across a wide range of contexts (Baron & Byrne, 1994).  
However, the strength of the link between attitude and behavior depends on several 
factors: is the attitude specific (stronger link) or general (Newcomb, Rabow, & 
24
 Hernandez, 1992); and to what extent does an individual have a vested interest in the 
subject matter of the attitude.  For example, students of color are more concerned 
about cross racial interactions and race relations, more so than White students (on 
campuses where White students are the majority) because they see it more directly 
impacting their lives (Fisher & Hartmann, 1995).  This might suggest that there will be 
stronger associations between attitudes toward cross racial interactions (behaviors) for 
students of color than for White students on majority White campuses. 
 Effect of prejudice on cross-racial interactions. There is evidence that if a 
person holds negative attitudes toward another group (prejudice), that person?s social 
interactions with people from that other group can be negatively impacted.  Harris et 
al. (1992) studied this effect in pairs of boys in third through sixth grade that were 
about to participate in a task together.   One of the boys in each pair was secretly 
informed (randomly) that the other boy had disruptive social skills and got into trouble 
a lot.  Boys that were told this  later reported that they had a more negative interaction 
with their partner (enjoyed the task less, liked their partners less and were less friendly 
toward their partners) than boys that were not told this about their partner.  In this case 
the experimentally induced prejudice led to a phenomenon sometimes referred to as 
the self-fulfilling prophecy (Baron & Byrne, 1994).  This phenomenon suggests that 
the positive or negative nature of cross racial interactions between college students 
may be impacted by the attitudes of the students toward each other?s group (e.g., 
racial, ethnic, religious, etc..). 
25
  Prejudice is an attitude (usually negative) held toward a specific group (Baron 
& Byrne, 1994).  It would be difficult to measure the amount of prejudice a person 
might harbor toward all of the possible racial targets of this prejudice (e.g., African 
American, White, etc..).  Another approach would be to examine the degree to which 
the person is generally more open toward accepting and appreciating people and 
cultures that are different than one?s own.  Miville et al. (1999) found that having an 
open and appreciating attitude toward diversity represented by all other persons (a 
universe-diverse orientation) was associated with lower levels of prejudice (e.g., 
homophobia and dogmatism) and higher levels of empathy.   
 Thus, if the quality of cross-racial interactions is to be examined, the possible 
impact of attitudes on these interactions ought to be examined as well.  Instead of 
trying to assess the specific prejudice or attitude each student has toward a student of 
another race, another approach would be to measure the general openness and 
appreciation for diversity that a student exhibits and determine if it has any influence 
on that student?s cross racial interactions.  There is some evidence to suggest that the 
association between attitude toward diversity and quality of cross racial interactions 
(behaviors) may be stronger for students of color, particularly Black students, because 
race is a more salient aspect of their lives. 
 In summary, if the positive or negative nature of cross racial interactions is 
going to be explored,  in addition to the situation in which the interaction takes place, 
there is the real possibility that the attitude of the individual may effect this 
26
 interaction.  No studies could be found that have attempted to measure the extent to 
which these  interactions are influenced by both the external situation and the internal 
attitudes of people. 
 Summary. Campus climate studies provide the general context within which 
to better understand cross racial interactions in a college campus environment.  Are 
the positive or negative nature of these interactions uniform across the variety of 
campus settings or are some situations better than others?  Some campus climate 
studies recognize that cross racial interactions may vary in different settings, but none 
explore this concept in a way that would enable comparison between settings on 
campus, for example, classrooms versus residence halls.  
 While all students seem to acknowledge the challenges for improving these 
interactions, studies suggest that students of color are experiencing the climate and 
cross racial interactions more negatively than their White counterparts. Is this uniform 
across all settings? Or would a closer examination show that in certain situations 
students of color are more comfortable in their relationships with White students than 
in other situations? 
 The contact hypothesis offers a theoretical basis to explain why the positive or 
negative nature of cross racial interactions may differ among settings.  There is also 
sufficient research to suggest that our internal attitudes can influence the quality of our 
cross racial interactions.    
 
27
  So to what extent can we better understand how the positive or negative nature 
of cross racial interactions varies among the different settings on a college campus 
and, at the same time, assess the extent these interactions are influenced by the internal 
attitudes of people toward diversity?  The current study will focus on these questions. 
Statement of Problem
 Universities provide an excellent opportunity to enhance positive cross racial 
interaction (American Commitments National Panel, 1995).  While students express a 
desire for better interactions, creating such an environment remains a challenge 
(Humphreys, 1998).   
 Campus climate studies have been helpful in understanding the general nature 
of cross racial interactions (Fisher & Hartmann, 1995).  Cotton et al. (2000) provides 
qualitative, empirical support for the idea that students perceive different and unique 
cross racial experiences across settings.  No campus climate studies have been found 
that focus on how the large variety of different situations on a college campus may 
influence such interactions.  
 In addition to the setting of cross racial interactions, there is evidence that 
peoples? predispositions toward others who are different can effect these interaction as 
well (Baron & Byrne, 1994).  For example the attitude of prejudice (usually negative) 
has been found to influence the positive or negative nature of the interaction between 
the person with the prejudice and a person who is the target of the prejudice (Harris et 
al., 1992). 
28
  To what extent does the link between attitude and behavior relate to the quality 
of cross racial interactions?  Campus climate studies have not examined the internal 
attitudes of students toward diversity as a variable related to the outcome of cross 
racial interactions.   
 The purpose of this study is to examine the different situations across a campus 
environment, the internal attitudes of students and their relationship with the positive 
or negative nature of interactions between different race students.  Therefore, the 
present study will test the following hypotheses:   
Hypothesis 1: Students will cluster in identifiable ways related to the similarity of their 
CRIS scores.
 It is clear that the college campus consists of a variety of different 
situations. The contact hypothesis provides a theoretical basis to suggest that 
different situations might influence cross racial interactions in different ways (Allport, 
1954; Brewer & Brown, 1998; Pettigrew,1994). Various studies of the contact 
hypothesis in and outside of college environments help to support the idea that cross 
racial interactions will vary across campus situations (Nesdale & Todd, 1998).  
Therefore, it is logical to hypothesize that the different situations on a college campus 
will manifest different characteristics, and this will influence cross racial interactions. 
 There is evidence that certain situations may foster the opportunity for more 
positive cross racial interactions. For example, situations with recurring contact, such 
as classrooms, versus one-time, such as parties, may be more likely to lead to a higher 
quality of cross racial interaction (Wilder, 1984). Similarly, structured situations such 
as classrooms or work places may provide more opportunities for contact hypothesis 
29
 criteria to come into play (Campus Assessment Working Group, 2003; Cotton et al., 
2000).  There is also evidence that the racial characteristics of students is related to 
their cross racial experiences (Fisher & Hartmann, 1995).  Therefore, hypothesis 1 will 
analyze both the quality of cross racial interactions and the demographic 
characteristics of students to identify identifiable clusters of students and/or situations. 
Hypothesis 2: The quality of cross-racial interactions for Black/African American, 
Asian Pacific American and Latino students will be lower than for White students.
 There is substantial research to support the idea that African American students 
are experiencing the campus environment more negatively than their White 
counterparts  (Ancis, Sedlacek, and Mohr, 1998; Fisher & Hartmann, 1995; 
McClelland & Auster, 1990; Student Affairs Research Services, 1994).  
Hypothesis 3: There will be a positive correlation between the perceived quality of 
cross-racial interactions and internal attitudes toward diversity.
 This correlation will reflect the positive relationship between a person?s 
internal disposition toward appreciating differences and similarities with other people 
and their level of positive experience with different race students irrespective of 
differing situations.  The research in ths area is equivocal.  On the one hand there is a 
clear link between attitudes and social behavior (Baron & Byrne, 1994).  On the other 
hand Powers and Ellison (1995) found no basis to believe that people with positive 
attitudes toward diversity sought out situations in which their experience would 
confirm their attitudes.  However, there appears a sufficient basis to hypothesize that 
attitudes toward diversity will have some affect on the interaction. 
 
30
 Additional Analyses
 Two research questions are posited that are intended to examine how gender 
and cohort relate to the cross racial experiences and how cross racial interactions differ 
from interactions in general (i.e., not specific to race).  
 The first research question relates to better understanding how gender and 
cohort relate to cross racial experiences. Are students of different genders and cohorts 
having different experiences across the settings?  Does gender interact with these 
experiences?  Are seniors, who have been on the campus for several years having 
cross racial interactions different than freshman who have just started?  
Research Question 1: Do the quality of cross racial interactions  vary by gender or 
cohort?
 Research into gender differences found that women were more likely to have 
more liberal racial attitudes then men (Schuman, Steen, Bobo, & Krysan, 1997; 
Springer, Palmer, Terenzini, Pascarella, & Nora, 1996).  The cohort, or class standing 
is an indicator of how long the student has been exposed to the university environment 
and presumably influenced by that environment.  This can vary between one and five 
years for the vast majority of undergraduate students.  There are several studies which 
suggest that residential students (living on campus) versus commuter students (living 
off-campus) are having different cross racial experiences (Campus Assessment 
Working Group, 2003; Cotton et al., 2000; Korgen et al., 2003; Nesdale & Todd, 
1998) . 
 The second research question relates to how cross racial interactions might 
differ from interactions in general (i.e., not specific to race).  It is difficult to assess the 
31
 meaning of positive or negative cross racial interactions in a given situation without 
comparing it to a norm or benchmark.  What are the quality of interactions in the same 
situation without specifically focusing on race?  One of the weaknesses with campus 
climate studies is that they don?t assess perceptions of the campus environment in 
relation to a broader context.  For example, students of color may report they are 
dissatisfied with the state of race relations on a campus, but they may also indicate that 
relations on campus are better than in society generally. While not eliminating the 
concern, it puts the concern in a context.  This can only be determined by establishing 
norms. 
 The current study attempts to determine benchmarks or norms related to the 
positive or negative nature of student interactions in situations across campus without 
focusing on race.  Against these benchmarks then, the positive or negative cross racial 
interactions can be compared for the same situations.  
Research Question 2: How does the quality of cross racial interactions between 
students differ from the quality of general interactions between students, without 
reference to race?
 The meaning of ratings or scores is enhanced when they can be viewed in 
relation to comparable ratings for an appropriate reference group (Hopkins, 1998).  It 
is important to understand that a norm is not necessarily a goal, but a measure of what 
is. 
 Summary. The current study extends the Cotton et al. (2000) study which 
explored the quality of cross racial interactions across the campus environment.  
32
 Extension of the Cotton et al. study enables greater generalization of its findings 
concerning the relationship of the different settings to the quality of the interactions.  
By adding a measure regarding a person?s attitude toward diversity the current study 
will extend the previous work by examining the interaction between environment and 
internal disposition in relation to a student?s cross racial interactions.  
 
