PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

We
FEB 1 9 2015

""30 1 r'
11u_ul

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IPrq/Acc&Pres/ClRC/DateDueAindd

INVESTIGATING THE MEDIATING VARIABLES OF INTERACTIONAL
DIVERSITY ON A UNIVERSITY CAMPUS

By

Jessica C. Mills

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Sociology

2009

ABSTRACT

INVESTIGATING THE MEDIATING VARIABLES OF INTERACTIONAL
DIVERSITY ON A UNIVERSITY CAMPUS

By
Jessica C. Mills

This dissertation is a study of student race relations on a university campus and
tests the opportunity hypothesis in a university setting. Research has determined that
students benefit most from interactional diversity (i.e. ongoing and meaningful contact
across race lines) in areas such as critical thinking, problem-solving skills, and improved
self-concept. While scholars agree on the benefits of interactional diversity, there has
been limited research on the most effective ways to facilitate it.

The literature suggests that students living in racially diverse residence halls are
more likely to interact cross-racially than students living in predominantly White
residence halls. An additional factor that contributes to students’ experiences on campus
is the college classroom. Research suggests that students exposed to diversified
curriculums and an active pedagogy, regardless of the course content, exhibit higher
levels of academic achievement, critical thinking, problem-solving skills, civic
participation, and cultural awareness. Most importantly, for the purposes of this study,
research has shown that opportunities for interaction in the classroom have proven to
contribute to interactional diversity outside of the classroom.

Using Michigan State University as a case study, this dissertation aims to identify
the patterns and settings that promote and deter interactional diversity in a campus setting
by focusing on the role of residence halls and classroom environment in students’

interactions across race lines. A 10-minute online survey was developed and distributed

to all undergraduate students currently living in a residence hall on the campus of
Michigan State University (n=l 3,935) between November and December 2006. The
primary dependent variable for this study was the Interactional Diversity scale—a 10-
point scale that addresses the frequency with which respondents engaged in a variety of
activities with students of different racial backgrounds (alpha=.916). Survey
participation was representative, and results are generalizable, to the on-campus
population of undergraduates at Michigan State University.

This research led to several key findings: (1) Levels of interactional diversity
differ significantly between racial groups. Levels of interactional diversity among White
students were more easily influenced than for other students. Among students of color,
Black students were the least likely to engage in interactional diversity. Mixed Race
students were the most likely to engage in interactional diversity. This finding is a new
contribution to the literature which has largely excluded this population of students.

(2) From this study is that levels of interactional diversity are significantly
affected by first-year roommates for all students. This is especially true for White and
Asian students who are more likely to engage in interactional diversity if their first-year
roommate is not someone that they requested and of a different racial background.

(3) Results confirm that the classroom is a critical component of the interactional
diversity puzzle. The Classroom Experience scale in this dissertation—which accounts
for both curriculum and pedagogy—proved to be a significant predictor of interactional
diversity for all students. The implications of these findings are discussed. The limitations

of this study and suggestions for future research are also outlined.

Cepyright by
JESSICA C. MILLS
2009

ACKNOWLEDGEMENTS

I am indebted to my advisor, Dr. Clifford Broman, for his guidance,
encouragement and unwavering support throughout my doctoral experience. I am equally
appreciative of my committee members, Drs. Steven Gold and Maxine Baca Zinn, for
their wisdom and expertise. I am especially grateful for the opportunity to work with
committee member, Dr. Jeanne Gaze], Director of the Multi-Racial Unity Living
Experience (MRULE). Dr. Gaze] inspired me to pursue this research subject and without
her mentorship this dissertation would not have been possible.

Moreover, the success of this study could not have been achieved without the
diligent efforts of the MRULE student leaders and members. MRULE is interactional
diversity at its best and I am honored to have had their support through this process.

I would also like to thank the following student organizations for their assistance
in promoting this project: All members of the Residence Halls Association General
Assembly, including their Executive Board; Black Caucuses from all five residence hall
complexes; Lesbian, Gay, Bisexual, Transgender and Ally Caucuses from all five
residence hall complexes; in-hall governments for all seventeen residence hall; the
Council of Racial and Ethnic Student Group (Asian Pacific American Student
Organization, Black Student Alliance, Culturas de las Razas Unidas, and North American
Indian Student Organization) and the Council of Progressive Students (The Alliance of
LBGTA Students, Arab Cultural Society, Council for Students with Disabilities, Jewish
Student Union, International Student Association, and Womyn’s Council); as well as the

College Assistance Migrant Program.

Additionally, I would like to extend my gratitude to the following departments at
Michigan State University for their approval of my research topic and also their
assistance: the Office of Student Affairs, the Department of Residence Life, the
Department of Housing, and the Office for Inclusion and Intercultural Initiatives
(formerly the Office of Affirmative Action Compliance and Monitoring).

Finally, I am grateful for the friends and family who stood beside me throughout
my graduate career. The Renegades: for their wisdom, laughter, and understanding. Tony
and Chester: for brightening every day. And my family: for loving me and believing in

me fiom the very beginning. Thank you.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. xii
LIST OF FIGURES ............................................................................. xxiv
CHAPTER 1: INTRODUCTION .................................................................. 1
INTRODUCTION .................................................................................. 1
LITERATURE REVIEW .......................................................................... 3
RESEARCH FOCUS .............................................................................. 11
OUTLINE OF CHAPTERS ...................................................................... 13
CHAPTER 2: METHODOLOGY ............................................................... 15
JUSTIFICATION OF RESEARCH DESIGN ................................................. 15
Data Sources .............................................................................. 16
Summary ................................................................................... 18
DATA COLLECTION ........................................................................... 19
Measures: Dependent Variables ........................................................ 30
Measures: Independent Variables ....................................................... 32
Measures: Other Control Variables .................................................... 34
Focus Groups .............................................................................. 42
ANALYTIC STRATEGY ........................................................................ 46
Gaining Access ........................................................................... 47
LIMITATIONS: DATA COLLECTION ...................................................... 47
CHAPTER 3: PRELIMINARY RESULTS ................................................... 54
RACIAL TENSION ON CAMPUS ............................................................. 54
Racial Tension: Race/Ethnicity ......................................................... 55
Racial Tension: Gender .................................................................. 56
Racial Tension: Class Standing ............................................................ 57
Racial Tension: First-Year Status ...................................................... 58

Racial Tension: Current Hall Percent White 60
Racial Tension: Current Hall Assignment 62
Racial Tension: F irst-Year Roommate ................................................ 64
Racial Tension: Current Roommate (Upperclassmen) .............................. 65
Racial Tension: Overall Racial Environment 67
Racial Tension: Classroom Experience 67

CURRENT RACIAL CLIMATE ............................................................... 68
Racial Climate: Race/Ethnicity ......................................................... 69
Racial Climate: Gender .................................................................. 71
Racial Climate: Class Standing ......................................................... 72
Racial Climate: First-Year Status ...................................................... 73

vii

Racial Climate: Current Hall Percent White ........................................... 74

Racial Climate: Current Hall Assignment ............................................. 77
Racial Climate: F irst-Year Roommate ................................................. 78
Racial Climate: Current Roommate (Upperclassmen) .............................. 79
Racial Climate: Overall Racial Environment ......................................... 81
Racial Climate: Classroom Experience ................................................ 81
BROADENED DIVERSITY OF FRIENDS ................................................... 82
Broadened Diversity: Race/Ethnicity .................................................. 83
Broadened Diversity: Gender ........................................................... 84
Broadened Diversity: Class Standing .................................................. 85
Broadened Diversity: First-Year Status ................................................ 87
Broadened Diversity: Current Hall Percent White ................................... 87
Broadened Diversity: Current Hall Assignment ...................................... 90
Broadened Diversity: First-Year Roommate .......................................... 91
Broadened Diversity: Current Roommate (Upperclassmen) ........................ 92
Broadened Diversity: Overall Racial Environment .................................. 94
Broadened Diversity: Classroom Experience ......................................... 94
AWARE OF PERSONAL PREJUDICES ...................................................... 95
Aware of Prejudices: Race/Ethnicity ................................................... 96
Aware of Prejudices: Gender ............................................................ 97
Aware of Prejudices: Class Standing .................................................. 99
Aware of Prejudices: F irst-Year Status ............................................... 100
Aware of Prejudices: Current Hall Percent White .................................. 101
Aware of Prejudices: Current Hall Assignment ..................................... 103
Aware of Prejudices: First-Year Roommate ......................................... 104
Aware of Prejudices: Current Roommate (U pperclassmen) ....................... 106
Aware of Prejudices: Overall Racial Environment ................................. 107
Aware of Prejudices: Classroom Experience ........................................ 107
SUMMARY ....................................................................................... 108
CHAPTER 4: INTERACTIONAL DIVERSITY SCALE ................................. 109
INTERACTIONAL DIVERSITY SCALE ITEM ANALYSIS ........................... 110
I Dine or Share Meals with Students of a Different Racial Background than
Myself ..................................................................................... 110
Dine/Eat: Race/Ethnicity ........................................................ l 11
Dine/Eat: Gender ............................................................... l 13
Dine/Eat: Class Standing ...................................................... 114
Dine/Eat: Current Hall Percent White ....................................... 115
I Visit with Residents of a Different Racial Background than Myself in Their
Room ...................................................................................... 118
Visit/Their Room: Race/Ethnicity ........................................... 119
Visit/Their Room: Gender .................................................... 120
Visit/Their Room: Class Standing ........................................... 121
Visit/Their Room: Current Hall Percent White ............................ 122
I Visit with Residents of a Different Racial Background in My Room .......... 125
Visit/My Room: Race/Ethnicity ............................................. 126

viii

Visit/My Room: Gender ...................................................... 127
Visit/My Room: Class Standing ............................................. 128
Visit/My Room: Current Hall Percent White .............................. 129
I Have Meaningful and Honest Discussions about Race and Ethnic Relations with
Students from Different Racial Backgrounds than Myself Outside of Class 132

Meaningful Discussions: Race/Ethnicity: ................................... 133
Meaningful Discussions: Gender ............................................. 134
Meaningful Discussions: Class Standing .................................... 135
Meaningful Discussions: Current Hall Percent White .................... 136
I Share Personal Feelings and Problems with Students of a Different Racial
Background than Myself ............................................................... 139
Share Feelings: Race/Ethnicity ................................................ 140
Share Feelings: Gender ........................................................ 141
Share Feelings: Class Standing ............................................... 142
Share Feelings: Current Hall Percent White ................................ 143
I Study or Prepare for Class with Students of a Different Racial Background than
Myself .................................................................................... 146
Study: Race/Ethnicity ......................................................... 147
Study: Gender .................................................................. 148
Study: Class Standing ......................................................... 149
Study: Current Hall Percent White .......................................... 150
I Socialize or Party with Students of a Different Racial Background than Myself
.............................................................................................. 1 53
Socialize/Party: Race/Ethnicity .............................................. 154
Socialize/Party: Gender ....................................................... 155
Socialize/Party: Class Standing .............................................. 156
Socialize/Party: Current Hall Percent White ............................... 157
I Have Intellectual Discussions with Students of a Different Racial Background
than Myself .............................................................................. 160
Intellectual Discussions: Race/Ethnicity .................................... 161
Intellectual Discussions: Gender ............................................. 163
Intellectual Discussions: Class Standing .................................... 164
Intellectual Discussions: Current Hall Percent White ..................... 165

I Attend Events with Students of a Different Racial Background than Myself . 168
Attend Events with Students of a Different Racial Background:
Race/Ethnicity .................................................................. 169
Attend Events with Students of a Different Racial Background: Gender
..................................................................................... 170
Attend Events with Students of a Different Racial Background: Class
Standing ......................................................................... 171
Attend Events with Students of a Different Racial Background: Current
Hall Percent White ............................................................. 173

1 Attend Events Sponsored by Other Racial/Ethnic Groups that Differ from My

Own Racial/Ethnic Background ...................................................... 175
Attend Events Sponsored by Other Racial Groups: Race/Ethnicity. . . .. 176
Attend Events Sponsored by Other Racial Groups: Gender .............. 178

ix

Attend Events Sponsored by Other Racial Groups: Class Standing. 179
Attend Events Sponsored by Other Racial Groups: Current Hall Percent

White ............................................................................. 180
INTERACTIONAL DIVERSITY SCALE ANALYSIS .................................... 183
Interactional Diversity Scale: Race/Ethnicity ....................................... 183
Interactional Diversity Scale: Gender ................................................ 184
Interactional Diversity Scale: Class Standing ....................................... 185
Interactional Diversity Scale: First-Year Status .................................... 186
Interactional Diversity Scale: Current Hall Percent White ........................ 186
Interactional Diversity Scale: Current Hall Assignment ........................... 188
Interactional Diversity Scale: F irst-Year Roommate ............................... 190
Interactional Diversity Scale: Current Roommate (Upperclassmen) ............. 191
Interactional Diversity Scale: Overall Racial Environment ........................ 192
Interactional Diversity Scale: Classroom Experience .............................. 193
SUMMARY ....................................................................................... 194
CHAPTER 5: MULTIVARIATE REGRESSIONS .......................................... 199
CORRELATIONS ............................................................................... 199
REGRESSIONS: CURRENT HALL PERCENT WHITE ................................. 202
Race ....................................................................................... 204
Gender .................................................................................... 204
Academic Standing ..................................................................... 204
Pre-College Demographics ............................................................ 205
Parents’ Background .................................................................... 205
Residence Hall ........................................................................... 205
Overall Racial Environment ........................................................... 206
F irst-Year Roommate ................................................................... 206
INTERACTIONS: CURRENT HALL PERCENT WHITE ............................... 206
Race ....................................................................................... 207
Academic Standing ..................................................................... 213
Hours Socializing ........................................................................ 213
Pre-College Demographics ............................................................ 214
Parents’ Background .................................................................... 214
OTHER INTERACTIONS: RACE ............................................................ 215
Academic Standing ..................................................................... 216
Hours Socializing ........................................................................ 216
Pre-College Demographics ............................................................ 216
Parents’ Background .................................................................... 217
Overall Racial Environment ........................................................... 218
First-Year Roommate ................................................................... 218
REGRESSION: CLASSROOM EXPERIENCE ............................................. 219
INTERACTIONS: CLASSROOM EXPERIENCE SCALE ............................... 221
Race ....................................................................................... 221
Academic Standing ..................................................................... 227
Hours Socializing ........................................................................ 228
Pre-College Demographics ............................................................ 228

Parents’ Background .................................................................... 228

OTHER INTERACTIONS: RACE ............................................................ 229
Gender .................................................................................... 229
Academic Standing ..................................................................... 229
Hours Socializing ........................................................................ 229
Pre-College Environment ............................................................... 230
Parents’ Background .................................................................... 230

SUMMARY ....................................................................................... 231

CHAPTER 6: CONCLUSION ................................................................. 233

DISCUSSION .................................................................................... 233
Implications .............................................................................. 251
Limitations and Direction for Future Studies ........................................ 254
Contributions of the Research ......................................................... 259

APPENDIX A: IPPSR NOTIFICATION LETTER .......................................... 261

APPENDIX B: IPPSR FIRST E-MAIL NOTIFICATION ................................. 263

APPENDIX C: IPPSR SECOND E-MAIL NOTIFICATION ............................. 264

APPENDIX D: THE 2006 MSU COMMUNITY SURVEY .............................. 265

APPENDD( E: MSU COMMUNITY SURVEY RECRUITMENT FLYER ........... 279

APPENDD( F: MSU COMMUNITY FOCUS GROUP QUESTIONS .................. 280

REFERENCES ................................................................................... 281

xi

LIST OF TABLES

Table 2.1 Race Coded * Race based on MSU Records Crosstabulation ................... 23
Table 2.2 Race based on MSU Records "' Skipped Q34 (racial/ethnic background)
Crosstabulation .................................................................................... 24
Table 2.3 Gender based on MSU Records * Skipped Q1 Crosstabulation ................. 29
Table 2.4 Chi-Square Tests ....................................................................... 29
Table 2.5 Current Residence Hall Required ................................................... 37
Table 2.6 Current Residence Hall Requested .................................................. 37
Table 2.7 First-year roommate ................................................................... 40
Table 2.8 Current Roommate Upperclassmen Only .......................................... 41
Table 2.9 Race based on MSU Records "‘ Self-Identification Race ........................ 45
Table 2.10 Race based on MSU Records "' Skipped last question Crosstabulation ...... 52
Table 2.11 Gender * Skipped Last Question Crosstabulation ............................... 53
Table 3.1 Racial Tension on Campus: Frequencies ........................................... 54
Table 3.2 Racial Tension on Campus: Chi-Square Test Frequencies .................. 54-55
Table 3.3 Racial Tension on Campus "‘ Self-Identified Race Crosstabulation ............ 56
Table 3.4 Chi-Square Tests ....................................................................... 56

Table 3.5 Racial Tension on Campus * Self-Identified Gender Crosstabulation 57
Table 3.6 Chi-Square Tests ....................................................................... 57

Table 3.7 Racial Tension on Campus * Self-Identified Class Standing Crosstabulation 58

Table 3.8 Chi-Square Tests ....................................................................... 58
Table 3.9 Racial Tension on Campus * First Year at MSU Crosstabulation .............. 59
Table 3.10 Chi-Square Tests ..................................................................... 59

xii

Table 3.11 Current Hall Percent White * Racial Tension Bivariate Crosstabulation. 60-61

Table 3.12 Chi-Square Tests ..................................................................... 62
Table 3.13 Racial Tension Bivariate AN OVA ................................................ 62
Table 3.14 Hall Assignment without Van Hoosen * Racial Tension on Campus

Crosstabulation .................................................................................... 63
Table 3.15 Chi-Square Tests ..................................................................... 64

Table 3.16 First-Year Roommate * Racial Tension on Campus Crosstabulation 65

Table 3.17 Chi-Square Tests ...................................................................... 65
Table 3.18 Current Roommate (upperclassmen only) * Racial Tension on Campus

Crosstabulation .................................................................................... 66
Table 3.19 Chi-Square Tests ..................................................................... 67
Table 3.20 Racial Tension on Campus ANOVA .............................................. 67
Table 3.21 Racial Tension on Campus AN OVA .............................................. 68
Table 3.22 Current Racial Climate .............................................................. 68
Table 3.23 Current Racial Climate: Chi-Square Test Frequencies .......................... 69
Table 3.24 Current Racial Climate * Self-Identified Race Crosstabulation ................ 70
Table 3.25 Chi-Square Tests ..................................................................... 71
Table 3.26 Current Racial Climate * Self-Identified Gender Crosstabulation ............. 71
Table 3.27 Chi-Square Tests ..................................................................... 72
Table 3.28 Current Racial Climate * Class Standing Crosstabulation ...................... 73
Table 3.29 Chi-Square Tests ..................................................................... 73
Table 3.30 Current Racial Climate * First Year at MSU Crosstabulation .................. 74
Table 3.31 Chi-Square Tests ..................................................................... 74