33
  Chapter 3 
 Method 
Design Statement
 This descriptive, correlational study will examine the quality of cross racial 
interactions in 13 situations in a college campus environment and internal attitudes 
toward diversity.  The study will use the Miville-Guzman Universality-Diversity 
Scale-Short (MGUDS-S), excerpts from the Cultural Attitudes and Climate 
Questionnaire (CACQ), and the author developed Cross Racial Interaction Scale 
(CRIS). 
Participants
 A stratified, random sample of 1,000 participants were randomly selected from 
the Fall, 2002 enrolled undergraduate population of the University of Maryland, 
College Park.  The sample was stratified by the four major racial/ethnic categories 
represented at the University of Maryland - White, African American/Black, Asian 
Pacific American, and Hispanic/Latino, with 250 students randomly selected from 
each category.    
 The Office of Institutional Research and Planning reported these as the four 
largest undergraduate student racial/ethnic categories at the University of Maryland in 
the previous year (Fall, 2001) as follows: White (59.4%), African American/Black 
(13.1%), Asian Pacific American (13.8%), Hispanic/Latino (5.1%).  Other categories 
that were not used in the current study included: International (2.9%), and Unknown 
(5.5%).  The American Indian population at the University of Maryland was recorded 
34
 at 0.3% and was not formally analyzed in this study due to the inability to draw a 
large enough sample. However, the Office of Institutional Research and Planning was 
able to provide the names of 25 American Indian students, so these were included in 
the sample, for a total of 1,025 participants.  Each student in the sample was randomly 
assigned to receive either version A (experiences with other races) or version B (no 
reference to race) of the questionnaire.   
Measures
 The Miville-Guzman Universality-Diversity Scale-Short (M-GUDS-S) 
(Fuertes, Miville, Mohr, Sedlacek & Gretchen, 2000) measures the construct 
Universal-Diverse Orientation (UDO).  Miville et al. (1999) defined UDO as an 
attitude toward other persons in that similarities and differences are both recognized 
and accepted.  The shared experience of being human results in a sense of connection 
with people and is associated with a diversity of interactions with others.          
 The M-GUDS-S is a 15 item questionnaire with items that are rated on a 6-
 point Likert-type scale ranging from strongly disagree (1) to strongly agree (6) (See 
Appendix D).  The M-GUDS-S yields a total score as well as scores from three 5-item 
subscales.  These subscales are: (a) relativistic appreciation of oneself and others, 
which involves recognition and acceptance of the similarities and differences among 
people, (b) diversity of contact, which assesses both previous and future intended 
behaviors relevant to interpersonal contact with people of different demographic 
backgrounds; and (c) sense of connection, involving the emotional bond one feels 
toward others.  These subscales reflect the cognitive, behavioral, and affective 
35
 components of UDO, respectively (Fuertes et al., 2000). 
 The M-GUDS-S is a short form of the original version, M-GUDS, which 
consists of 45 items.  Administration of the original version to four separate samples 
resulted in internal consistency and retest-reliability ranging from .89 to .95.  The 
measure significantly correlated in theoretically predicted ways with measures of 
racial identity, empathy, healthy narcissism, feminism, androgyny, homophobia, and 
dogmatism. The original version displayed discriminant validity by not correlating 
with SAT verbal scores; however, mixed results were obtained with social desirability 
(Miville et al., 1999).  
 The short form M-GUDS-S was derived by selecting the five highest loading 
items on each of the three factors of the original scale.  The correlation between the 
total score of the short form and the long form was found to be .77 (p<.001) (Fuertes, 
et al. 2000).  The short form correlated in theoretically predicted directions with items 
from a new student orientation survey  related to religious tolerance, attitudes toward 
gay and lesbian persons, and having close friends of another race (Fuertes, et al. 
2000). 
 The Cross Racial Interaction Scale (CRIS) is a new instrument developed by 
the author to measure the extent students report positive or negative experiences with 
students of races other than their own in situations across the campus environment.  
The scale includes 13 situations grouped into four major settings: (a) academic setting 
- classrooms, small group/lab discussion sections, study groups and group projects 
outside the classroom; (b) campus employment setting - on campus and off-campus 
36
 jobs; (c) residential life setting - planned programs as well as informal interaction and 
friendships; and (d) social setting - broad category that includes student groups, 
parties/social events; sports & recreation, fraternities and sororities, large campus 
events (e.g., first look fair, basketball game) and off-campus socializing stemming 
from on-campus relationships.   The CRIS uses a 6 point Likert-type scale from 
1=very negative to 5=very positive, and includes a not applicable option (0) for each 
situation to indicate that the respondent had no experience with students of other races 
in that situation. 
 The CRIS was drawn from the results of the qualitative study by Cotton et al. 
(2000) in which 75 semi-structured interviews of undergraduate students, stratified by 
race and randomly selected at a large, public university, were used to gain student 
input about the situations in which they have had positive and negative experiences 
with same race and different race students.  The purpose of this qualitative study was 
to (1) better understand the situations in which students have positive and negative 
racial experiences, and (2) attempt to uncover common characteristics of those 
situations to guide strategies to enhance positive contact.  The study used the critical 
incidents technique (Flanagan, 1954), a strategy of interviewing to gather the critical 
characteristics of a situation or behavior.  An inductive approach was used to analyze 
the responses and to develop a classification structure (Cotton et al., 2000).  Multiple 
raters were used to enable assessment of inter-rater reliability.  Analysis of the 
responses led to the identification of four major settings where experiences were 
taking place: Academic, campus employment, residential and social.  Within each of 
37
 these major settings, specific situations were repeatedly mentioned.  The CRIS 
settings and situations were drawn directly from the outcome of this earlier study.  
 Two versions of the CRIS were administered, each to one-half of the sample. 
Version A (see Appendix A) asked about experiences with students of other races 
[emphasis added] in different situations on campus (e.g., classrooms, residence halls).  
Version B (see Appendix B) asked the same question but left out the reference to 
other races.  Version B served as a control.  This allowed comparisons, by race, 
between the control group and the test group to establish benchmarks, i.e., norms 
against which to compare experiences across races.  This methodology is drawn from 
the Situational Attitude Scale (SAS) originally developed by Sedlacek and Brooks 
(1970).  Sedlacek (1996) elaborates on the method which is to use experimental and 
control forms of a questionnaire in order to assess prejudice without social desirability 
bias. 
 The CRIS was pilot tested with a small convenience sample of undergraduate 
students in order to check face validity and understandability.  It was also reviewed 
with a researcher widely published in the areas of race, race relations, and campus 
climate and who was involved in the qualitative study (Cotton et al., 2000).  He 
confirmed the CRIS to be an adequate reflection of the results of the qualitative study.  
 The Cultural Attitudes and Climate Questionnaire (CACQ) (Helm, Sedlacek, 
& Prieto, 1998) is a 100 item question designed to measure students? perceptions and 
experiences of the university racial and ethnic climate (see Appendix C). Using a 4-
 point and 5-point Likert-type scale, 560 first and third year college students reported 
38
 their level of agreement with statements regarding the campus climate (Ancis, 
Sedlacek, & Mohr, 1998). The overall coefficient alpha reliability of the questionnaire 
was .81 (Helm, Sedlacek, & Prieto, 1998).  Eleven factors were identified using 
principal axis factor analysis and varimax rotation, accounting for 48% of the total 
variance.  The factors were labeled Racial Tension ( r=.73); Cross-Cultural Comfort ( 
r=.73); Diversity Awareness ( r=.67); Racial Pressures ( r=.60); Residence Hall 
Tension ( r=.69); Fair Treatment ( r=.74); Faculty Racism ( r=.77); Respect for Other 
Cultures ( r=.62); Lack of Support ( r=.63); Comfort With Own Culture ( r=.55); 
Overall Satisfaction ( r=.78).   
 The CACQ is one of very few campus climate surveys for which psychometric 
characteristics are available.  Eight items were selected from the CACQ by the author 
that seemed most likely to correlate in theoretically predictable ways with the CRIS. 
While the entire CACQ would be valuable to administer, it would more than  triple 
the length of the current survey (from 29 items to 129 items, plus demographics).  
Consultation with researchers in the campus Office of Institutional Research and 
Planning (responsible for administering numerous surveys) strongly suggested that 
increasing the length would further exacerbate what have become smaller and smaller 
response rates in recent years among general student mailings.   
The CACQ was reviewed by the author to identify the items that should have a 
strong relationship with the CRIS.  A student?s average CRIS score across all 
situations on campus should reflect their overall experience, and correlate to three 
CACQ questions that measure overall campus climate: ?There is racial conflict on 
39
 campus?, ?There is racial/ethnic separation on campus? and ?There are a great deal of 
friendships between students of different racial and ethnic groups?.  
 Other CACQ items were selected that related to specific situations, for 
example, ?There is interracial tension in the residence halls? should correlate with the 
CRIS scores about experience in the residence halls. Other CACQ items related to the 
classroom situation were selected that should correlate with CRIS scores specific to 
the classroom. 
 Open Ended Question. One open ended question was asked: ?Describe an 
activity or environment on campus where you have had positive relationships with 
students who were different than you (e.g., race, ethnicity, religion, sexual 
orientation...)?.  This question was intended to obtain qualitative to better understand 
the situations on campus where positive relationships across racial lines are occurring.  
Ultimately, the goal is to foster environments that will increase such experiences.   
The open ended question also reflects an extension of the qualitative information 
gathered by Cotton, et al. (2000).  The results of this question were not used in the 
current study. These data will be analyzed in a future study. 
 Demographics. Gender, race/ethnicity, and cohort were primary sorting keys. 
Cohort was included because of the known developmental changes associated with 
the college years. Also, living on or off-campus, employment on or off-campus, part 
or full-time status and current college (major) were obtained for possible significance 
in examining differences in the experiences with students of other races (see 
Appendix E). 
40
 Procedures
 The data that was used for this study was archival data, having been collected 
as  part of a funded research grant project in which the author was a co-principal 
investigator.   
 The following process was used to collect the data.  The sample of 1,000 
students was randomly selected from the Fall, 2002 undergraduate population. 
Twenty five Indian American students were then added for a total of 1,025 students.  
Each student was then randomly assigned to receive either version A or version B of 
the questionnaire.  Version A asked about experiences with students of other races 
[emphasis added] in different situations on campus (e.g., classrooms, residence halls).  
Version B asked the same question but left out the reference to other races.  Version B 
was a control.  All statements concerning anonymity, voluntary participation and 
freedom to stop at anytime were the exact same for both versions A and B.   All of the 
procedures below were the exact same for respondents whether they received version 
A or B.  
 After Institutional Research Board approval was obtained, an email was sent 
to each of the selected students inviting them to participate in the study (see 
Appendices F and G).  They could then click on a web URL to complete a 
questionnaire if they wished. 
 Version A of the questionnaire included the following sections (a) what has 
been their experience with students of other races in different settings on campus 
(CRIS, 13 items), (b) their attitude towards diversity issues  
41
 (M-GUDS-S, 15 items), ( c) their opinion about the racial climate on campus (CACQ 
excerpted, 8 items), (d) open-ended question about where they have had positive 
experiences on campus, and (e) demographic information.  Version B, since it 
eliminated the reference to race, include sections (a) what has been their experience 
with other students in different settings on campus (eliminated reference to race) 
(CRIS-non race version, 13 items), and (d) and (e) which remained the same. 
 A $10 (for version A) and $5 (for version B) gift certificate to the University 
Book Center were offered as an incentive to complete the questionnaire.  In order to 
protect anonymity, after completing a questionnaire the respondents were sent to a 
web site separate from the questionnaire web site and could either enter a mailing 
address for the gift certificate or read instructions on how to pick up their certificate 
on campus.  
 Ten days following the initial email, a followup email was sent to the entire 
sample (since responses were anonymous) asking them to complete the survey if they 
had not already done so.  Five days following the second email, the web site receiving 
surveys was closed.   
 A total of 381 surveys were completed (37.2% response rate).   One response 
was removed since it included no responses to the instruments in the questionnaire.  
Of the 380 useable responses, 195 completed version A (reference to experiences with 
other races) and 185 completed version B (no reference to race).      
 A set of steps were taken in order to obtain the best response rate possible.  A  
significant incentive was provided (every respondent was guaranteed a $10 or $5 gift 
42
 certificate to the university book store), a short survey was constructed (10 minutes to 
complete), an initial invitation email was reviewed first with students to maximize its 
potential to persuade students to complete the questionnaire, and a reminder followup 
was sent approximately 2 weeks after the initial email..  A third reminder was not 
possible since it was the end of the semester. 
 
43
  Chapter 4 
 Results 
 This section is divided into preliminary analysis, hypothesis and research 
question analysis, and additional analysis.  The preliminary analysis describes the 
results of the surveyed sample in terms of demographics and comparison of the 
respondents to the version A of the questionnaire which referenced race, and version 
B of the questionnaire which did not.  
Preliminary analysis
 Respondents. Three demographic profiles were examined.  First, the 
demographics of the respondents representing the entire sample themselves, second, a 
comparison of the respondents who took version A versus those who took version B 
of the questionnaire and third, a comparison of the entire sample (respondents) to the 
mailing list (all students sent the questionnaire).  Version A included the reference to 
race in the Cross Racial Interaction Scale, CRIS, while version B did not include the 
reference.  These two versions will be referenced as version A (race) and version B 
(non-race).     Table 1 compares the demographic characteristics of these different 
samples including race, gender, cohort (e.g., junior, senior), housing (living on 
campus/off campus),  and employment.  
 Comparison of version A and version B respondents. Out of 1,025 students 
sent the questionnaire, there were 381 completed questionnaires.  One response  had 
no data and was eliminated, leaving 380 useable responses, or a 37 percent response 
rate.  Of the 380 valid responses, 195 responded to version A (race) and 185 
44
 responded to version B (non race). These are comparable response rates, with version 
A having a 5.4% higher return rate than version B. Comparing the demographic 
characteristics of the respondents to the two versions exhibits some differences 
(adjusted for the 5.4% difference, which would be an expected difference between 
Version B to version A).  For example, the response rate increased substantially for 
students of color from version B to version A, while the response rate for White 
students decreased. The Black, APA and Latino student respondents increased from 
90 to 118, or 25.7% (adjusted by the 5.4% expected increase), while White student 
respondents decreased from 59 to 56, or  -10.5% (adjusted).  Therefore, students of 
color were 36.2% more likely to respond to a questionnaire about race relations versus 
a questionnaire about student relations without reference to race than White students.  
 Non-response bias. Table 1 shows the percentage breakdown of demographic 
characteristics for all of  the students sent the questionnaire as well as those for the 
respondents.  There are several significant differences in these characteristics.  
 Cross tabs and chi square analyses were run with race, gender, cohort and 
housing as criteria.  Significant differences were found for Gender,  X2 (1, N = 1280) 
= 3.91, p = .048. Significant differences were found for Race (Black, APA, Latino, 
and White)  X2 (3, N = 1280) = 15.42, p = .001.  No significant differences were 
found for Cohort (First year students, Sophomore, Junior and Senior)  X2 (3, N = 
1280) = 2.429, p = .488.   Significant differences were found for Housing  (living on-
 campus or off-campus)  X2 (1, N = 1280) = 13.75, p < .000.   
 The previous findings suggest significant non-response bias, however, several 
45
 factors might explain some of the difference.  For example, 13.9% of the respondent 
pool self-identified as either multirace or unknown.    Adding the multirace students 
proportionately back into the original race categories would change their percentages.  
For example the number of multirace students self-identifying as part Latino would 
have increased the Latino count by 15.5%.  
 In comparing Gender, 11.8% of the respondent pool elected to not identify 
gender. Thus, while the percentage of females was 52.9% for the respondent pool and 
52.1% for the original sample, males were 35.3% of the respondent pool while 
comprising 47.9% percent of all of the students sent the questionnaire.  What is 
unclear, is to what extent the 11.8% of respondents who did not report gender are 
causing the non-response bias conclusion for gender. 
 In conclusion, even though there may be reasons other than non-response bias 
that can explain the differences between the response pool and all of the students sent 
the questionnaire, based on the differences in the  Chi square analyses, it is likely that 
this sample is not representative of the population of all undergraduate students at the 
University of Maryland during the Fall, 2002 semester.  
 Use of selective sample sets. Table 1 includes all categories of respondents. 
Some of the analyses in this study exclude respondents depending on the intent.  For 
example, in terms of race, this study is limited to investigating the four major race 
categories at the University of Maryland - Black, APA (Asian Pacific American), 
Latino, and White.  Therefore, the categories of Multirace and American Indian will 
be excluded from some analyses.  The category of Not reported may be excluded for 
46
 Race, Gender and Cohort. 
 Most of the analyses also focus on the respondents to version A (race) since 
this measures cross racial interaction.  Therefore, most of the analyses will be limited 
to these 195 respondents. 
 Cross Racial Interaction Scale. The CRIS measures the quality of cross racial 
experiences in 13 different situations on campus.   Table 2 provides descriptive data 
on the results of the CRIS scores, including the N, M and SD for each situation, as 
well as Pearson?s product moment correlation matrix.  The mean CRIS score varied 
from a high of 3.99 (SD = .81) for classrooms, to a low of 1.17 (SD = 1.74) for Greek 
life.   
 The CRIS uses a 6 point Likert-type scale of 0 = No Experience, 1 = Very 
Negative, 2 = Negative, 3 = Neutral, 4 = Positive and 5 = Very Positive. Some 
analyses will exclude the No Experience selection in order to more accurately reflect 
the quality of the cross racial interaction.  
 
47
 Table 1.  Demographic Characteristics of Sample 
 
Respondents  
Version A 
(race) 
Version B 
(non-race) 
Total 
Resp?dents 
 All Students 
 Sent  
Survey 
N % N % N % N %
 Race Black 37 19.0 30 16.2 67 17.6 250 24.4
 APA 50 25.6 33 17.8 83 21.8 250 24.4
 Latino 31 15.9 27 14.6 58 15.3 250 24.4
 Am. Indian 0 0 4 2.2 4 1.1 25 2.4
 White 56 28.7 59 31.9 115 30.3 250 24.4
 Multirace 10 5.1 16 8.6 26 6.8 - -
 Not rep. 11 5.6 16 8.6 27 7.1 - -
 Gender Female 104 53.3 97 52.4 201 52.9 534 52.1
 Male 69 35.4 65 35.1 134 35.3 491 47.9
 Not rep. 22 11.3 23 12.4 45 11.8 - -
 Cohort 1st year 35 17.9 46 24.9 81 21.3 236 23.0
 Soph 50 25.6 31 16.8 81 21.3 246 24.0
 Junior 43 22.1 53 28.6 96 25.3 271 26.4
 Senior 65 33.3 51 27.6 116 30.5 272 26.5
 Not rep. 2 1.0 4 2.2 6 1.6 - -
 H?sing On-campus 85 43.6 103 55.7 188 49.5 388 37.9
 Off-campus 110 56.4 82 44.3 192 50.5 637 62.1
 Job On-campus 38 19.5 43 23.2 81 21.3
 Off-campus 70 35.9 58 31.4 128 33.7
 Both 11 5.6 8 4.3 19 5.0
 No job 75 38.5 75 40.5 150 39.5 
Totals  195 100% 185 100% 380 1,025
48
  
Table 2 
Pearson Correlation Coefficients for CRIS Scores 
 
N M SD 
class 
rm 
lab/ 
disc- 
uss. 
study 
group 
on  
camp. 
job 
off  
camp. 
job 
classrm 194 3.99 .81 1.00     
lab/discuss. 195 3.81 1.09 .45** 1.00    
study groups 193 3.69 1.26 .40** .25** 1.00   
on camp. job  192 1.83 1.98 .01 -.03 .11 1.00  
off camp. job 193 2.71 1.97 .00 .11 .17* .07 1.00 
RL programs 195 2.33 1.90 .11 .19** .05 .13 -.09 
RL social 195 2.89 1.90 .06 .17* .07 .13 -.12 
student grps 193 3.22 1.73 .12 .27** .20** .12 .04 
parties 195 3.22 1.62 .26** .26** .19** -.02 .09 
sports/rec. 195 2.68 1.95 .13 .19** .19** .21** .07 
Greek life 194 1.17 1.74 -.04 -.05   .05 .17* .06 
camp. events 195 3.50 1.46 .21** .24** .13 -.03 -.08 
off-campus 194 2.98 1.90 .18* .03 .25** .03 .17* 
** correlations significant at the 0.01 level (2-tailed) 
 
* correlations significant at the 0.05 level (2-tailed) 
 
Note. Includes ?no experience? CRIS scores. 
49
 Table 2 (Continued) 
 
Pearson Correlation Coefficients for CRIS Scores 
 
RL 
prog. 
RL 
social 
stud?t 
grps 
parties sports/ 
recr. 
Greek 
life 
camp. 
events 
off 
campus 
classrm         
lab/discuss.         
study groups         
on camp. job          
off camp. job         
RL programs 1.00        
RL social .68** 1.00       
student groups .18* .09 1.00      
Parties .12 .18* .15* 1.00     
sports/rec. .22** .13 .24** .18*  1.00    
Greek life .24* .13 .18* .22** .16* 1.00   
camp. events .28** .23** .26** .24** .21** .17* 1.00  
off-campus .12 -.01 .20** .25** .19** .13 .28** 1.00 
** correlations significant at the 0.01 level (2-tailed) 
 
* correlations significant at the 0.05 level (2-tailed) 
Note. Includes ?no experience? CRIS scores. 
50
  
Hypothesis 1: Students will cluster in identifiable ways related to the similarity of their 
CRIS scores.
 This hypothesis was tested using a hierarchical cluster analysis.  Cluster 
analysis is generally used to identify homogeneous subtypes within a complex data set 
and typically, there is no a priori knowledge about natural groupings (Borgen & 
Barnett, 1987) . The analysis clustered cases (students) based on the degree of 
similarity between their CRIS scores.  The Ward?s minimum variance method (Ward, 
1963) was used as the clustering technique, with the proximity matrix calculated using 
the Squared Euclidean Distance measure.   Ward?s clustering method has been widely 
used in behavioral science research and has been found to be one of the more effective 
methods for recovering underlying structure (Borgen & Barnett).  Out of the 195 
respondents to the CRIS Version A (race), 185 respondents were valid cases for the 
hierarchical cluster analysis. Ten cases were not valid because at least one score among 
the 13 CRIS variables was missing (i.e., the respondents did not fill in a CRIS score).  
 Based on visual inspection of the resulting dendogram, a seven cluster solution 
was chosen. Choosing a smaller number of clusters would have increased the in-cluster 
variance or dissimilarity (because the number of students in each cluster must increase).  
This would decrease the ability to describe unifying characteristics of the clusters. 
 Clusters and CRIS scores. Table 3 shows the number of students N in each 
cluster, with means and standard deviations for their CRIS (cross racial interaction 
scale) scores.  It also displays the mean CRIS score for each cluster for  the 13 
51
 situations.   The clusters range from a high of 44 students (cluster 5)  to a low of 14 
(cluster 6).   Mean CRIS scores ranges from a high of 4.20 (SD = .63) for cluster 2 to a 
low of 3.47 (SD = .56) for cluster 7.  A one-factor, fixed effect ANOVA indicated 
significant differences between the cluster mean CRIS scores F(6,178) = 2.45, p = .027.  
Tukey?s HSD post hoc test for pairwise comparisons found significant differences 
between cluster 7 and cluster 1 (M difference = -.58, p = .043), and cluster 7 and cluster 
2 (M difference = -.74, p = .008).   The unweighted mean of the cluster CRIS scores is 
shown in the right hand column.   
 Clusters and demographics. Crosstabs were run with the seven clusters of 
students and the variables of race, gender, living on/off campus, cohort, and 
employment.  Table 4 displays this information.   The percentage of each characteristic 
(e.g., race, gender) is displayed within each cluster of students, as well as the 
percentage of that characteristic for the total valid sample of 185 cases (far right 
column).  Table 4 excludes selective demographic sub-categories in order to simplify 
the table.  This means the percentages reflected do not sum to 100%.  For example, the 
Race category excludes Multirace, American Indian, and No report.  If these 
subcategories were added back to each cluster, Race would sum to 100%.  Similarly, 
the No-report subcategory count is not shown for Gender, Cohort, Housing and Job. 
CRIS and demographic characteristics of the seven clusters of students.
 The following section analyzes the seven clusters in terms of the demographic 
profile of the students (Table 4), and characteristics of their CRIS scores (Table 3).   
 Two criteria were established to identify distinguishing characteristics of the 
52
 clusters. Using Table 4, the demographic percentages within a cluster were compared 
with expected percentages for the entire sample as shown in the right hand column.  
Wherever there was a difference of greater than 20 percent between the expected 
percentage and the cluster percentage for a given characteristic this was identified as a 
distinguishing characteristic of that cluster and is shown in bold in Table 4.   Twenty 
percent was chosen after examining the size of variances between the cluster 
demographic percentages and selected as a reasonable cut-off.    
 Using Table 3, the CRIS scores were compared with the unweighted mean of 
clusters (CRIS) scores shown in the right hand column for each situation.  Wherever 
there was a difference of greater than 1.0 full point between the cluster CRIS score and 
the unweighted mean CRIS this was identified as a distinguishing characteristic of that 
cluster and is shown in bold in Table 3.  One full point was chosen after examining the 
size of variances between the cluster CRIS scores and selected as reasonable cut-off. 
 The seven clusters were analyzed using this operational definition of 
distinguishing characteristics.   Each cluster is discussed below, with demographic 
percentages shown in parenthesis (cluster percent/expected percent).  CRIS scores are 
also shown in parentheses. 
 Cluster 1 (26 students). This group is overwhelmingly living off campus 
(96.2%/54.6%) and working off-campus (65.4%/34%).  They are predominantly 
upperclass students (juniors and seniors, 88.5%/53.5%) and include more minority 
students (Black, APA & Latino, 80.8%/59.5%).   The lowest CRIS score is RL 
(resident life) programs (.15). This latter score is indicative of the off-campus living 
53
 characteristic, with almost all respondents in this cluster indicating ?no experience? for 
RL programs. This, combined with the low score for RL social (.27) and high scores for 
off-campus job (3.85) and off-campus activities (3.81) are consistent with the off-
 campus characteristics of this group. 
 Cluster 2 (18 students). This group is reflective of White (50.0%/29.7%), male 
(61.1%/34.1%), first year and sophomore students (66.6%/45.4%) and not employed 
(72.2%/39.5%).  There are no APA students (0%/24.9%). They have relatively high 
CRIS scores, compared to the average, for RL social (4.22), RL programs (4.33), 
parties (4.17), sports/recreation (4.17), and Greek life (2.17).      
 Cluster 3 (38 students). The only distinguishing characteristic of this racially 
and gender representative group is that it lives more off-campus (78.9%/54.6%).  They 
have the lowest CRIS score for sports/recreation (.55) among the clusters and low 
scores for RL programs (.47) and RL social (1.39).  
 Cluster 4 (30 students). The primary distinguishing characteristic of this 
group is living on-campus (70.0%/44.3%).  The group is mostly  first year and 
sophomore students (66.6%/45.4%).    They have the second lowest CRIS score for 
classrooms (3.80). Interestingly, as an on-campus population, they do not work on 
campus, so their on-campus CRIS score is low (.70), while they work more off campus 
and their CRIS score for off campus work is high (3.87). 
 Cluster 5 (44 students). Demographically this group stands out in having no 
distinguishing demographic characteristic that exceeds the 20% threshold of difference.  
They have relatively high CRIS scores, exceeding 1 point higher than the average for 
54
 on-campus jobs (3.25), off-campus jobs (3.80), resident life programs (3.75), Greek life 
(2.59) and off-campus activities (3.98).  Their CRIS score exceeds the average for 12 
out of 13 situations. 
 Cluster 6 (14 students). These students primarily live on-campus 
(78.6%/44.3%), and work on-campus (64.3%/20.0%).   They have the highest CRIS 
score for on-campus jobs (4.21) and high score for resident life social (4.36).  A a low 
score for off-campus jobs; can be explained primarily by the low percentage working 
off-campus, 7.1%, and most likely checking ?no experience?.  
 Cluster 7 (15 students).    This group lives predominantly on-campus 
(73.3%/44.3%), are first year students (46.7%/18.9%) and not employed 
(60.0%/39.5%).  The are overrepresented in Black students (40.0%/18.9%) while 
underrepresented in APA students (0%/24.9%).  Their CRIS scores are low in almost 
all areas, below the average in 11 of 13 situations.  Their CRIS scores are at least 1 
point lower than average in study groups (2.07), off-campus jobs (.80), student groups 
(1.20), parties (1.13), sports/recreation 1.33), and off-campus activities (1.07).  Though 
not differing by 1 point, they also  have below average scores in the academic areas 
(classrooms, lab/discussion sections)..  
 In summary, hypothesis 1 was tested by first running a hierarchical cluster 
analysis which led to the identification of seven clusters of students based on their 
similar CRIS scores. These clusters were then analyzed using Crosstabs and Chi square 
analysis to determine distinguishing demographic characteristics, and CRIS descriptive 
statistics.   This analysis found that all the clusters exhibited distinguishing 
55
 characteristics along either demographic and/or CRIS score continua, however, some 
clusters exhibited a clearer profile of characteristics than others.  An ANOVA and post 
hoc analysis found significant difference between the mean CRIS scores for three 
clusters. These analysis support hypothesis 1.   
56
  