Table 3.32 Current Hall Percent White "' Current Racial Climate Bivariate
Crosstabulation ................................................................................. 75-76

xiii

 

 

Table 3.33 Chi-Square Tests ..................................................................... 76

Table 3.34 Current Racial Climate Bivariate ANOVA ....................................... 77
Table 3.35 Hall Assignment without Van Hoosen * Current Racial Climate
Crosstabulation .................................................................................... 77
Table 3.36 Chi-Square Tests ..................................................................... 78
Table 3.37 First-year Roommate * Current Racial Climate Crosstabulation .............. 79
Table 3.38 Chi-Square Tests ..................................................................... 79
Table 3.39 Current Roommate (Upperclassmen only) * Current Racial Climate
Crosstabulation .................................................................................... 80
Table 3.40 Chi-Square Tests ..................................................................... 80
Table 3.41 Current Racial Climate AN OVA ................................................... 81
Table 3.42 Current Racial Climate ANOVA ................................................... 81
Table 3.43 Broaden Diversity of Friends ....................................................... 82
Table 3.44 Broaden Diversity Friends: Chi-Square Test Frequencies .................. 82-83
Table 3.45 Broaden Diversity of Friends * Self-Identified Race Crosstabulation ........ 83
Table 3.46 Chi-Square Tests ..................................................................... 84

Table 3.47 Broaden Diversity of Friends * Self-Identified Gender Crosstabulation 85

Table 3.48 Chi-Square Tests ..................................................................... 85
Table 3.49 Broaden Diversity of Friends * Class Standing Crosstabulation ............... 86
Table 3.50 Chi-Square Tests ..................................................................... 86
Table 3.51 Broaden Diversity of Friends * First Year at MSU Crosstabulation .......... 87
Table 3.52 Chi-Square Tests ..................................................................... 87
Table 3.53 Current Hall Percent White * Broaden Diversity of Friends Bivariate

Crosstabulation ................................................................................. 88-89
Table 3.54 Chi-Square Tests ..................................................................... 89

xiv

Table 3.55 Broaden Diversity of Friends Bivariate ANOVA ............................... 90

Table 3.56 Hall Assignment without Van Hoosen * Broaden Diversity of Friends

Crosstabulation .................................................................................... 91
Table 3.57 Chi-Square Tests ..................................................................... 91
Table 3.58 First-year Roommate * Broaden Diversity of Friends Crosstabulation ...... 92
Table 3.59 Chi-Square Tests ..................................................................... 92
Table 3.60 Current Roommate (Upperclassmen Only) * Broaden Diversity of Friends
Crosstabulation .................................................................................... 93
Table 3.61 Chi-Square Tests ..................................................................... 94
Table 3.62 Broaden Diversity of Friends ANOVA ........................................... 94
Table 3.63 Broadened Diversity of Friends ANOVA ........................................ 95
Table 3.64 Aware Personal Prejudices ......................................................... 95
Table 3.65 Aware Personal Prejudices: Chi—Square Test Frequencies ..................... 96
Table 3.66 Aware Personal Prejudices * Self-Identified Race Crosstabulation ........... 97
Table 3.67 Chi-Square Tests: Racial/Ethnic background .................................... 97
Table 3.68 Aware Personal Prejudices * Self-Identified Gender Crosstabulation ........ 98
Table 3.69 Chi-Square Tests ..................................................................... 98
Table 3.70 Aware Personal Prejudices * Class Standing Crosstabulation ................. 99
Table 3.71 Chi-Square Tests .................................................................... 100
Table 3.72 Aware Personal Prejudice "' First Year at MSU Crosstabulation ............. 100
Table 3.73 Chi-Square Tests .................................................................... 101
Table 3.74 Current Hall Percent White * Aware of Personal Prejudices Bivariate
Crosstabulation .............................................................................. 101-102
Table 3.75 Chi-Square Tests .................................................................... 103
Table 3.76 Aware Prejudices Bivariate ANOVA ............................................ 103

XV

Table 3.77 Hall Assignment without Van Hoosen * Aware Personal Prejudices

Crosstabulation ................................................................................... 104
Table 3.78 Chi-Square Tests .................................................................... 104
Table 3.79 F irst-year Roommate * Aware Personal Prejudices Crosstabulation ........ 105
Table 3.80 Chi-Square Tests .................................................................... 106
Table 3.81 Current Roommate (Upperclassmen Only) * Aware Personal Prejudices

recoded Crosstabulation ......................................................................... 106
Table 3.82 Chi-Square Tests .................................................................... 107
Table 3.83 Aware Personal Prejudices ANOVA ............................................. 107
Table 3.84 Aware Personal Prejudices ANOVA ............................................. 108
Table 4.1 Interaction Scale Statistics .......................................................... 109
Table 4.2 Test Statistics ......................................................................... 110
Table 4.3 Dine/Eat Different Racial Background ............................................ 111
Table 4.4 Dine/Eat Different Racial Background: Chi-Square Test Frequencies ....... 111

Table 4.5 Dine/Eat Different Racial Background * Race/Ethnicity Crosstabulation 112

Table 4.6 Chi-Square Tests ..................................................................... 112
Table 4.7 Dine/Eat Different Racial Background * Gender Crosstabulation ............ 113
Table 4.8 Chi-Square Tests ...................................................................... 114

Table 4.9 Dine/Eat Different Racial Background * Class Standing Crosstabulation 114

Table 4.10 Chi-Square Tests .................................................................... 115
Table 4.11 Current Hall Percent White without Van Hoosen * Dine/Eat Bivariate

Crosstabulation .............................................................................. 1 16-1 17
Table 4.12 Chi-Square Tests .................................................................... 117
Table 4.13 Dine/Eat Bivariate ANOVA ...................................................... 117
Table 4.14 Visit Different Racial Background Their Room ................................ 118

xvi

Table 4.15 Visit Different Racial Background Their Room: Chi-Square Test Frequencies

................................................................................................. 118-119
Table 4.16 Visit Different Racial Background Their Room * Race/Ethnicity

Crosstabulation ................................................................................... 120
Table 4.17 Chi-Square Tests .................................................................... 120

Table 4.18 Visit Different Racial Background Their Room * Gender Crosstabulation 121

Table 4.19 Chi-Square Tests .................................................................... 121
Table 4.20 Visit Different Racial Background Their Room * Class Standing
Crosstabulation ................................................................................... 122
Table 4.21 Chi-Square Tests ................................................................... 122
Table 4.22 Current Hall Percent White without Van Hoosen * Visit/Their Room
Bivariate Crosstabulation .................................................................. 123-124
Table 4.23 Chi-Square Tests .................................................................... 124
Table 4.24 Visit/Their Room Bivariate AN OVA ............................................ 125
Table 4.25 Visit Different Racial Background My Room .................................. 125
Table 4.26 Visit Different Racial Background My Room: Chi-Square Test Frequencies126
Table 4.27 Visit Different Racial Background My Room * Race/Ethnicity
Crosstabulation ................................................................................... 127
Table 4.28 Chi-Square Tests .................................................................... 127

Table 4.29 Visit Different Racial Background My Room * Gender Crosstabulation 128

Table 4.30 Chi-Square Tests .................................................................... 128
Table 4.31 Visit Different Racial Background My Room * Class Standing

Crosstabulation ................................................................................... 129
Table 4.32 Chi-Square Tests .................................................................... 129

Table 4.33 Current Hall Percent White without Van Hoosen * Visit/My Room Bivariate
Crosstabulation .............................................................................. 130-131

xvii

 

Table 4.34 Chi-Square Tests ..................................................................... 131

Table 4.35 Visit/My Room Bivariate AN OVA .............................................. 132
Table 4.36 Meaningful Honest Discussions About Race ................................... 132
Table 4.37 Meaningful Honest Discussions About Race: Chi-Square Test Frequencies

....................................................................................................... 133
Table 4.38 Meaningful Honest Discussions About Race * Race/Ethnicity Crosstabulation
....................................................................................................... l 34
Table 4.39 Chi-Square Tests .................................................................... 134

Table 4.40 Meaningful Honest Discussions About Race * Gender Crosstabulation 135

Table 4.41 Chi-Square Tests .................................................................... 135
Table 4.42 Meaningful Honest Discussions About Race * Class Standing Crosstabulatilo3n6
. Table 4.43 Chi-Square Tests .................................................................... 136
Table 4.44 Current Hall Percent White without Van Hoosen * Meaningful Discussions
Bivariate Crosstabulation .................................................................. 137-138
Table 4.45 Chi-Square Tests .................................................................... 139
Table 4.46 Meaningfirl Discussions Bivariate ANOVA .................................... 139
Table 4.47 Share Feelings ....................................................................... 140
Table 4.48 Share Feelings: Chi-Square Test Frequencies .................................. 140
Table 4.49 Share Feelings with Students of Different Racial Background *
Race/Ethnicity Crosstabulation ................................................................. 141
Table 4.50 Chi-Square Tests .................................................................... 141
Table 4.51 Share Feelings * Gender Crosstabulation ....................................... 142
Table 4.52 Chi-Square Tests .................................................................... 142
Table 4.53 Share Feelings * Class Standing Crosstabulation .............................. 143
Table 4.54 Chi-Square Tests .................................................................... 143

xviii

Table 4.55 Current Hall Percent White without Van Hoosen "' Share Feelings Bivariate

Crosstabulation .............................................................................. 144-145
Table 4.56 Chi-Square Tests .................................................................... 145
Table 4.57 Share Feelings Bivariate AN OVA ................................................ 146
Table 4.58 Study ................................................................................. 146
Table 4.59 Study: Chi-Square Test Frequencies ............................................. 147

Table 4.60 Study with Students of Different Race * Race/Ethnicity Crosstabulation .. 148

Table 4.61 Chi-Square Tests .................................................................... 148
Table 4.62 Study * Gender Crosstabulation .................................................. 149
Table 4.63 Chi-Square Tests .................................................................... 149
Table 4.64 Study * Class Standing Crosstabulation ......................................... 150
Table 4.65 Chi-Square Tests .................................................................... 150
Table 4.66 Current Hall Percent White without Van Hoosen * Study Bivariate
Crosstabulation .............................................................................. 1 5 1-152
Table 4.67 Chi-Square Tests ................................................................... 152
Table 4.68 Study Bivariate ANOVA .......................................................... 153
Table 4.69 Socialize/Party ...................................................................... 153
Table 4.70 Socialize/Party: Chi-Square Test Frequencies .................................. 154
Table 4.71 Socialize/Party * Race/Ethnicity Crosstabulation .............................. 155
Table 4.72 Chi-Square Tests .................................................................... 155
Table 4.73 Socialize/Party * Gender Crosstabulation ....................................... 156
Table 4.74 Chi-Square Tests .................................................................... 156
Table 4.75 Socialize/Party * Class Standing Crosstabulation .............................. 157
Table 4.76 Chi-Square Tests .................................................................... 157

xix

Table 4.77 Current Hall Percent White without Van Hoosen * Socialize/Party Bivariate

Crosstabulation ............................................................................. 158-159
Table 4.78 Chi-Square Tests ................................................................... 160
Table 4.79 Socialize/Party Bivariate AN OVA .............................................. 160
Table 4.80 Intellectual Discussions ............................................................ 161
Table 4.81 Intellectual Discussions: Chi-Square Test Frequencies ....................... 161
Table 4.82 Intellectual Discussions with Students of Different Race * Race/Ethnicity
Crosstabulation ................................................................................... 162
Table 4.83 Chi-Square Tests .................................................................... 162
Table 4.84 Intellectual Discussions * Gender Crosstabulation ............................. 163
Table 4.85 Chi-Square Tests .................................................................... 163
Table 4.86 Intellectual Discussions * Class Standing Crosstabulation ................... 164
Table 4.87 Chi-Square Tests .................................................................... 165
Table 4.88 Current Hall Percent White without Van Hoosen * Intellectual Discussions
Bivariate Crosstabulation .................................................................. 166-167
Table 4.89 Chi-Square Tests .................................................................... 167
Table 4.90 Intellectual Discussions Bivariate ANOVA ..................................... 167
Table 4.91Attend Events Different Racial Backgrounds ................................... 168
Table 4.92 Attend Events Different Racial Backgrounds: Chi-Square Test Frequencies
.................................................................................................. 168-169
Table 4.93 Attend Events with Students of Different Racial Backgrounds *
Race/Ethnicity Crosstabulation ................................................................. 170
Table 4.94 Chi-Square Tests .................................................................... 170
Table 4.95 Attend Events with Students of Different Racial Backgrounds * Gender
Crosstabulation ................................................................................... 171
Table 4.96 Chi-Square Tests .................................................................... 171

XX

Table 4.97 Attend Events with Students of Different Racial Backgrounds * Class

Standing Crosstabulation ........................................................................ 172
Table 4.98 Chi-Square Tests .................................................................... 172
Table 4.99 Current Hall Percent White without Van Hoosen * Attend Events With

Bivariate Crosstabulation .................................................................. 173-174
Table 4.100 Chi-Square Tests .................................................................. 175
Table 4.101 Attend Events With Bivariate ANOVA ........................................ 175
Table 4.102 Attend Events Sponsored by Other Racial Groups ........................... 176

Table 4.103 Attend Events Sponsored by Other Racial Groups: Chi-Test Frequencies 176

Table 4.104 Attend Events Sponsored by Other Racial Groups * Race/Ethnicity

Crosstabulation ................................................................................... 1 77
Table 4.105 Chi-Square Tests ................................................................... 177
Table 4.106 Attend Events Sponsored by Other Racial Groups * Gender Crosstabulaticln; 8
Table 4.107 Chi-Square Tests .................................................................. 178
Table 4.108 Attend Events Sponsored by Other Racial Groups * Class Standing
Crosstabulation ................................................................................... 179
Table 4.109 Chi-Square Tests .................................................................. 180
Table 4.110 Current Hall Percent White without Van Hoosen "‘ Attend Events Sponsored
Bivariate Crosstabulation .................................................................. 181-182
Table 4.111 Chi-Square Tests ................................................................... 182
Table 4.112 Attend Events Sponsored Bivariate ANOVA ................................. 183
Table 4.113 Interactional Diversity Scale Descriptives: Race/Ethnicity .................. 184
Table 4.114 Interactional Diversity Scale ANOVA ......................................... 184
Table 4.115 Interactional Diversity Scale Descriptives: Gender ........................... 184
Table 4.116 Interactional Diversity Scale AN OVA ......................................... 185

xxi

Table 4.117 Interactional Diversity Scale Descriptives: Class Standing .................. 185

Table 4.118 Interactional Diversity Scale ANOVA ......................................... 185
Table 4.119 Interactional Diversity Scale Descriptives: First-Year Status ............... 186
Table 4.120 Interactional Diversity Scale ANOVA ......................................... 186

Table 4.121 Interactional Diversity Scale Descriptives: Current Hall Percent White 188
Table 4.122 Interactional Diversity Scale ANOVA ......................................... 188

Table 4.123 Interactional Diversity Scale Descriptives: Current Hall Assignment 189

Table 4.124 Interactional Diversity Scale ANOVA ......................................... 189
Table 4.125 Interactional Diversity Scale Descriptives: F irst-Year Roommate ......... 190
Table 4.126 Interactional Diversity Scale AN OVA ......................................... 191
Table 4.127 Interactional Diversity Scale Descriptives: Current Roommate

(U pperclassmen) ................................................................................. 192
Table 4.128 Interactional Diversity Scale ANOVA ......................................... 192
Table 4.129 Interaction Scale AN OVA: Overall Racial Environment .................... 193
Table 4.130 Interactional Diversity Scale ANOVA: Classroom Experience ............ 193
Table 5.1 Interactional Diversity Scale Correlations ........................................ 202
Table 5.2 Coefficients ............................................................................ 203
Table 5.3 Coefficients ........................................................................... 208
Table 5.4 Coefficients ........................................................................... 209
Table 5.5 Coefficients ........................................................................... 210
Table 5.6 Coefficients ........................................... . ............................... 211
Table 5.7 Coefficients ........................................................................... 212
Table 5.8 Coefficients ........................................................................... 221
Table 5.9 Coefficients ........................................................................... 223

xxii

Table 5.10 Coefficients .......................................................................... 224

Table 5.11 Coefficients .......................................................................... 225
Table 5.12 Coefficients .......................................................................... 226
Table 5.13 Coefficients .......................................................................... 227

xxiii

LIST OF FIGURES

Figure 4.1 Classroom Experience Scale * Interactional Diversity Scale .................. 194

xxiv

CHAPTER 1: INTRODUCTION

Introduction

This dissertation is a study of student race relations on a university campus. More
specifically, this dissertation considers how the cross-racial friendships of undergraduates
living on the campus of Michigan State University are affected by such things as the
residence halls in which they live and the classes which they attend. The primary
dependent variable for this study is the Interactional Diversity scale—a lO-point scale
which addressed the frequency with which respondents engaged in a variety of activities
with students of a different racial background than themselves.

I chose to focus my research on undergraduates at Michigan State University for a
number of reasons. First, segregation experienced early in life, particularly in educational
settings, tends to be perpetuated throughout the lifecourse (Braddock 1985; Braddock,
Crain, and McPartland 1984; Gurin 1999; Milem, Umbach, and Liang 2004).
Furthermore, students are increasingly likely to attend racially segregated schools.
Research indicates that segregation across the country, both residential and educational,
is increasing (Orfield and Lee 2004; Orfield and Whitla 2001). The likelihood of
interaction across racial lines, both casual and intimate, decreases as racial/ethnic groups
become further separated from one another spatially (McPherson, Smith-Lovin, and Cook
2001; Sigelman, Bledsoe, Welch, and Combs 1996). For a growing number of students,
college is likely to be their only opportunity to experience an integrated educational
setting, making it a vital location for race scholars to consider.

Second, the value of affirmative action has recently come under fire in states

across the country. Scholars that are committed to maintaining affirmative action policies

in college admissions have worked to develop a body of scientific literature that confirms
the educational benefits of diversity on college campuses. Although the structural
diversity of an institution does play a role in student outcomes, the key finding from these
studies is that students benefit most from interactional diversity—ongoing and
meaningful contact across race lines—in areas such as critical thinking, problem-solving
skills, and improved self-concept (Antonio 2001; Antonio, Chang, Hakuta, Kenny, and a1
2004; Astin 1993; Chang 1999; Gurin, Dey, Hurtado, and Gurin 2002; Gurin, Dey,
Gurin, and Hurtado 2003; Gurin, Nagda, and Lopez 2004; Hurtado 2001; Hurtado 2003;
Hurtado, Dey, Gurin, and Gurin 2003; Hurtado 2005).