Table 3:  Clusters Mean CRIS Scores per Situation 
 
Clusters 
1 2 3 4 5 6 7
 Unwgt. 
Mean 
of  
Clstrs 
 
N 26 18 38 30 44 14 15  
CRIS Scores
 M 4.05 4.20 3.91 3.94 3.91 4.08 3.47  
SD .56 .63 .76 .47 .57 .44 .56  
Classrm 4.08 4.28 4.03 3.80 4.00 4.29 3.73 4.03 
lab/discuss. 3.62 4.06 3.79 4.07 3.95 4.07 3.13 3.81 
Study groups 3.92 4.00 3.50 3.87 3.95 4.29 2.07 3.66 
on campus 
job  
2.04 .61 .89 .70 3.25 4.21 1.47 1.88 
off campus 
job 
3.85 .28 2.53 3.87 3.80 .50 .80 2.23 
RL programs .15 4.33 .47 3.13 3.75 1.93 3.27 2.43 
RL social .27 4.22 1.39 3.80 4.02 4.36 3.53 3.08 
student grps 3.54 4.00 2.50 3.50 4.00 2.93 1.20 3.10 
Parties 3.35 4.17 3.13 3.27 3.57 3.50 1.13 3.16 
Sports/rec. 4.00 4.17 .55 2.63 3.43 3.29 1.33 2.77 
Greek life .69 2.17 .45 .60 2.59 .14 .00 0.95 
Campus 
events 
2.85 4.39 3.05 4.07 3.93 3.29 2.93 3.50 
off-campus 
activities 
3.81 2.89 2.71 2.57 3.98 2.07 1.07 2.73 
Note. Bold items exceed a 1.0 full point variance from the unweighted mean of clusters. 
 
57
  
Table 4. 
 
Demographic Characteristics Of the Seven Clusters 
 
Percentage w/in Clusters 
N 1 2 3 4 5 6 7
 % of
 total 
Expect 
Race
 Black 35 19.2 16.7 13.2 13.3 15.9 35.7 40.0 18.9 
APA 46 30.8 0 28.9 40.0 25.0 28.6 0 24.9 
Latino 29 30.8 16.7 23.7 13.3 9.1 0 6.7 15.7 
White 55 19.2 50.0 18.4 23.3 40.9 28.6 33.3 29.7 
Gender
 Female 100 50.0 27.8 63.2 56.7 54.5 50.0 66.7 54.1 
Male 63 38.5 61.1 28.9 23.3 34.1 42.9 20.0 34.1 
Cohort
 First year 35 0 33.3 7.9 33.3 13.6 21.4 46.7 18.9 
Soph 49 3.8 33.3 23.7 33.3 31.8 42.9 20.0 26.5 
Junior 37 30.8 5.6 26.3 13.3 20.5 14.3 20.0 20.0 
Senior 62 57.7 27.8 42.1 20.0 34.1 21.4 13.3 33.5 
Housing
 On-campus 82 0 55.6 21.1 70.0 47.7 78.6 73.3 44.3 
Off-campus 101 96.2 38.9 78.9 30.0 52.3 21.4 26.7 54.6 
Job
 On-campus 37 11.5 5.6 10.5 6.7 31.8 64.3 26.7 20.0 
Off-campus 63 65.4 16.7 47.4 30.0 29.5 7.1 13.3 34.1 
Not 
employed 
73 11.5 72.2 42.1 53.3 27.3 28.6 60.0 39.5 
Note. Bold items exceed a 20% variance from the expected percentages. 
58
  Hypothesis 2: The quality of cross-racial interactions for Black/African 
American, Asian Pacific Am. and Latino students will be lower than White students.
 A one-factor, fixed effects ANOVA was used to analyze this hypothesis. The 
four major racial groupings represented the levels of the independent factor. Each 
participant?s average CRIS score (i.e., the mean across all 13 situations) was the 
dependent variable. This is a between-subjects design.   Differences among the actual 
13 situations between the racial groups was not analyzed since this hypothesis relates to 
the overall quality of experience and not any particular situation.  
 For this analysis, the CRIS scores were limited to the 5 point Likert-type 
options of: Very Negative, Negative, Neutral, Positive, Very Positive. The 6th selection 
option of No Experience was excluded, since the purpose of this analysis is to 
determine the quality of the cross racial interactions, where the No Experience selection 
is not a measure of this quality.  Since the hypothesis was addressing the quality of 
cross racial interactions, analysis was limited to respondents to Version A (race).  
 Table 5 shows the N  for each race group, with their CRIS M, and SD.   Results 
of the ANOVA indicated significant differences F(3,170) = 4.495, p = .005.  Tukey?s 
HSD post hoc test of pairwise comparisons found significant differences within Black 
students being lower than White students (M difference = -.419, p = .005), and Black 
students being lower than Latino students (M difference = -.396, p = .031). 
 These results provide partial support for hypothesis 2.  The quality of cross-
 racial interactions for Black students was found to be lower than White students,  
however,   the quality of cross-racial interactions for Asian Pacific American and 
Latino students was not found to be significantly lower than for White students. 
59
 Table 5. 
 
Mean CRIS scores by race.  
 
N M SD 
Black 37 3.64 .67 
APA 50 3.83 .51 
Latino 31 4.03 .62 
White 56 4.06 .59 
Total 174 3.90 .61 
60
  Hypothesis 3: There will be a positive correlation between the perceived quality 
of cross-racial interactions and internal attitudes toward diversity.
 Because the relationship is hypothesized to be positive, a one tailed Pearson 
correlation was  run for the overall coefficient as well as by race, gender, cohort, living 
situation and several combinations of demographic characteristics (e.g., White Males).  
The Miville-Guzman Universality-Diversity Scale-Short (M-GUDS-S) was used to 
assess internal attitudes toward diversity, and the mean of the CRIS score (i.e., the 
mean across all 13 situations) was used to assess the perceived quality of cross-racial 
interactions.    
 As in hypothesis 2, for this analysis, the CRIS scores excluded the ?No 
Experience? option,  since the purpose of this analysis is to determine the quality (i.e., 
negative to positive) of the cross racial interactions, where the No Experience selection 
is not a measure of this quality. The analysis is limited to respondents to Version A - 
race based CRIS. 
 Results of the correlation between CRIS and M-GUDS-S scores is shown in 
Table 6.  For the total sample, the correlation is r = .26, p < .01 level.  There are large 
differences between subsets of the sample.  For Race the largest difference was 
between White students ( r = .55, p < .01) and APA students ( r = -.01, not significant, 
ns).  For Gender, the correlation for Male students ( r = .43, p < .01) was much higher 
than Female students ( r = .17, p < .05).   Even for Housing, the correlation for students 
living off-campus ( r = .38, p < .01) was much higher than students living on-campus ( r 
= .09, ns).  
 
61
  Further subsets were chosen to explicate the different correlations.  Females by 
race are shown in Table 6.  White Female students had the highest correlation ( r = .58, 
p < .01) while Black Female students were the farthest away with a negative correlation 
( r = -.34, ns).   White and Black Female students were further divided into on-campus 
and off-campus, since housing showed a significant effect on the CRIS/M-GUDS-S 
correlation in Table 6.  On campus White Female students dropped but still had a 
relatively high correlation ( r = .40, p < .05) and on-campus Black Female students 
showed an increasing negative correlation ( r = -.64, p < .01).  Other comparisons were 
not attempted because of the small cell sizes.  
 The overall Pearson?s coefficient r = .26 ( p < .01) confirms the hypothesis that 
there is a positive correlation between the perceived quality of cross racial interactions 
and internal attitudes toward diversity.  Additionally, there are intriguing variations in 
the correlation across sub-communities of students.   
 
62
 Table 6 
 
Pearson Correlation Coefficient for CRIS and M-GUDS-S scores 
 
Demographic  
characteristic 
 
N r
 Overall  195 .26** 
Race Black 37 .10    
APA 50 -.01 
Latino 31 .41* 
White 56 .55** 
Gender Female 104 .17* 
Male 69 .43** 
Housing On-campus 85 .09 
Off-campus 110 .38** 
White Females 28 .58** 
APA Females 26 -.09 
Latina Females 15 .44* 
Black Females 22 -.34 
On campus 
Black Females 
14 -.64** 
On campus 
White Females 
19 .40* 
** correlations significant at the 0.01 level (1-tailed) 
 *   correlations significant at the 0.05 level (1-tailed) 
 
63
 Additional Analyses
 Research Question 1: Do the quality of cross racial interactions  vary by gender 
or cohort?
 A 2x4 factor univariate analysis of variance was performed on gender (male, 
female) and cohort (first year students, sophomore, junior, senior). Each participant?s 
average CRIS score (i.e., the mean across all 13 situations) was the dependent variable. 
This was a between-subjects design.   Differences among the actual 13 situations were 
not be analyzed since this hypothesis related to the overall quality of experience and not 
any particular situation. Cases with missing data on gender or cohort identification were 
excluded.  
 As in hypothesis 2 and 3, for this analysis, the CRIS scores excluded the ?No 
Experience? option,  since the purpose of this analysis is to determine the quality (i.e., 
negative to positive) of the cross racial interactions, where the No Experience selection 
is not a measure of this quality. 
 No significant differences were found, either in main effects or interactions.  
Gender F(1,163) = 1.357, p = .246, Cohort F(3,163) = 1.767, p = .156, Gender x Cohort 
F(3,163) = 1.454, p = .229.   
Research Question 2: How does the quality of cross racial interactions between students 
differ from the quality of general interactions between students, without reference to 
race?
 This research question compared the responses to the version A CRIS (race 
included) and the responses to version B CRIS (race not included). 
 A fixed effect, single factor ANOVA was run to compare the overall CRIS 
64
 scores for Version A (race) with the overall CRIS scores for Version B (non-race) All 
cases were included, 380, but the ?no experience? option was excluded to accurately 
reflect the quality of cross racial interactions. Descriptive data comparing the CRIS 
scores from both the race based version and non-race based version follows in Table 7.   
The cases for multirace, American Indian, and not reported were excluded to improve 
the performance of the ANOVA in terms of unequal cell sizes. The F test found no 
significant difference, F(1, 378) = .162, p = .687. 
 There was other evidence, however, that there were differences between the two 
CRIS scores.  A fixed effect ANOVA was run on version B (race) and found no 
significant differences among the racial groups F(3,145) = .522, p = .668, that is, when 
race was not considered, students did not differ across racial groups in their overall 
CRIS scores. However, as shown in hypothesis 2, when race was considered (Version 
A) there were significant differences between the CRIS scores for Black and Latino 
students (M difference = -.396, p = .031) and between Black and White students (M 
difference = -.419, p = .005).    
 The differences in the M for Version A and Version B are shown for the race 
groups.  Curiously, for Black students, the M for Version A is -.21 less than Version B.  
For White students the direction reverses, with the M for Version A +.21 higher than 
Version B.   
 
65
  
Table 7 
 
CRIS Scores (Versions A and B) by Race 
 
CRIS Version A  
(race) 
 CRIS Version B  
(non-race) 
N M SD N M SD
 M
 Diff.* 
 
Overall 195 3.92 .62  185 3.89 .44 +.03 
Black 37 3.64 .67  30 3.85 .36 -.21 
APA 50 3.83 .51  33 3.86 .48 -.03 
Latino 31 4.03 .62  27 3.98 .45 +.05 
White 56 4.06 .59  59 3.86 .49 +.20 
Total Race 174 3.90 .61  149 3.88 .46 +.02 
* (Version A minus Version B)  
 
66
 Preliminary validity check of CRIS.
 Pearson?s product moment correlation was used to check the construct validity 
of the CRIS instrument with a selective items from the  the Cultural Attitudes and 
Climate Questionnaire (CACQ) (Helm, Sedlacek, & Prieto, 1998).   Selected items 
from the CACQ have been used to confirm the overall ability for CRIS to measure the 
quality of cross racial interactions, as well as specific items related to the residential 
setting and the academic setting.  All respondents to the Version A (race) questionnaire 
also completed eight items from the CACQ.  In order to consistently report the degree 
to which responses are measuring the quality of cross racial interactions, the following 
responses have been excluded from this analysis: from the CACQ ?not applicable? and 
?no interaction? and from the CRIS ?no experience?.   Therefore, the number of 
respondents will vary in this analysis.  
 Overall CRIS. There should be a positive correlation between the mean CRIS 
score and the eight selected CACQ items, which are measures of racial tension, conflict 
and perceptions of fairness. These eight items were reverse scored as necessary so that 
a higher number meant the perception of a better racial climate (e.g., lesser racial 
tension/conflict).  A mean of the CRIS score (an average across the 13 situations) 
should correlate positively with means of the eight CACQ scores. A Pearson?s 
correlation matrix is shown in Table 8 with the CRIS mean and the eight CACQ scores.  
As predicted, the CRIS mean is significantly correlated (p < .01, 1 tailed) with all eight 
CACQ items. 
 Residence halls. Two of the CRIS items relate to cross racial experiences in the 
residence halls, and two of the CACQ items relate to the perception of the climate in 
67
 the halls.  The Pearson correlation matrix (not shown) found all four factors to be 
significantly correlated.  Specifically, CRIS (resident life programs) with CACQ4 
(dorm racial tension) r = .48, p < .01 and with CACQ6 (dorm staff are fair) r = .22, p < 
.01. CRIS (resident life social) with CACQ4 (dorm racial tension) r = .48, p < .01 and 
with CACQ6 (dorm staff are fair) r = .21, p < .01.  Note that CACQ4 was reverse 
scored. 
 Academic settings. Two of the CRIS items relate to cross racial experiences in 
the classroom and lab/discussion sections, and two of three of the CACQ items related 
to perceptions of the climate in the academic setting.  The Pearson?s correlation matrix 
found all four factors to be significantly correlated.  Specifically, CRIS (classrooms) 
with CACQ5 (classroom race tensions) r = .23, p < .01 and with CACQ7 (faculty are 
fair) r = .29, p < .01 and with CACQ8 (teaching assistants are fair) r = .32, p < .01.  
CRIS (lab/discussion section)  with CACQ5 (classroom race tensions) r = .22, p < .01 
and with CACQ7 (faculty are fair) r = .25, p < .01 and with CACQ8 (teaching 
assistants are fair) r = .33, p < .01.  Note that CACQ5 was reverse scored. 
 
68
 Table 8   
 
Pearson Correlation Coefficients for CACQ Items 
 
N M SD 
CRIS 
Mean 
 
CACQ1 
 
CACQ2 
CRIS Mean 195 3.92 .62 1.00   
CACQ1 - racial conflict 190 2.98 .84 .28** 1.00  
CACQ2 - race separation 195 2.15 .95 .31** .47** 1.00 
CACQ3 - cross race friends 193 2.82 .85 .18** .04 .25** 
CACQ4 - dorms race 
 tensions 
141 3.36 .85 .51** .59** .34** 
CACQ5 - classroom race 
 tensions 
188 3.60 .68 .28** .47** .30** 
CACQ6 - dorm staff are fair 134 4.10 .98 .33** .07 -.01 
CACQ7 - faculty are fair 194 4.04 .87 .34** .18** .04 
CACQ8 - teaching ass?ts are 
 fair 
193 4.09 .83 .33** .06 .02 
** correlations significant at the 0.01 level (1-tailed) 
 *   correlations significant at the 0.05 level (1-tailed) 
 
69
  
Table 8 (Continued)  
 