While the evidence to support this relationship is mounting, there has been limited
research on the context in which interactional diversity is most likely to occur. What
research has been done suggests that the residence halls and the classroom play a
significant role in students’ participation in interactional diversity (Antonio 2004;
Feldman 1981; Feldman and Newcomb 1969; Hallinan and Williams 1989; Hurtado
2001; Milem, Umbach, and Liang 2004; Pogrebin 1987; Saenz, Ngai, and Hurtado 2007;
Slavin and Cooper 1997). I felt it was important to begin moving the literature in that
direction in order to provide college administrators, faculty, and staff with guidance on
how best to maximize student relationships and outcomes on their campuses.

Finally, I chose to focus on undergraduates at Michigan State University because
of my ties to the school. Having attended the university during my undergraduate years,
and again as a graduate student, it is familiar to me and I had at least a basic
understanding of the campus climate and history. My involvement with the student group

MRULE (the Multi-Racial Unity Living Experience) furthered my understanding of the

undergraduate experience, particularly as it relates to the racial dynamics on campus. For
the purpose of this dissertation, I felt that a case study would provide the most
comprehensive, in-depth analysis. I chose to limit the study to this campus in order to
utilize my connections, through MRULE, to the student body and campus administrators.

Ultimately, this dissertation tests the opportunity hypothesis in a university
setting. Are students more likely to engage in meaningful ways across race lines if they
are in an environment where they regularly come into contact with peers of different
racial backgrounds? The short answer is, yes; however, context is an important

determinant.
Literature Review

One of the challenges that this research topic presents is the disconnect within the
literature. Research on the relationship between racial attitudes and housing, pedagogy,
and interactional diversity rarely overlap. As a result, compiling a review of this literature
and organizing it in a meaningful way was quite difficult. Part of the goal for this
dissertation, then, is to bridge these disparate topics. I will first review the key findings
from the interactional diversity literature. I will then compare the literature on housing
with that of student housing on college campuses. Finally, I will summarize the role of
curriculum and pedagogy on students’ race relations.

Research on interactional diversity consistently finds that it promotes complex
thinking that involves the integration of multiple perspectives (Antonio 2004). It has also
been linked to increased retention rates, overall college satisfaction, intellectual self-
concept, and social self-concept (Chang 1996; Chang 1999). Finally, interactional

diversity has been connected to positive learning and job-related outcomes such as

critical thinking, problem-solving skills, and the ability to work cooperatively with others
(Hurtado 2001).

The effects of interactional diversity are found to be beneficial for all students, but
especially for White students (Chang 1996; Gurin, Nagda, and Lopez 2004; Hu and Kuh
2003). Research on interactional diversity finds that these experiences increase Whites
students’ perceptions of commonality among African American and Latino students.
Students of color with the greatest experience interacting with peers fiom different racial
backgrounds report an increased interest in learning about groups other than their own
and a reduced sense of difference between racial/ethnic groups (Gurin, Nagda, and Lopez
2004).

Students report that interactional diversity leads to an increase in cultural
awareness and a greater commitment to racial understanding (Astin 1993). Research on
interactional diversity has shown strong support for the development of civic outcomes,
such as increased tolerance of people with different beliefs and leadership ability
(Hurtado 2001). In a longitudinal study conducted at the University of California, Los
Angeles, Van Laar, Levin, Sinclair, and Sidanius (2005) found that first-year students
who were randomly assigned a roommate with a different racial background from
themselves exhibited lower levels of prejudice and increased affect levels for all racial
groups, and especially for the racial group that the roommate belongs to. These findings
were consistent for all pairings except for students who were assigned an Asian
roommate. In addition, students assigned an outgroup roommate reported an increase in

perceived competence in interacting with members of different racial groups. Van Laar et

al. found that the effects of the roommate’s race were greater for Black and Latino
students than for White and Asian students.

While scholars agree on the benefits of interactional diversity, there has been
limited research on the most effective ways to facilitate it. More specifically, under what
circumstances are students most likely to establish meaningful relationships—in other
words, to become friends with—peers of a different racial background? There are a
number of explanations for how and why individuals choose their fiiends. Much of the
literature points to homophily, the idea that friendships are based primarily on
similarities: race, gender, age, values, interests, etc. (Antonio 2004; Joyner and Kao 2000;
McPherson, Smith-Lovin, and Cook 2001; Merton 1949; Pogrebin 1987). However,
similarities—and differences—are to a certain extent perceived. Racial segregation and
stratification can lead to assumptions of others being different when, in fact, they might
be very similar. Bonilla-Silva argues that extensive isolation of Whites from non-Whites
contributes to the development of the White habitus which typically results in a
heightened sense of difference and negative perceptions of minorities (2003). As the
contact theory asserts, the most effective way to combat this tendency is through
sustained interracial contact between individuals of equal status (Allport 1954; Pettigrew
1998; Pettigrew and Tropp 2000). In order for this to occur, however, the individuals of
different racial backgrounds must first have the opportunity to meet.

The opportunity hypothesis suggests that racial diversity is a necessary
prerequisite for cross-racial interactions (Blau 1994). Even in a diverse environment,
ingroup friendships are the most common type among all racial-ethnic groups; however

the racial composition of schools does play a role in predicting interracial fi'iendships

(Bonilla-Silva 2003; Joyner and Kao 2000). Research has shown that the likelihood of
students interacting across race lines increases as the racial diversity of the student body
increases; this is particularly true for White students (Antonio 2001; Chang 1996; Chang
1999; Chang 2001; Chang, Astin, and Kim 2004; Joyner and Kao 2000). However,
people have a tendency to expend minimal effort in establishing ties outside their local
areas (Zipf 1949). Therefore, depending on the size of the campus, student demographics
may not be enough to facilitate interactional diversity.

The organizational perspective on friendship finds that when people are grouped
together, separate from the larger population, they are more likely to meet (Feldman
1981; Moody 2001). For example, research on the effects of tracking in secondary
schools finds that as propinquity increases (i.e. the proximity and familiarity between
students) the likelihood of friendship among students also increases (Kubitschek and
Hallinan 1998). Additionally, research on interracial friendships has found that students
are most likely to become close fiiends with those peers they see on a daily basis
(Antonio 2004). This suggests that residence halls and the classroom may play a critical
role in the development of friendships in a college setting across all racial groups.

Research on campus housing has found that students who live in closer proximity
to one another are more likely to develop deeper and more meaningful friendships (Hu
and Kuh 2003; Pascarella, Terenzini, and Blimling 1994; Pogrebin 1987). Moreover, the
residence halls have been pointed to as the most likely environment for students to
develop cross-racial friendships (Antonio 2004; Pogrebin 1987). Similarly, research has
shown that roommates are an important source of fiiendships for all students (Tyson

2004). However, the opportunity for these relationships to develop may be diminishing.

While residential segregation at the national level can be attributed, in large part,
to individual discrimination (e.g., residential steering) or institutional discrimination (e.g.,
unfair housing practices, redlining), housing options available to students on college
campuses are theoretically equal across racial and ethnic lines. Despite this fact, many
campuses across the country have observed that the racial distribution of students within
their residence halls is beginning to mirror that of the general public across (Koch 1999;
Koehler 2001; Tyson 2004). More specifically, campuses with laissez-faire housing
policies, whereby students’ preferences for residence halls are given weight in the
housing assignments, are more likely to experience a gradual process of segregation,
either by race or specific student interests (Koch 1999). One popular explanation for this
de facto segregation is the Schelling model. Schelling (1971; 1972) theorized that
individual preferences for living among one’s own racial group would eventually result in
extreme segregation, even in the absence of prejudice. In other words, when students’
housing choices are based, even slightly, on racial preferences, these preferences
accumulate and result in significant segregation over time.

Research on attitudes toward integration has found support for Schelling’s model.
Whites’ support for the principle of integration of neighborhoods and schools has
increased substantially in the last sixty years. Black support for integration has always
been high (Bobo 2001; Feagin 2000; Schuman and Steeh 1996). Level of support for
integration in the abstract is nearly identical among Whites and Blacks, with
approximately three-fourths of both groups expressing support (Cashin 2004). The
meaning of integration for Blacks and Whites, however, is often very different. For most

Whites, the integration of schools and neighborhoods implies a dominant White

population with only a small number of people of color. For most Blacks, an integrated
school or neighborhood includes a substantial number of Blacks in the population (Bobo
2001; Cashin 2004; Schuman and Steeh 1996). Additionally, Blacks’ desire to live in
integrated neighborhoods results primarily from their interest in increased access to the
opportunities and resources that are not available to them in segregated Black
communities (Bobo 2001; Cashin 2004). Interestingly, despite these differences in
interpretation, Whites and Blacks are equally resistant to government-induced residential
integration (Cashin 2004).

Whites’ willingness to live in integrated neighborhoods is negatively related to
the number of African Americans living in the neighborhood, such that willingness
decreases as the proportion of Black residents increases (Farley and Schuman 1978).
Research suggests that negative stereotyping of African Americans plays a significant
role in this relationship. These results persist independent of perceptions about the
average class status of Blacks (Bobo 2001). These findings also apply to Whites’
negative stereotypes of Hispanics and Asians and the prospect of living in an integrated
neighborhood with Hispanic or Asian neighbors.

It appears that Whites are not the only racial group resistant to living among
African Americans. Using data from the 1990 General Social Survey, Herring and
Amissah (1997) determined that all racial-ethnic groups (Asian Americans, Latinos,
Western and Eastern Europeans, Native Americans, and Jewish Americans) are
significantly more likely than Afiican Americans to oppose living in predominantly
Black neighborhoods. In contrast, all racial-ethnic groups other than Native Americans

are equally willing to live in predominantly White neighborhoods.

 

While the majority of African Americans anticipate, and most research supports,
that residential integration would increase the opportunities and resources available to
them, there is some evidence that their desire to integrate is slipping as a growing number
have begun to view the costs as outweighing the benefits (Cashin 2004).