Pearson Correlation Coefficients for CACQ Items 
 
CACQ3 CACQ4 CACQ5 CACQ6 CACQ7 
 
CACQ8
 CRIS Mean       
CACQ1 - racial 
 conflict 
 
CACQ2 - race 
 separation 
 
CACQ3 - cross race 
 friends 
1.00      
CACQ4 - dorms race 
 tensions 
.13 1.00     
CACQ5 - classroom 
 race tensions 
.05 .56** 1.00    
CACQ6 - dorm staff 
 are fair 
.06 .25** .11 1.00   
CACQ7 - faculty are 
 fair 
.03 .20** .26** .61** 1.00  
CACQ8 - teaching 
 ass?ts are fair
 .03 .19* .13* .61** .70** 1.00 
** correlations significant at the 0.01 level (1-tailed) 
 *   correlations significant at the 0.05 level (1-tailed) 
70
  Chapter 5 
 Discussion 
 This section first examines the results of the hypotheses and research questions. 
Then, a summary of the potential meaning of these results is provided, followed by 
limitations and implications for future research. 
Hypothesis 1: Students will cluster in identifiable ways related to the similarity of their 
CRIS scores.
 The cluster analysis represents an interesting way to uncover relationships 
among a large set of both demographic and situational data.  The analysis identified 
seven clusters based on the similarity of CRIS scores for students within each cluster.  
The clusters were then distinguished from each other by examining both the variance in 
CRIS scores (1 point +/- from the mean for each CRIS situation) and variance in 
demographic profile (20% +/- from the expected percentage for each demographic 
characteristic).  
 Clusters 2, 5 and 7 seem to stand out the most in terms of providing the most 
information about cross racial experiences and student demographic data.   Clusters 2 
and 5 are high in quality of cross racial experiences and Cluster 7 is low in this quality.  
 Cluster 2 seems to be first year and sophomore students , White and Male, and 
seems particularly positive about the social environment, high in residence hall 
programs and social life, parties, sports and recreation, and Greek life.  It may be too 
simplistic to conjecture that this cluster reflects students who are somewhat insulated 
from the experiences of being a minority student on campus, and have not yet been 
confronted with the difficult (and potentially developmental) dimensions of a 
71
 multicultural environment. Ancis, Sedlacek, and Mohr (1998) found similar results with 
White students consistently reporting less racial tension, an experience of being treated 
fairly and a climate characterized by respect for diversity and having the highest overall 
satisfaction.  Findings by Mack et al. (1997) and Fisher and Hartmann (1995) also 
support this finding.. 
 Cluster 5 seems also very satisfied with their cross racial experiences, high in on 
campus and off campus jobs, residence hall programs, Greek life and off campus 
activities.  This cluster, the largest numerically, is the only one without any 
demographically distinguishing characteristics, i.e., it is most representative of the 
sample.  Cluster 5 provides evidence that the quality of cross racial experiences are not 
starkly divided along racial lines.  It represents a cross section of students, perhaps the 
largest segment of the student population (since it is the largest cluster), that reflects the 
racial and gender diversity of the campus and is generally satisfied with their 
interactions across racial lines. This is consistent with the widely held impression that 
the University of Maryland is not a racially troubled campus (Campus Assessment 
Working Group, 2003, March), indeed it has been recognized as a leader nationally in 
the area of diversity (Office of Human Relations Programs, 1998).  
 Cluster 7 reflects students who are dissatisfied, relative to the other clusters, 
with their cross racial interactions low in a remarkably large number of areas.  These 
include off campus jobs, student groups, parties, sports and recreation, study groups, 
and off campus activities.   This group also had a relatively large amount of ?no 
experience?, which explains to some extent the low CRIS scores, it also suggests that 
this group is disengaged in a large number of campus activities.  It also is the only 
72
 Cluster with Black students as a distinguishing characteristic.  Unfortunately, this 
cluster tempers the positive outcome of Cluster 5, since Cluster 7 reinforces what a 
large number of studies have found related to Black students having less positive and 
more alienating experiences in college (Ancis, Sedlacek, & Mohr, 1998; McClleland & 
Auster, 1990).     In some ways Cluster 7 stands in contrast to Cluster 2 also.  Both 
clusters are comprised primarily of first year students.   Cluster 2 is distinguished by the 
high White student count, while Cluster 7 is distinguished by the high Black student 
count (interestingly, both are absent any APA students).  Recalling that the basis for the 
clusters was their similarity in CRIS scores, perhaps APA students are neither at the 
very high end or at the very low end in this regard. 
 So these three clusters provide evidence of distinguishable communities of 
students who share demographic traits and cross racial experiences.  While the other 
clusters also have unique characteristics, these three seem the most pronounced. 
 In assessing the clusters, it is helpful to keep in mind several aspects of the data, 
specifically that certain factors covary in predictable ways. For example, far fewer first 
year students are employed while a large number of seniors are employed.  So in 
clusters 2 and 7 which have high first year students and/or sophomores, it is 
understandable that they have high unemployment.  Another consistent correlation is 
living off campus and having jobs off campus as well as the reverse.  So, cluster 1 is 
high in off-campus living and high in off-campus employment.  Another set of 
covarying factors relate demographic characteristics with CRIS scores. Since the cluster 
analysis necessarily had to include the ?no experience? selection (marked as 0 on a six 
point scale with very negative as 1 and very positive as 5), CRIS scores high in ?no 
73
 experience? should correlate with certain demographic characteristics.  For example in 
Clusters 1 and 3, high off campus living means a large number of these students 
selected ?no experience? for CRIS scores in residence life programs and 
social/friendships, as opposed to having negative experiences.  Cluster 2 has high 
unemployment, leading to low CRIS scores for on and off-campus jobs.  Recognizing 
these overlaying relationships helps to accurately assess the underlying data.  These 
relationships were examined as part of the interpretation of analysis and discussion for 
hypothesis 1. 
Hypothesis 2: The quality of cross-racial interactions for Black/African American, 
Asian Pacific American and Latino students will be lower than for White students.
 This hypothesis was partially supported in that Black students were found to 
have lower quality cross racial interactions than White and Latino students.  This 
finding  is not surprising and augments previous research which has found that Black 
students experience the college environment in less positive ways across a number of 
dimensions.   A number of studies present findings of different perceptions between 
White students and minority students, particularly African American students.  
Minority students feel more socio-cultural alienation (Mack et al., 1997); African 
American students feel the climate is less friendly (Ancis, Sedlacek, and Mohr, 1998; 
McClleland & Auster, 1990); and significantly less satisfying than White students 
(Fisher & Hartmann, 1995; Student Affairs Research Services, 1994).  The present 
study extends these findings to the specific area of cross racial interactions.  It was 
expected that a greater degree of dissatisfaction would be evidenced by the Latino and 
74
 APA students than was reflected in the ANOVA.  The Latino students had a mean 
CRIS of 4.03, comparable to a mean CRIS of 4.06 for White students.  It would prove 
interesting to find out if such a similarity in experiences between White and Latino 
students would be evident in other universities in other parts of the country since other 
research suggests greater dissatisfaction.  Vernez and Mizell (2001), for example, 
report the outcome of a focus group study involving Latino students at a number of 
universities in which social adjustment was identified as a major problem.   Latino 
students in focus groups at the University of Maryland expressed negative experiences 
as compared to other racial groups (Campus Assessment Working Group, 2003).  
Perhaps, Latino students have satisfactory experiences with one-one relationships with 
other students, while their dissatisfaction may revolve around academic retention and 
preparedness, financial aide, etc.., i.e., issues other than cross racial interactions. 
Hypothesis 3: There will be a positive correlation between the perceived quality of
 cross-racial interactions and internal attitudes toward diversity.
 The central point of this hypothesis was to test the degree to which internal 
attitudes, a relatively stable characteristic, might correlate with perceptions of the 
quality of cross racial interactions in different situations, a changing dimension, in a 
college environment.     The overall Pearson?s correlation for CRIS scores (quality of 
cross racial interactions) and M-GUDS-S (internal attitudes toward diversity) was r = 
.26, p < .01.  The positive finding lends support to the literature that contends that 
attitudes influence behaviors (Baron & Byrne, 1994), i.e., that perhaps the internal 
attitude toward diversity might predict the quality of cross racial interactions, although 
75
 it must be stressed that this study did not test for a causal relationship.  While Pearson 
coefficient was significant this overall correlation masks surprisingly wide variations in 
the correlation for sub communities of students.   
 White students have a much higher correlation when compared to APA and 
Black students.  Male students have a much higher correlation when compared with 
Female students. This latter finding might be examined in light of  Schuman et al. 
(1997) findings that women were more likely to have more liberal attitudes then men. 
Perhaps this leads to greater incongruity with the reality of the racial environment.  The 
correlations for White males are somewhat consistent with the cluster analysis of 
hypothesis 1.  Cluster 2 has distinguishing characteristics of White and Male, while 
Cluster 7 has a distinguishing characteristic of Black students. Cluster 2 reflects a high 
quality of cross racial interactions, while Cluster 7 reflects a low quality of cross racial 
interactions.  Since MG-GUDS-S scores are generally consistent across racial groups, 
and generally high, Cluster 7 seems to reflect a greater disparity between M-GUDS-S 
and CRIS than Cluster 2.   
 A greater disparity than either race or gender alone is evident when these factors 
are combined. White females have an r = .58, p< .01 while Black females have a 
negative correlation of r = -.34, p = n/s.  There are many anecdotal stories of Black 
students, particularly females, who arrived at college fully expecting a rich, diverse 
environment in which significant cross racial interaction would occur, and been 
extremely disappointed at what they find. This might lead to the ironic situation where 
those with the highest M-GUDS-S scores and the highest expectation of positive cross 
76
 racial interactions are the ones who are most disappointed and critical of their 
experiences.  
 The variation due to living situation is interesting, with on-campus students 
having an r = .09, ns and off-campus students having an r = .38, p < .01.  Perhaps those 
who live on campus are exposed to a wider variety, both good and bad, of social 
experiences than those who live off campus.  Those who live off campus have more 
limited social interactions than students who live in the residence halls and eat in the 
dining halls.  The interactions for off-campus students may revolve principally around 
academic settings, classrooms, lab/discussions, etc... and these measure fairly high in 
CRIS scores.  There is literature support for the different living experiences of on-
 campus and off-campus.  Korgen, Mahon, and Wang (2003) found that residential 
students in a very racially diverse setting were far more likely to report higher racial 
tension than their commuter counterparts. The authors conjectured that this might be 
evidence of a ?tipping effect? (Glazer, 1995, as cited in Korgen et al., 2003), that as the 
diversity of a community grows, without intentional intervention strategies, the amount 
of racial tension can also grow.  If true, this might suggest the residential environment 
warrants special attention.  It could represent the greatest potential for expanding and 
deepening positive cross racial interactions, while it might also represent the greatest 
risk for negative experiences (Newswanger, 1996). 
Research Question 1: Do the quality of cross racial interactions  vary by gender or 
cohort?
77
  The CRIS did not vary by gender or cohort (Gender F(1,163) = 1.357, p = .246; 
Cohort F(3,163) = 1.767, p = .156).   In terms of gender, on the one hand, the cluster 
analysis in hypothesis 1 only had one cluster where gender was a distinguishing 
characteristic. That means the other six clusters reflected the gender distribution of the 
sample.  On the other hand data from hypothesis 3 suggests that males and females are 
differing, at least in the extent that their internal attitudes are congruent with their cross 
racial experiences.  The lack of significant difference among the cohorts indicates that 
the quality of cross-racial interactions isn?t tied to a development dimension.  
Research Question 2: How does the quality of cross racial interactions between students 
differ from the quality of general interactions between students, without reference to 
race?
 This question was meant to primarily establish a non-race benchmark (version 
B).  Doing so would enhance the value of the race based version A, by providing a 
reference group (Hopkins, 1998).  The ANOVA did not find significant difference 
between the CRIS scores for the race-based version A questionnaire and non-race based 
version B questionnaire.  On the one hand this would suggest that race as a factor 
doesn?t substantially influence the perceived quality of the interactions between 
students. On the other hand there is clear evidence that the race-based version A created 
differences in CRIS scores that were not evident in the non-race version B.  The 
ANOVA for version A found the CRIS scores for Black students to be significantly 
below White and Latino students while no differences were found in the non-race 
version B. There is also an interesting interaction (not significant) in mean CRIS score 
78
 and race.  In Table 9, the mean CRIS score for Black students is .21 lower in the race-
 based version A than the non-race version B.  The exact opposite change occurs for 
White students, that is, the mean CRIS score increases by .21 for the race-based version 
A.   The latter finding would not be expected, unless White students actually have 
better interactions with students of other races than with students without reference to 
race.  It does suggest the possibility of social desirability affecting the CRIS scores for 
White students.  
Summary of findings
 The primary purpose of this study was to better understand the quality and 
context surrounding cross racial interactions on a college campus in order to search for 
ways in which to improve these interactions. Review of relevant literature suggested 
several directions in which to investigate.  Campus climate studies provided one line of 
research that tended to reinforce the differential nature of the experience along racial 
lines. Specifically, these studies suggested that  students of color, and Black students in 
particular, were having less positive and more alienating experiences in college as 
opposed to White students (Ancis, Sedlacek, & Mohr, 1998; Fisher & Hartmann, 1995; 
Mack et al., 1997).    
 Another set of studies focused on intergroup dynamics, emanating from the 
contact hypothesis, that reinforced the idea that the characteristics of different situations 
can have differing influences on the quality of cross racial interactions (Allport, 1954; 
Cook, 1985; Gaertner, Dovidio & Bachman, 1996; Marcus-Newhall & Heindl, 1998; 
Pettigrew, 1998).   The present study continued this line of research in seeking to find 
79
 evidence of the influence of different situations in the college environment on cross-
 racial interactions. 
 The present study also attempted to determine if the quality of cross racial 
interactions were influenced by the internal attitude toward diversity.  There is evidence 
in the literature that such a relationship should exist (Baron & Byrne, 1994; Harris et 
al., 1992).   
 So prior literature supports the idea that race matters, the situation matters, and 
internal attitudes matter.  The results of the present study were consistent with the 
findings of these previous studies.  In addition, the present study provided further 
information regarding the relationship of these factors to the quality of cross racial 
interactions on a college campus. 
 Not only did situations as measured by the CRIS vary differently in terms of the 
quality of cross racial interactions, but this study found that students could be clustered 
according to the similarity of their experiences.  These clusters yielded evidence of 
three distinguishable ?communities? of students which were reporting different cross 
racial experiences.  Cluster 2 might be labeled ?Naive Young White Males?. 
Characterized by being primarily first year and sophomore students, White and male, 
these students have high perceived quality of cross racial interactions across a number 
of situations. ?Naive? is not meant pejoratively, but to reflect the general sentiment that 
everything is fine racially, everybody gets along and what is the problem (Mack et al., 
1997)?  Cluster 7 might be labeled ?Disengaged Young Blacks?.   These are the 
classmates and roommates of the previous cluster, yet their cross racial experiences are 
80
 either low in quality or non-existent (i.e., no experience) for a large number of 
situations.  This might be reflective of those students that don?t make the successful 
transition to college that is further exacerbated by racial alienation (Humphreys, 1998; 
Reisberg, 1999).  Cluster 5 might be labeled ?Silent Satisfied Majority?.  These students 
reflect the full spectrum of campus diversity, and are in general pretty satisfied with 
their cross racial interactions. The study by the Campus Assessment Working Group 
(2003, March) found supporting data in their qualitative focus groups on the University 
of Maryland campus.  The University has a reputation of being somewhat of a leader in 
this regard and perhaps there is some basis for it (Office of Human Relations Programs, 
1998).    
 The present study further supports the literature and the cluster above in regards 
to Black students having cross racial experiences that are less positive than Whites (as 
well as Latino students in this study). The cluster analysis, however, points out the 
complexity of this finding since a large number of Black students are not having poor 
cross racial experiences.   The correlation between the internal attitude toward diversity 
and cross racial experiences provides, perhaps, the starkest reminder that clearly 
defined segments of the campus population are experiencing incongruence between 
what they may want things to be like and the way they perceive things to be.  White and 
Black females reflect the widest divergence with a .92 swing from r = .58, p< .01 to r = 
-.34, ns, respectively.  This is an added lense into better understanding cross racial 
interactions. For a  number of students, Black females in this case, the higher their 
openness toward diversity, perhaps the higher their dissatisfaction with the cross racial 
81
 environment they are experiencing.   Why would Black females have a negative 
correlation and White females have a positive correlation?  Perhaps White students, 
majority students, have more agency, and perceive a greater ability to seek out and/or 
initiate cross racial interactions that are consistent with their attitudes toward 
minorities, whereas minority students may approach the university environment more 
as an outsider perceiving their ability to shape that environment more passively. It must 
be noted, however, that this would be inconsistent with Powers and Ellison (1995) who 
disconfirmed evidence of selectivity bias, intentionally seeking out situations that 
confirm positive cross racial experiences.    Perhaps many Black females arrive at 
campus with high hopes for new horizons and interactions with people, and find it not 
much different from their previous social experiences with Whites and others.  There is 
evidence of this phenomenon in the literature (Reisberg, 1999, Smith, 1997). 
 The incongruence between expectations and reality as well as the clustering of 
students seems to suggest that the situation itself, at least as defined in this study, may 
not be a paramount factor that influences the quality of cross racial interactions.  In 
other words, for example, this study did not find that students always experience a high 
quality of cross racial interactions in classrooms and  always a low quality of cross 
racial interactions at parties.   The quality of cross racial interactions are mixed across 
the situations, mixed across the race, gender and cohort groups, and mixed across 
internal attitudes.   This by no means dismisses the idea that the situation matters. 
Clearly, there is variance in the quality of cross racial interactions across situations and 
in some cases consistent across the entire sample.  For example the academic situations, 
82
 classroom, lab/discussion, and study groups, had the three highest mean CRIS scores 
for the entire sample (see Table 2).  This study, however, simply has not pinned down 
the basis for this variance.  What this study has done is to show a better understanding 
about interrelationships between situations, demographic variables, and attitudes 
toward diversity.   
 The results of this study also confirm that much of the student population is 
fairly satisfied with the quality of cross racial interactions evident by generally high 
CRIS scores.  This should not be a reason for complacency, rather, affirmation that the 
myriad of efforts across departments, faculty and staff and the initiatives of students 
themselves ultimately have an effect.  It is through these continued efforts, perhaps 
guided by studies such as the present study, to target communities of students or 
specific environments that cross racial interactions can be improved further.  Racial 
tensions are never far from the surface.  Students naturally affiliate along lines of 
shared experience, so diligence in this area is necessary.   
Limitations
 There are several limitations of this study.  The relatively low response rate of 
37.2% increases the risk of the effect of non-response bias.  However, in recent years, 
random mailings of questionnaires to students from the Office of Institutional Research 
and Planning and also through the Campus Assessment Working Group, two units 
which develop and administer student surveys routinely, have had comparable results  
with response rates.  The current study was informed by best practices in survey 
administration and included a significant incentive (every respondent was guaranteed a 
83
 $10 or $5 gift certificate to the university book store), a short survey (10 minutes to 
complete), an initial invitation email that was first reviewed with students to maximize 
its potential to persuade students to complete the questionnaire, and a reminder 
followup.  Even so, this does not reduce the problem of systematic self selectivity in 
respondents (reduces the randomization of the sample) that is possible with a low 
response rate.   
 A second limitation in the sample is the low comparability in selected 
demographic variables between the random sample of students mailed a survey and 
those who responded.   One standard way to check non-response bias is to compare the 
demographic profile of the respondents with the full random sample.  Chi square 
analyses found significant differences in race, gender and housing.  While it is not 
known what factors would actually lead to non-response bias, the lack of comparability, 
combined with the relatively low response rate, make it likely that this sample is not 
representative of the population of all undergraduate students at the University of 
Maryland during the Fall, 2002 semester.  Therefore, great caution must be exercised in 
using the results of this study in connection with the full student population.  
 A significant limitation is that the validity and reliability of the Cross Racial 
Interaction Scale has not been established.  The reliability of this instrument is 
undetermined because it has not been tested against other factors that might explain 
changes in the quality of cross racial interactions.  For example, classes and jobs are 
typically during the week and during the day, while residence hall interactions and 
many social events (e.g., parties, sports) are more associated with the evening and to 
84
 some degree the weekends.  This and other confounds have not been explicated.  Does 
time of day make a difference?  In terms of stability, a test-retest should be used to 
ascertain the degree to which the recalled quality of cross-racial experiences are 
consistent over time. 
 A specific limitation of the CRIS was discovered after its administration.  
Respondents were allowed to select ?no experience? intended for situations in which 
they had no cross racial interactions.  However, it can also mean the participant had no 
experience with the situation (e.g., on-campus job).   
 A method bound limitation relates to the cluster analysis, which  had the 
limitation of requiring all variable values to be filled out, therefore, the ?no experience? 
selection had to be counted in the proximity matrix.  This had the effect of confounding 
the extent to which the cluster analysis was actually measuring the quality of cross 
racial experiences, since the data included ?no experience?. The effect on the cluster 
analysis was mitigated by visual inspection and where ?no experience? counts were 
significantly skewing the quality number downwards this was accounted for in the 
interpretation of the data.  
 Social desirability is always a factor that needs to be considered when asking for 
self-report data on a social related topic.  While anonymity was promised and fully half 
of the sample served as a control, the degree to which social desirability may have 
influenced responses for both the CRIS and the MGUDS-S is unknown. 
 
85
 Implications for Future Research
 Future research should focus on understanding why cross racial interactions are 
poor for some and better for others.  Focus groups may be an effective way to target, 
for example, White females with a high attitude/experience correlation, and Black 
females with a low attitude/experience correlation.  Similarly, White male first year 
students with high quality cross racial experiences and Black first year students 
students with low quality cross racial experiences might provide rich, contrasting 
qualitative data about their differences. Another approach would be to  conduct mixed 
focus groups (e.g., White and Black females together).  This might be structured more 
like an intergroup dialogue program rather than a research study.   Another area of 
future study would be to investigate the cross racial experiences of the graduate student 
body.  These students comprise almost 30 percent of the student population at the 
University of Maryland and, therefore, have a significant stake in, and can significantly 
affect, the cross racial climate on campus.  
 Since a key purpose of this study is to further knowledge about the factors that 
influence the development and maintenance of positive cross racial interactions, an 
interesting future research effort would be to focus on a representative sub sample such 
as Cluster 5 (?Silent Satisfied Majority?) and investigate further the factors they share in 
common that may be relevant to these positive outcomes.  This would pave the way to 
develop intentional interventions and to foster environments that promote such factors  
more widely. 
 Another interesting direction of further research would be to use a different 
construct to measure the internal attitudes toward diversity. Specifically,  Helms and  
86
 Carter?s (1990) White Racial Identity Attitude Scale is a measure of White people?s 
racial identity and may be a more accurate way to examine the relationship between 
cross racial interactions and internal attitudes.  
 In summary, as college campuses continue to become more diverse it becomes 
ever more important to foster environments that lead to improved race relations.  One 
key aspect to improving race relations is to improve the quality of cross racial 
interactions students have on a day to day basis.  These interactions can occur in a 
variety of settings.  This study examined the quality of cross racial interactions in 13 
situations on a college campus, and further examined the influence of internal attitudes 
and of demographic factors such as race and gender on these interactions.  This study 
found that the quality of cross racial interactions did vary by situation and that students 
clustered in identifiable ways related to the similarity of their cross racial experiences. 
This study found evidence in support of previous research that Black students continue 
to perceive their cross racial experiences less positive than their White counterparts, 
and added to this research by finding large discrepancies between Black students? 
perceived cross racial experiences and their internal attitudes toward diversity.   These 
results suggest that the quality of cross racial interactions is related to the setting in 
which they take place and the internal attitudes and demographic characteristics of the 
people involved.  However, the study falls short of uncovering an overall pattern or 
coherency in these relationships that point clearly to strategies that can foster improved 
cross racial interactions.   These findings are important because they add to our 
understanding of cross racial interactions.  Future research should focus on finding the 
factors that explain the differences observed in this study, that is, why certain 
87
 communities of students cluster in their cross racial experiences.  Find out what are the 
factors influencing those who are having positive interactions and what are the factors 
influencing those who are not. Implications to counseling psychology include being 
better informed about an important dimension of  interpersonal relationships of our 
student clients, and being better able to provide consulting services to the institution 
surrounding the psycho-social dynamics of race relations.  
 
88
  Appendix A 
 
Cross Racial Interaction Scale (CRIS)  
 Version A - references race 
 
In the following settings, my experiences with students of other [emphasis added] races 
has been (Very Negative-Negative-Neutral-Positive-Very Positive-No Experience): 
 
In the classroom 
 
In my lab/discussion sections  
 
In study groups & group projects outside the classroom 
 
On-campus employment 
 
Off-campus employment 
 
Residence hall sponsored programs 
 
Residence hall socializing; friendships 
 
Student groups 
 
Parties 
 
Participating in sports/recreation 
 
Fraternities/Sororities 
 
Campus events (e.g., First Look Fair, basketball games...) 
 
Off-campus activities stemming from on campus relationships 
 
(The actual web survey can be seen at: http://toto.umd.edu/survey/surveyer.asp)
89
  Appendix B 
 
Cross Racial Interaction Scale (CRIS)  
Version B - no reference to race 
 
In the following settings, my experiences with other students  has been (Very Negative-
 Negative-Neutral-Positive-Very Positive-No Experience): 
 
In the classroom 
 
In my lab/discussion sections  
 
In study groups & group projects outside the classroom 
 
On-campus employment 
 
Off-campus employment 
 
Residence hall sponsored programs 
 
Residence hall socializing; friendships 
 
Student groups 
 
Parties 
 
Participating in sports/recreation 
 
Fraternities/Sororities 
 
Campus events (e.g., First Look Fair, basketball games...) 
 
Off-campus activities stemming from on campus relationships 
 
90
  Appendix C 
 
Cultural Attitudes and Climate Questionnaire (CACQ)  
 Selected items 
To what extent do you believe each of the following is present at the University of 
Maryland (Little or none- Some-Quite a bit-  A great deal-  Not applicable)
 Racial conflict on campus 
 
Racial/ethnic separation on campus: 
 
Friendship between students of different racial and ethnic groups: 
 
Interracial tensions in the residence halls: 
 
Interracial tensions in the classroom: 
 
How fairly do you believe you have been treated by the following (Very unfairly-
 Unfairly-Neutral- Fairly- Very Fairly-No interaction):
 Residence hall personnel 
 
Faculty 
 
Teaching Assistants 
 
91
  Appendix D 
 
Miville-Guzman Universality-Diversity Scale-Short 
 (MGUDS-S) 
Please indicate below how descriptive each statement is of you (Strongly disagree-
 Disagree-Disagree a little bit-Agree a little bit-Agree-Strongly agree): 
 
I would like to join an organization that emphasizes getting to  know people from 
different countries. 
 
I would like to go to dances that feature music from other countries. 
 
I often listen to music of other cultures. 
 
I am interested in learning about the many cultures that have existed in this world. 
 
I attend events where I might get to know people from different racial backgrounds. 
 
Persons with disabilities can teach me things I could not learn elsewhere.    
 
I can best understand someone after I get to know how he/she is both similar and 
different from me 
 
Knowing how a person differs from me greatly enhances our friendship.    
 
In getting to know someone, I like knowing both how he/she differs from me and is 
similar to me.        
Knowing about the different experiences of other people helps me understand my own 
problems better.  
 
Getting to know someone of another race is generally an uncomfortable experience for 
me (reversed scored) 
 
I am only at ease with people of my race (reversed scored) 
 
It?s really hard for me to feel close to a person from another race (reversed scored)  
It is very important that a friend agrees with me on most issues (reversed scored).  
 
I often feel irritated by persons of a different race. (reverse scored).  
 
92
  Appendix E 
 
Demographic information: 
Gender 
 
Age 
 
Race/Ethnicity - Hispanic/Latino 
- Asian/ Pacific Islander 
- White 
- African American/Black 
- Native American 
- Please specify if none of the above 
 
Live    - On-campus   
- Off-campus. If off-campus:  
 - Live alone  
 - Share with other students 
 - Fraternity/Sorority 
 - Live with parents 
 - Other  
 
Employment:   - On-campus   
- Off-campus  
- Not Employed 
 
Class Standing - Freshman             
- Sophomore             
- Junior    
- Senior 
- Other Status 
 
Current Enrollment Status:     
- Full time student  
- Part time student 
 
Current college: 
- Arts and Humanities  - Agriculture & Natural Resources 
- Letters and Sciences  - Health and Human Performance 
- Architecture   - School of Public Affairs 
- Journalism   - Behavioral and Social Sciences 
- Education    - Library and Inf. Sciences 
- Engineering   - Computer, Math & Phys. Sciences 
- Life Sciences  - Business & Management 
- Undergraduate Studies - None 
93
  Appendix F 
 
Initial Email to one-half of the sample 
 Version A  
 (race included) 
 
Hello, 
You are one of a small number of selected students we are asking to complete a brief 
questionnaire about your experiences with students of other races on campus.  It will 
take less than 10 minutes. 
IN RETURN YOU WILL RECEIVE A $10 GIFT CERTIFICATE TO THE 
UNIVERSITY   BOOK CENTER.   
After completing the questionnaire, we will ask you where to mail the $10 gift 
certificate.  This mailing information is completely separate from your responses to the 
questionnaire. 
Your participate ?on is completely voluntary. Your questionnaire response is 
completely ANONYMOUS.  
Thank you very much for helping us to make the University of Maryland a better place 
for all students. 
Click here to continue: [URL address] 
Sincerely, 
Warren Kelley, Division of Student Affairs 
94
  
Appendix G 
 
Initial Email to one-half of the sample 
 Version B 
 (race not included) 
 
Hello, 
You are one of a small number of selected students we are asking to complete a brief 
questionnaire about your experiences with other students on campus.  It will take less 
than 5 minutes. 
IN RETURN YOU WILL RECEIVE A $5 GIFT CERTIFICATE TO THE 
UNIVERSITY BOOK CENTER.   
After completing the questionnaire, we will ask you where to mail the $5 gift 
certificate.  This mailing information is completely separate from your responses to the 
questionnaire. 
Your participation is completely voluntary. Your questionnaire response is completely 
ANONYMOUS.  
Thank you very much for helping us to make the University of Maryland a better place 
for all students. 
Click here to continue: [URL address] 
Sincerely, 
Warren Kelley, Division of Student Affairs 
95
  