Although Schelling’s model has been applied successfully at the neighborhood
level, there is minimal support for it at the campus level. In her analysis of a large

southeastern university where increased residential segregation had been documented,
Koehler (2001) determined that preferences for racial composition in a given residence
hall did not result in significant student movement. Koehler suggests that Schelling’s
model is more appropriate for living environments that are sustained over a greater period
of time, such as residential neighborhoods, rather than short-term, temporary placements,
such as campus housing. Koehler did find evidence, however, of individual preferences
for racial compositions that reflect the respondent’s own racial background. Consistent
With previous studies on residential integration, Koehler found that Black students
Preferred more integrated residence halls than White students. Additionally, students who
rePorted having positive experiences interacting with persons of a different racial
background than themselves also expressed more tolerant racial preferences, regardless
of the student’s race. The majority of White and Black students reported their discomfort
With the possibility of being assigned a roommate of a different racial background than
themselves. White females expressed the greatest discomfort with this possibility, while
Black males were most open to the idea.
Although access to resources, neighborhood stability and safety, and housing

quality are essentially equal across campus, segregation in campus housing appears to be

on the rise. As a result, the experience of living within these halls is quite different. At
the time of this study, there was only one project that directly examined the relationship
between campus housing and interactional diversity. Tyson (2004) explored the effects
of the race of first-year roommates and racial composition of the first-year residence hall
on interracial fi'iendships at Duke University. Tyson found that for White students, the
race of the first-year roommate and the racial composition of their first-year residence
hall had little effect on their likelihood for establishing interracial friendships. The
majority of White students with interracial roommates were close friends with their
roommates, but had they had few interracial fi'iends beyond the roommate. For students
of color, however, race of their first-year roommate and the racial composition of their
first-year residence hall were strong predictors of their interracial friendships. Minority
Students with same-race roommates were significantly less likely to have interracial
friends than those students with interracial roommates. Similarly, minority students in
residence halls with higher proportions of students from their same racial background
Were less likely to have interracial friends. Tyson concludes that because White students
are in the majority on the Duke campus, the racial composition of the residence halls does
no‘ impede their ability to meet other same-race students to the extent that it affects
Students of color. Consistent with research on interactional diversity, Tyson found that
Students who had interracial friendships prior to attending college were more likely to
have interracial fi'iends during their first year at Duke University.

An additional factor that contributes to students’ experiences on campus is the
c0“98¢ Classroom. Research suggests that students exposed to diversified curriculums,

(e.g. integrated diversity courses that typically include readings on different racial/ethnic

10

 

is;

I-R

groups; participating in intergroup dialogue that lasts throughout the semester) show
significantly higher levels of civic participation, cultural awareness, critical drinking, and
satisfaction with college (Astin 1993; Gurin 1999; Hurtado 2005; Milem 1994). In
addition, students exposed to an active pedagogy (e. g. opportunities to interact with peers
fiom different backgrounds through class discussions, collaborative learning methods,
and group projects), regardless of the course topic, exhibit higher levels of academic
achievement, critical thinking, problem-solving skills, civic participation, and cultural
awareness (Chang, Astin, and Kim 2004; Cohen, Bianchini, Cossey, Holthuis, Morphew,
and Whitcomb 1997; Hurtado 2001; Slavin 1995; Wolfe and Spencer 1996). Most
importantly, for the purposes of this study, research has shown that opportunities for
interaction in the classroom have proven to contribute to interactional diversity outside of

the classroom (Hallinan and Williams 1989; Hurtado 2001; Milem, Umbach, and Liang

2004; Saenz, Ngai, and Hurtado 2007; Slavin and Cooper 1997).

Research Focus

Using Michigan State University as a case study, this dissertation aims to identify
the Patterns and settings that promote and deter interactional diversity in a campus setting
by focusing on what role, if any, the residence halls and classroom environment play in
StudentS’ interactions across race lines by addressing the following questions:

1- How do patterns of racial interaction differ between residence halls?

2- Are residence halls the greatest predictor of students’ experiences with across
racial lines? What role does the student’s race and gender play?

3- Do experiences inside of the classroom contribute to experiences outside of
the classroom with regard to cross-racial interactions among students?

11

More specifically, this dissertation will test the opportunity hypothesis by examining the
particular racial composition of the residence balls, as well as students’ experiences in the
classroom, to determine how patterns of student interaction across racial/ethnic lines are
affected. The literature suggests that students living in racially diverse residence halls
will be more likely to interact cross-racially than students living in predominantly White
residence halls due to the increased opportunity to cross paths with students of different
racial backgrounds, as well as the convenience of meeting friends within the residence
hall. Because White and Black students make up the two largest racial groups on the
campus of Michigan State University, non-Black students of color are expected to exhibit
the highest rates of interactional diversity. Residence halls are predicted to be the most
important factor in determining levels of interactional diversity; however, a number of
additional independent variables are likely to be significant as well.

It is unclear what role, if any, gender will play on students’ interactional diversity.
Similarly, the experiences of students may vary along both race and gender lines. For
example, the level of interactional diversity among White women compared with women
of color or women of color compared with men of color. The likelihood of interactional
diversity is expected to increase as class standing increases. Based on the contact
literature, consistent exposure to diverse peers should reduce students’ doubts and fears
Ofgroups they believe to be different from themselves. The first-year roommate, whether
the r00mutate was requested and the racial background of the roommate compared with
that of the participant, is also expected to play a meaningful role in the level of
interactional diversity. Students who requested their first-year roommate and who share

the Same racial background as that roommate are expected to have the lowest rates of

12

interactional diversity. In contrast, students who requested their first-year roommate and
that roommate is of a different racial background from them are expected to have the
highest rates of interactional diversity. Among students who did not request their first-
year roommate, those students whose roommate is of a different racial background than
themselves are predicted to exhibit higher levels of interactional diversity. Similar
patterns are also expected for upper-classmen relating to their current roommate. Since
the literature has not considered the role of the residence halls on interactional diversity,
it is unclear whether students who requested their current hall will differ from those who
were required to live in their current hall because of their participation in a living-
learning program and those who were randomly assigned to their hall. Finally, with
regard to the classroom experience, it is expected that students who learn about groups
Who are different from themselves and are given the opportunity to interact with diverse

peers within the classroom, regardless of subject matter, will be more likely to engage in

interactional diversity outside of the classroom.

Outline of Chapters

Each chapter of this dissertation explores the potential predictors of interactional
diversity with increasing complexity. Chapter 2 outlines the methodology of this study,
the data collection process, and describes the potential limitations of this study. Chapter 3
focuses on the preliminary results related to several one-item measures dealing with
respondents’ perceptions of the racial environment on campus as well as their personal
experiences with diversity while attending Michigan State University. Chapter 4
addresses a series of questions related specifically to respondents’ engagement with

Intel"aetional diversity which are used to compute an Interactional Diversity scale.

13

Crosstabulations are presented for each question within the scale using key independent
variables. Crosstabulations are then presented for the scale itself on these independent
variables. Chapter 5 provides the regression analysis for the Interactional Diversity scale
Possible interactions between independent variables are also examined during this
chapter. Chapter 6 explains the main findings from the analysis, addressing where and
how these findings tie into the existing literature. The implications of these findings are

also addressed. Finally, the limitations of this study are reviewed with suggestions for

future steps provided.

14

CHAPTER 2: METHODOLOGY

Justification of Research Design

As stated earlier, many colleges maintain housing policies that allow their
students to request a roommate and/or give weight to students’ requests for specific
residence halls. While the flexibility in housing assignments remains limited, a growing
number of institutions are experiencing residential isolation of students along racial and
ethnic lines (Koch 1999; Koehler 2001; Tyson 2004). Michigan State University is
among those campuses whose residence halls are becoming increasingly segregated.
Home to more than 45,000 students, including over 35,000 undergraduates, Michigan
State University has the largest single-campus residence hall system in the country with
23 undergraduate halls (at the time of this study, one of these halls was unoccupied as it
was being renovated), one graduate hall, and three apartment villages. Fifty-four percent
of students at Michigan State University are female. Racial demographics for the students
at Michigan State University show that 75% of all students identify as non-Hispanic
White, 7.9% as African American, 5.3% as Asian Pacific Islander, 3% as Chicano/other
HiSPaIIiC, and 0.8% as Native American (Newsroom 2007).

Housing data indicates that approximately one-third of all students at Michigan
State University live on campus and generally reflect gender and racial demographics
Similar to the total student body composition. During the Fall semester of 2006, non-
Hispanic Whites comprised 76.5% of the on-campus undergraduate population
(IMO-663), non-Hispanic Blacks comprised 9.6% of the on-campus undergraduate
Population (n=1340), Chicano/Latino students comprised 3.1% (n=432) of the on-campus

undergraduate population, American Indian and Alaskan Natives comprised 0.6% (n=86)

15

of the on-campus undergraduate population, Asian American and Pacific Islanders
comprised 6.0% (n=834) of the on-campus undergraduate population, and International
students comprised 2.9% (n=409) of the on—campus undergraduate population. The
racial/ethnic background for the remaining students (1.2%) was not available. When
housing data is broken down by residence halls, however, strong differences can be found
in the racial composition of specific halls. Of the 22 residence halls that undergraduates
currently reside in, non-Hispanic Whites comprise more than 85% of the population in
eight of the halls; Afiican Americans comprise more than 25% in four additional halls.
Furthermore, those halls that are predominantly White are centrally located on campus
whereas those halls with a high percentage of students of color are located along the

outskirts of campus. As a result, students are both racially segregated and spatially

isolated from one another.

Data Sources

As the purpose of this dissertation is exploratory, only those students at Michigan
State University were surveyed. Data were collected from the on-campus undergraduate
Population of Michigan State University (n=13,936). The entire population was surveyed
in an attempt to maximize the generalizability of the results. The survey yielded a 32.8
percent completion rate (n=4,576), and participation was representative by race, current
residence hall, and class standing. The response rate of men was somewhat under-
representative for all races (35.6 percent of respondents compared to 45 percent of the
Orr-campus population).

Overall, results from this study are generalizable to the on-campus undergraduate

Population at Michigan State University for these demographic variables. However, the

16

K!

3

L)

”I.

results of this survey are not generalizable to the national population of on-campus
undergraduates, as there are many unique aspects of this population. For example, the

vast majority of students (over 80%) at Michigan State University hail from within the
state of Michigan, the most segregated state in the country. Several communities and
metropolitan areas within the state of Michigan are listed as the most highly segregated in
the country (e. g. Detroit, Livonia, Sterling Heights, Warren, and Flint). Finally, Detroit is
the most segregated school district in the nation (US Census Bureau 2002). As a result,
students coming to Michigan State University are far more likely to have been raised in
segregated communities and to have attended segregated schools than students from
anywhere else. Other aspects that differentiate Michigan State University are that it is one
of the largest institutions in the country, with the total number of undergraduates
exceeding 35,000 students, and that it is a public, state-funded, research intensive
institution. Students enrolled at Michigan State University are likely to have a different
experience, once arriving on campus, than students from smaller, private, and/or liberal
arts schools.

Despite these qualities which make Michigan State University distinct, the issues
related to campus housing and student race relations that are explored in this study are
Common across campuses nationwide. In fact, concerns over student segregation in
campus housing, either through deliberate “special programs” housing options or
Students, choice, have been raised for many years now on campuses such as Duke,
Cornell, Harvard, Yale, Bryn Marwr, Dartmouth College, Louisiana State University and
Others (1994; 1995; 1996; 1998; 2001; Clark and Meyers 1995; Koch 1999; Koehler

2001; O'Neill 2003; Pinsker 1999; Rios 1994; Shea 1993; Siegel 1997; Staples 1993;

17

Staples 1996; Tyson 2004). Similarly, trends relating to the self-segregation of students
more generally, sometimes referred to as “balkanization”, have been observed from the
UC school system (Duster 1991) to the University of Pennsylvania (Staples 1993).
Although the topics of campus housing and race have been broached, there has yet to be a
study that focuses directly on this relationship. With the largest single residence hall
system in the country, Michigan State University offers a significant opportunity for
researchers to explore this issue on a broad scale.

Regarding the classroom experience, many researchers have pointed to the

benefits of diverse classrooms on a host of student outcomes. There is limited

infomation, however, on whether these experiences translate to social outcomes.
Because this study draws from the on-campus population at Michigan State University,
findings that relate to the classroom experience will not be generalizable to its entire
Student body. Nonetheless, as an exploratory study, these results should provide a strong

indicator of whether this is a relationship that should be explored in greater depth in

filtlue studies.
SW]?

A 10-minute online survey was developed and distributed to all undergraduate
students currently living in a residence hall on the campus of Michigan State University
(n§ 13,935.) Data was collected during November and December 2006. In total, 5,184
reSpondents (37.2%) “clicked” on the link and were taken to the website. The survey was
Q(>lnpleted entirely or partially by 4,576 students (32.8%). Participation was

l‘eIDresentative by race, current residence hall, and class standing. The response rate of

IIIen was somewhat under-representative for all races (35.6% of respondents compared to

18

45 percent of the on-campus population). Because the entire population was sampled, and
the proportion of participants is representative, the data should be generalizable to the
undergraduate residents at Michigan State University.
Data Collection
The survey was distributed by the Institute of Public Policy and Social Research
(IPPSR) at Michigan State University. Data was collected using WebSurveyor (V ovici)
Software. A unique, custom designed data collection instrument was created for this
project according to the researcher’s specifications. The data collection instrument was
thoroughly tested to ensure the question wording, skip patterns, and data collection
format were accurate and met the researcher’s specifications. Access to the names, basic
demographics, and e-mail accounts of on-campus residents was provided to IPPSR by the
Office of Student Affairs at Michigan State University. All students at Michigan State
University have an e-mail account and free access to the intemet. Every student in the
population received a personalized prenotification letter from IPPSR via campus mail
informing them about the upcoming survey and its release date. Research on personalized
cOntact with participants has been found to significantly increase response rates
(I‘Ieerwegh 2005; Heerwegh, Vanhove, Matthijs, and Loosveldt 2005). Research on Web
Slll'vey response rates has found that when surveys are preceded by prenotification letters
in the mail, they can achieve similar response rates as hard copy questionnaires delivered

through mail (Kaplowitz, Hadlock, and Levine 2004).

Each prenotification letter included five pre-incentive coupons for local

lDIISinesses:

$1.00 Off Ice Cream Creation from Cold Stone Creamery
Free and Discounted Food Items from Bruggers Bagels

19

10% Off Custom Printing from Campus Corner

7 VIP Day Memberships to Powerhouse Gym in East Lansing

A Hand or Chair Massage or Mini Make-Up Session from Douglas J Salon.

Research on pre-incentives finds that monetary incentives significantly increase
response rates compared with no incentives. Additionally, respondents’ answers are not
significantly affected by incentives such that answers are largely consistent whether

incentives are provided or not (James and Bolstein 1990).

Because of the size of the population being surveyed, the sample was divided into

three groups by residence hall, as equally as possible, based on where each hall is located

on campus.

Group 1 (North Campus): Brody Complex, West Circle, Van Hoosen, and Mason-

Abbott (n=4705)
Group 2 (East Campus): Hubbard, Akers, Holmes, and McDonel (n=4264)

Group 3 (South/Central Campus): Wonders, Case, Wilson, Holden, and Shaw
(n=4967)

Letters were mailed to the respondents in each group so that they would arrive 2 to 3 days

Prior to receiving the email invitation. The letters were sent on the following dates:

Group 1 (North Campus): November 8, 2006

Group 2 (East Campus): November 9, 2006

Group 3 (South/Central Campus): November 10, 2006
El‘nail invitations were sent to each respondent and delivered on the day that the online
SIll'vey opened for that group. By splitting the respondents into three groups, the volume
of students attempting to participate in the survey was distributed and overloading the

Sel'ver was avoided. The email invitations were distributed on the following dates:

Group 1 (North Campus): November 14, 2006
Group 2 (East Campus): November 15, 2006
Group 3 (South/Central Campus): November 16, 2006

20

Using WebSurveyor (V ovici) software, IPPSR sent each student a personalized
email invitation (and reminders) with unique links for the survey that allowed data to be
passed into the survey instrument. Each student’s race, gender, current residence hall,
and class standing were embedded within the personal link. This link served two

important purposes. First, it ensured that each student only participated in the survey one
time. WebSurveyor allows only one response per unique id, email, or IP address. In
addition, the software contains a unique tracking feature: If a respondent does not
“submit” their survey at the end, they are considered incomplete until it is submitted or
data collection period ends. Once a respondent “submits” their survey, the data is
collected for that respondent. While the students could take the survey an unlimited
number of times, only their first submission was recorded. The second benefit of the
embedded information is that it enabled us to accurately gauge participation rates for
Specific populations (race, gender, class standing, and current residence hall) during the

Conrse of data collection. As a result, recruitment efforts could be targeted to those

populations.
Participants were asked to provide their racial/ethnic background (an open-ended

ghestion), gender, class standing, and current residence hall within the survey. They were

l'1(>t informed that this information was embedded in their links. However, only self-
I‘eported answers for these questions are used in the data analysis. Since the embedded
information was only used for recruitment purposes, and not in the analysis stage, the
ic1entities of the students have not been compromised. Aside from the ethical issues that

Would be raised by using the embedded information for anything more than recruitment,

t1"left: is an additional reason why it would be inappropriate to use it, especially the

21

4 .

ur-

information on race. It appears that the university records are not accurate representations
of how respondents self-identify: a number of participants’ responses to the racial/ethnic
question in the survey are not consistent with how the university has them listed. Perhaps
this is because the university does not provide enough racial categories for students to
choose fiom, or because the university does not allow students to choose multiple
categories. An additional possibility is that when students first enter the university,
typically at 18-years-old, their racial identity has not been established. By the time of this
survey—for some this may be four or five years later—their racial identity has changed
so that university records no longer reflect how they identify. Whatever the reason, there
i s substantial variation between university records and participants’ self-identified
racial/ethnic labels (see table 2.1). This is particularly true for students categorized as
Afi'ican American, Native American, and Asian American by the university but identify
here as being multiracial (coded as Mixed Race by the researcher).