References 
 Allport G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley. 
 American Commitments National Panel (1995). American Pluralism and the 
College Curriculum - Higher Education in a Diverse Democracy.(Report prepared for 
American Commitments).  Washington D.C: Association of American Colleges and 
Universities. 
 Amir, Y. (1969). Contact hypothesis in ethnic relations. Psychological Bulletin, 
71, 319-342. 
 Ancis, J. R., Sedlacek, W. E., & Mohr, J. J. (1998). Student perceptions of the 
campus cultural climate by race (Research Report # 1-98). College Park, MD: 
University of Maryland, Counseling Center. 
 Astin, A. W. (1993). Diversity and multiculturalism on campus: How are 
students affected? Change 25, 44-49. 
 Atkinson, D. R. Morten, G., & Sue, D. W. (1998). Counseling American 
minorities. (5th ed.). Boston: McGraw-Hill. 
 Baron, R. A., & Byrne, D. (1994). Social psychology: Understanding human 
interaction (7th ed.). Boston: Allyn and Bacon. 
 Barrett, D. M., & Trickett, E. J. (1979). Change strategies and perceived 
environment data: Two conceptual models and their applications. Journal of 
Community Psychology, 7, 305-312. 
 Borgen, F. H., & Barnett, D. C. (1987).  Applying cluster analysis in counseling 
psychology research.  Journal of Counseling Psychology, 34, 456-468. 
96
  Brewer, M. B., & Brown, R. J. (1998).  Intergroup relations. In D. T. Gilbert, S. 
T. Fiske, G. Lindzey (Eds.), The handbook of social psychology (pp.554-594). Boston: 
McGraw-Hill.  
 Campus Assessment Working Group (2003, March).  Campus climate and 
community: What students say. Available: www.umd.edu/cawg. College Park, MD: 
University of Maryland. 
 Carter, D. J., & Wilson, R. (1998).  Minorities in higher education: fifteenth 
annual status report. Washington D.C: American Council on Education. 
 Cook, S.W. (1985). Experimenting in social issues: The case of school 
desegregation. American Psychologist, 40, 452-460. 
 Cook, S. W., & Pelfrey, M. (1985). Reactions to being helped in cooperating 
interracial groups: a context effect.  Journal of Personality and Social Psychology, 49,
 1231-1245. 
 Cotton V., Kelley W., & Sedlacek W. (2000). Situational characteristics of 
positive and negative experiences of same race and different race students. In NASPA: 
Student Affairs Administrators in Higher Education (Ed.), Diversity on campus: 
Reports from the field (pp. 16-17, full text available at: www.naspa.org) .  Washington 
DC: NASPA.   
 Duster, T. (1991, September 25). Understanding self-segregation on the campus 
[Letter to the editor]. The Chronicle of Higher Education, p. B1. 
 Duster, T. (1993). The diversity of California at Berkeley: An emerging 
reformulation of competence in an increasingly multi cultural world. In B. W. 
Thompson & S. Yagi (Eds.), Beyond A Dream Deferred: Multi cultural Education and 
97
 the Politics of Excellence (pp. 231-256). Minneapolis: University of Minnesota Press. 
 DYG, Inc. (1998). The Ford Foundation Campus Diversity Initiative: National 
Survey of Voters.  
Available:http://www.inform.umd.edu/EdRes/Topic/Diversity/Response/Web/NewsRo
 om/polloverview.html. 
Fisher B. J., & Hartmann D. J. (1995). The impact of race on the social 
experiences of college students at a predominantly White university. Journal of Black 
Studies, 26, 117-133. 
 Flanagan, J. (1954). The critical incident technique.  Psychological Bulletin, 51,
 327-358. 
 Fuertes, J. N., & Gelso, C. J. (1998). Asian-American, Euro-American, and 
African-American students? universal-diverse orientation and preferences for 
characteristics of psychologists.  Psychological Reports, 83, 280-282.  
 Fuertes, J. N, Miville, M. L., Mohr, J. J, Sedlacek, W. E., & Gretchen, D. 
(2000). Factor structure and short form of the Miville-Guzman Universality-Diversity 
Scale. Measurement and Evaluation in Counseling and Development, 33, 157-169. 
 Fuertes, J. N, Sedlacek, W. E., Roger, P. R., & Mohr, J. (1998). College 
students? universal-diverse orientation and its relation to help-seeking behaviors, 
academic self-confidence, religiosity, and attitudes toward diversity. (Research Report 
#9-98). College Park, Maryland: University of Maryland, Counseling Center. 
 Gaertner, S. L., Rust, M. C., Dovidio, J. F., Bachman, B. A., Anastasio, P. A. 
(1994).  The contact hypothesis: the role of a common ingroup identity on reducing 
intergroup bias. Small Group Research, 25, 224-249. 
98
  Gaertner, S. L., Mann, J., Murrell, A., & Dovidio J. F. (1989). Reducing 
intergroup bias: the benefits of recategorization. Journal of Personality and Social 
Psychology, 57, 239-249. 
 Gaertner, S,L., Dovidio, J.F., & Bachman, B.A. (1996). Revisiting the contact 
hypothesis: the induction of a common in-group identity. International Journal of 
Intercultural Relationships, 20, 271-290. 
 Green, C. W, Adams, A. M., & Turner, C. W. (1988). Development and 
validation of the school interracial climate scale. American Journal of Community 
Psychology, 16, 241-259. 
 Harris, M. M., Milich, R., Corbitt, E. M., Hoover, D. W., & Brady, M. (1992). 
Self-fulfilling effects of stigmatizing information on children?s social interactions. 
Journal of Personality and Social Psychology, 63, 41-50. 
 Helm, E. G., Sedlacek, W. E., & Prieto, D. O. (1998). The relationship between 
attitudes toward diversity and overall satisfaction of university students by race.  
Journal of College Counseling, Fall 1998, 111-120. 
 Helms, J. E., & Carter, R. T. (1990). Development of the white racial identity 
inventory. In J. E. Helms (Ed.). Black and White racial identity: Theory, research, and 
practice (pp. 67-80). Westport, CT: Greenwood. 
 Hopkins, K. D. (1998). Educational and psychological measurement and 
evaluation. (8th ed.). Boston: Allyn and Bacon 
 Humphreys, D. (1998). The impact of diversity on college students: The latest 
research. In  Campus Diversity Initiative, American Association of State Colleges and 
Universities and Ford Foundation [On-line]. Available: 
99
 http://www.inform.umd.edu/diversityweb/NewsRoom/studbenefit.html 
 Hurtado, S., Dey, E. L., & Trevino, J. G. (1994, April). Exclusion or self-
 segregation?: Interaction across racial/ethnic groups on college campuses. Paper 
presented at the annual meeting of the American Educational Research Association, 
New Orleans, LA. 
 Jenkins, J. R. (1990). The college as a racial model for society. In R. Hively 
(Ed.), The Lurking Evil: Racial and Ethnic Conflict on the College Campus (pp. 25-28). 
Washington DC: American Association of State Colleges and Universities. 
 Johnson-Durgans, V. D. (1994). Perceptions of racial climates in residence halls 
between african-american and euroamerican college students. Journal of College 
Student Development, 35, 267-274. 
 Korgen, K. O, Mahon, J., & Wang, G. (2003).  Diversity on college campuses 
today: The growing need to foster campus environments capable of countering a 
possible ?tipping effect?. College Student Journal, 37, 16-27. 
 Kropp, A. J. (1992, April 22). Colleges must find ways to eradicate racial 
divisions [Letter to the editor].  The Chronicle of Higher Education, p. B3. 
 Lederman, D. (1993, October 20). Old times not forgotten. The Chronicle of 
Higher Education, p A51. 
 Lomax, R. G. (1992). Statistical concepts: A second course for education and 
behavioral sciences. Mahwah, NJ: Lawrence Erlbaum Associates. 
 Loo, C. M., & Rolison, G. (1986). Alienation of ethnic minority students at a 
predominantly white university. Journal of Higher Education, 57, 58-77. 
100
  Mack, D. E., Tucker, T. W., Archuleta, R., DeGroot, G., Hernandez, A. A., & 
Cha, S. O. (1997). Interethnic relations on campus: Can?t we all get along? Journal of 
Multicultural Counseling and Development, 25, 256-268. 
 Magen, Z. (1985). Cross-cultural personality correlates of intensity and content 
category of positive experiences. Journal of Personality and Social Psychology, 49,
 1631-1642. 
 Magner, D. K. (1990, June 6). Racial tensions continue to erupt on campuses 
despite efforts to promote cultural diversity. The Chronicle of Higher Education, p. A1. 
 Magnus son, D. (Ed.). (1981). Toward a psychology of situations: An 
interactional perspective. Hillsdale, NJ: Lawrence Erlbaum Associates. 
 Marcus-Newhall, A., & Heindl, T. R. (1998). Coping with interracial stress in 
ethnically diverse classrooms: How important are Allport?s contact conditions? Journal 
of Social Issues, 54, 813-830. 
 McClelland, K. E., & Auster, C. J. (1990). Public platitudes and hidden 
tensions. Journal of Higher Education, 61, 607-642. 
 Miville, M. L., Gelso, C. J., Pannu, R., Liu. W., Touradji, P., Holloway, P., & 
Fuertes, J. (1999). Appreciating similarities and valuing differences: the Miville-
 Guzman Universality-Diversity Scale. Journal of Counseling Psychology, 46, 291-307. 
 Moody, J. (1992, December 2). Racial Separatism on Campus [Letter to the 
editor]. The Chronicle of Higher Education, p. B3. 
 Moos, R. H. (1973). Conceptualizations of human environments. American 
Psychologist, 652-665. 
101
  Moos, R. H. (1996). Understanding environments: The key to improving social 
processes and program outcomes. American Journal of Community Psychology, 24,
 193-201. 
 Nesdale, D., & Todd, P. (1998). Intergroup ratio and the contact hypothesis. 
Journal of Applied Social Psychology, 28, 1196-1217. 
 Newswanger, J. F. (1996). The relationship between white racial identity 
attitudes and the experience of having a black college roommate. Journal of College 
Student Development, 37, 536-542. 
 Office of Human Relations Programs (1998). Diversity blueprint: A planning 
manual for colleges and universities. Washington D.C.: Association of American 
Colleges and Universities. 
 Pascarella, E. T., Edison, M., Nora, A., Hagedorn, L., & Terenzini, P. T. (1996). 
Influences on students? openness to diversity and challenge in the first year of college. 
Journal of College Student Development, 67, 174-195. 
 Peterson, D. R. (1992). Interpersonal relationships as a link between person and 
environment. In W. B. Walsh, K H. Craik, & R. H. Price (Eds.), Person-environment 
psychology: Models and perspectives (pp. 127-155). Hillsdale, NJ: Lawrence Erlbaum 
 Pettigrew, T. F. (1994). Prejudice and discrimination on the college campus. 
The Higher Education Extension Service Review - on line version, 6, 1-12. 
 Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of 
Psychology, 49, 65-85. 
 Ponterotto, J. G., & Casas J. M. (1991). Handbook of racial/ethnic minority 
102
 counseling research. Springfield, IL: Charles C.Thomas.  
 Powers, D. A., & Ellison, C. G. (1995). Interracial contact and black racial 
attitudes: the contact hypothesis and selectivity bias. Social Forces, 74, 205-226. 
 Reisberg, L. (1999, April 2). In bids to increase minority enrollments, colleges 
deal with reality and perceptions. The Chronicle of Higher Education, p. A49. 
 Rudenstine, N. L. (1996, April 19). Why a diverse student body is so important 
[Letter to the editor]. The Chronicle of Higher Education, p. B1. 
 Schneider, B., & Bartlett, C. J. (1970). Individual differences and organizational 
climate II: Measurement of organizational climate by the multi-trait, multi-rater matrix. 
Personnel Psychology, 23, 493-512. 
 Schuman, H., Steeh, C., Bobo, L., & Krysan, M. (1997).  Racial attitudes in 
America: Trends and interpretations. Cambridge, MA: Harvard University Press. 
 Sedlacek, W. E. (1996). An empirical method of determining nontraditional 
group status. Measurement and Evaluation in Counseling and Development, 28, 200-
 210. 
 Sedlacek, W. E., & Brooks, G. C., Jr. (1970). Measuring racial attitudes in a 
situational context. Psychological Reports, 27, 971-980. 
 Sigelman, L., & Welch, S. (1993). The contact hypothesis revisited: black-white 
interaction and positive racial attitudes. Social Forces, 71, 781-795. 
 Smith, D. G.(1997). Diversity works: The emerging picture of how students 
benefit. Washington D.C: Association of American Colleges and Universities. 
 Smith, E. R. (1993). Social identity and social emotions: Toward new 
103
 conceptualizations of prejudice.  In D. M. Mackie & D.L. Hamilton (Eds.), Affect, 
cognition, and stereotyping: Interactive processes in group perception (pp. 297-313). 
San Diego: Academic Press. 
 Springer, L., Palmer, B., Terenzini, P. T., Pascarella, E. T., Nora, A. (1996).  
Attitudes toward campus diversity: Participation in a racial or cultural awareness 
workshop.  The Review of Higher Education, 20, 53-68. 
 Stephan, W. G., & Stephan, C. W. (1985). Intergroup anxiety. Journal of Social 
Issues, 41, 157-175. 
 Student Affairs Research Services (1994). Building community through 
diversity (Survey - Executive summary). CO: University of Colorado at Boulder. [On-
 line]. Available: http://www.colorado.edu/SARS/divrsty/execsum.htm. 
 Southern Poverty Law Center (1999). The year in hate: Hate group count tops 
500, Internet sites soar.  In by Mark Potok (Ed.), Intelligence Report [On-line]. 
Available: http://www.splcenter.org/intelligenceproject/ip-4i1.html. 
 Trope, Y & Pomerantz, E. M. (1998).  Resolving conflicts among self-
 evaluative motives: Positive experiences as a resource for overcoming defensiveness.  
Motivation and Emotion, 22, 53-72. 
 Vernez, G. & Mizell, L. (2001). Goal: To double the rate of Hispanics earning a 
bachelor?s degree. Center for Research on Immigration Policy.  Santa Monica: RAND. 
 Wilder, D. A. (1984). Intergroup contact: the typical member and the exception 
to the rule. Journal of Experimental Social Psychology, 20, 177-194. 
 Wright, S. C., Aron, A., McLaughlin-Volpe, T., & Ropp, S. A. (1997).  The 
104
 extended contact effect: knowledge of cross-group friendships and prejudice. Journal of 
Personality and Social Psychology, 73, 73-90. 
 
105