Students were not asked to identify their domestic status (i.e. International
Stl-ldents). As a result, 3.9 percent of the participants are distributed within these groups
based on the racial/ethnic groups they identified with. The majority of these students
now appear in the Asian category. This researcher does not assume that the experiences
01‘ International students are consistent with those of domestic students. However, in his
reSearch at Duke University, Tyson (2004) found that International students did not differ
Significantly from domestic students in their likelihood of having interracial fiiends.
Sil'lce the total number of International students in this sample is relatively small, the

OV-r'erall effect of removing this variable should not be significant. Using racial/ethnic

22

- 1
13'4““
"b b- «M

‘1

labels that participants currently identify with provides a clearer picture of their

 

 

 

 

 

 

 

 

 

perspective.
Table 2.1 Race Coded * Race based on MSU Records Crosstabulation
Race based on MSU Records
American Asian
Non- Non- Indian, American,
Hispanic Hispanic Chicano/ Alaskan Pacific
Race Coded White Black Latino Native Islander International T0131
White/ Count 2898 5 8 3 5 16 2935
Caucasian % 99.0% 1.7% 7.5% 13.6% 2.2% 13.0% 79.5%
Black Count 4 256 l 0 0 5 266
% .1% 88.3% .9% .0% .0% 4.1% 7.2%
Latino Count 5 0 97 2 0 10 114
% .2% .0% 90.7% 9.1% .0% 8.1% 3.1%
Native Count 2 0 O 6 2 0 10
American % .1% .0% .0% 27.3% .9% .0% 3%
Asian Count 0 0 0 0 201 92 293
% .0% .0% .0% .0% 88.9% 74.8% 7.9%
Mixed Count 17 29 1 ll 18 0 76
Race % .6% 10.0% .9% 50.0% 8.0% .0% 2.1%
Total Count 2926 290 107 22 226 123 3694
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

 

 

 

One final note related to the difference between embedded and self-identified
information: a number of participants chose not to answer the question regarding
I‘aCial/ethnic background, either because they did not complete the survey to that point or
l)ecause they did not want to provide that information (see Table 2.2). The vast majority
of those students, 74.3 percent, were non-Hispanic White (according to university
reCords). Students the university identified as Chicano/Latino were most likely to skip
tl'lis question: 64.3 percent of all students identified by the university as Chicano/Latino
did not answer the race/ethnicity question. Non-Hispanic Whites and Asian
Atnerican/Pacific Islanders were the least likely groups to skip this question (17.1% and

l 6-0% respectively)-

23

Table 2.2 Race based on MSU Records * Skipped Q34 (racial/ethnic background)

 

 

 

 

 

 

 

 

 

_ Crosstabulation
Skipped Q34
Completed Skipped race
Race based on MSU Records race question question Total

Non-Hispanic White Count 2921 603 3524

/ % within MSU race 82.9% 17.1% 100.0%
% within skipped34 81.6% 74.3% 80.2%
Non-Hispanic Black Count 290 87 377

% within MSU race 76.9% 23.1% 100.0%

% within skipped34 8.1% 10.7% 8.6%

Chicano/Latino Count 10 18 28
% within MSU race 35.7% 64.3% 100.0%

% within skipped34 .3% 2.2% .6%

Anterican Indian, Count 20 7 27
Alaskan Native % within MSU race 74.1% 25.9% 100.0%
% within skipped34 .6% .9% .6%

Asian American, Count 226 43 269
Pacific Islander % within MSU race 84.0% 16.0% 100.0%
% within skipped34 6.3% 5.3% 6.1%

International Count 1 1 3 54 167
% within MSU race 67.7% 32.3% 100.0%

% within skipped34 3.2% 6.7% 3.8%

TOtal Count 3580 812 4392
% within MSU race 81.5% 18.5% 100.0%

% within skipped34 100.0% 100.0% 100.0%

 

 

 

 

 

 

A number of security measures were taken to protect the rights and privacy of

each student in the sample. All emails were sent individually so no other respondent

lmantles appear on any email. WebSurveyor has other security features that ensure the

integrity of the data and meet federal research standards. The Office for Survey Research

(08R) stores all identifying and survey data associated with the web survey on its own

S"~"<-‘-ure server. WebSurveyor employs 128 bit data encryption and all data is downloaded

onto a password protected computer. Finally, all research personnel having any contact

with research participants or the data have signed pledges of confidentiality assuring that

fFe<1eral standards for protecting the rights and privacy of research participants are met.

'1'." "
53.5. S

..
'1 9"
1‘ .
it W

..

FD} V

Students had the option to skip any questions and were permitted to leave the survey at
any time. The tracking system of WebSurveyor allowed reminders and follow-ups to only
be sent to those students who had not responded or completed the survey. If a respondent
did not “submit” their survey, they received subsequent reminders until it was submitted
or the data collection period ended. Students who did not complete the survey within the
first week received a follow-up personalized email reminder from IPPSR. These
reminders were divided in the same manner as the notification letters and email

invitations so that students who had not yet participated were contacted in three groups.

Group 1 (North Campus): November 18, 2006
Group 2 (East Campus): November 19, 2006
Group 3 (South/Central Campus): November 20, 2006.

A second reminder was sent to select students who had not completed the survey
by the second week. Because of the embedded information, response rates could be
tracked based on their demographics: race, gender, and current hall. It was determined
that participation of students of color, men, and students residing in particular halls was
loV's/er than ideal. In an effort to boost participation in these populations, the second
l‘elninder email was sent on November 30, 2006 to all men on campus and all women
1‘ esiding in the following halls: Bailey, Butterfield, Emmons, Hubbard, Mayo, Wonders,
and Wilson (n=6,654). When we compare participation rates before this email was sent
Out to the final rates for each group, it appears that this was a useful strategy. An

additional 254 men participated, bringing their proportion of respondents up nearly 3
percentage points. Increases were also seen in participation among students of color,
particularly among Afiican Americans and Latinos. Targeting specific halls also proved

beneficial: the percentage of residents in each hall increased so that no hall showed less

25

 

 

 

 

than a 25 percent response rate. Wonders Hall had the lowest participation rate of 25.4
percent (incidentally, this is also the hall with the highest percentage of White residents

on campus) and Van Hoosen Hall had the highest participation rate of 53.1 percent (this

is an all-female hall).

A post-incentive was also employed in this survey. Respondents were offered a
chance to win an iPod 2GB or one of two iPod shuffles (1GB). The drawings were staged
to encourage respondents to respond quickly to email invitations. Research on promised
incentives has found that prize draws significantly increase response rates as well as

completion rates of web surveys (Bosnjak and Tuten 2003).

Compared to other student surveys that were conducted at Michigan State
University during the 2006-07 academic year, this survey yielded higher response rates.
For example, the surveys conducted through the Department of Residence Life resulted
in response rates lower than 20 percent (Residence Life First-Time First-Year First-Week

Survey 17%; Residence Life First-Time First-Year January Survey 12%) (Lange 2006;

Lange 2007).

In addition to the incentives and the email reminders provided by IPPSR, a variety
of additional methods were used to increase student participation in the survey. Flyers
Were posted across campus, both in residence and academic halls. The Residence Halls
Association also posted these flyers in their advertisement rotation that runs on on-
ciatrrpus television channels and prior to on-campus film showings. A large number of
s‘llcfent organizations on campus also encouraged their constituents to participate through
faCe-to-face contact and e-mails. All members of the Residence Halls Association

General Assembly helped to promote this research, including their Executive Board;

26

 

Black Caucuses from all five residence hall complexes; Lesbian, Gay, Bisexual,
Transgender and Ally Caucuses from all five residence hall complexes; in-hall
governments for all seventeen residence hall; the Council of Racial and Ethnic Student
Group (Asian Pacific American Student Organization, Black Student Alliance, Culturas
de las Razas Unidas, and North American Indian Student Organization) and the Council
of Progressive Students (The Alliance of LBGTA Students, Arab Cultural Society,
Council for Students with Disabilities, Jewish Student Union, International Student
Association, and Womyn’s Council). Other organizations that were involved in recruiting
Participants were the Multi-Racial Unity Living Experience (MRULE) and the College
Assistance Migrant Program. In addition to these organizations, many of the complex
directors and residence hall mentors encouraged their residents, through face-to-face
Contact as well as by providing additional incentives such as points toward their hall
1‘ eWards system, to participate in the study.

The intent of including all of these student organizations in the recruitment
process was to increase the overall participation rate. Research finds that college students
are more likely to hold a particular attitude as the percentage of peers who hold that value
increases (Pascarella and Terenzini 1991). Given the significance of peer influence on
(>011ege students, it was expected that students would be more responsive to solicitation
fi‘om their own peers than from administrators or the researcher. In addition, the range of
student groups and mentors involved hopefully diminished perceptions that this study
was simply a “race thing”. This was an especially important message to convey to the

large proportion of White students on campus who may not have understood that their

participation was as critical as the participation from students of color.

27

 

Unfortunately, there is no way to be certain that these additional measures
significantly improved the overall response rate. However, the racial demographics of
other student surveys at Michigan State University suggest that these recruitment efforts
helped to increase minority participation. For example, 82 percent of participants in
surveys conducted through the Department of Residence Life during the 2006-07
academic year were non-Hispanic White compared to the actual 76.5 percent that they
represent of on-campus undergraduates. In contrast, 77 percent of the participants for this
Study were non-Hispanic White. Compared to other surveys conducted during the 2006-
07 academic year, this survey did not exhibit an appreciable difference in participation
rates based on gender.

An additional benefit to having the embedded race and gender of the participants
is that it gives us a better idea of who did not complete the survey, that is who “clicked”
the survey but did not answer any of the questions. There are 413 respondents who
oI)ened the survey but failed to answer even the first question. Using the embedded
university records, we can tell that the majority of each racial/ethnic group completed the
first question (see Table 2.3). American Indian/Alaskan Native students, as identified by
tlite university, failed to answer this question at a far greater rate than any other group
followed by International students and Asian American/Pacific Islanders. Despite their
lower rate of completion than other groups, the number of participants in these
DOpulations who completed this survey is proportionate to the on-campus population.
Non-Hispanic White students (per university records) were the most likely to complete

this first question. This fact adds to the overrepresentation of Whites in the survey.

Although non-Hispanic Whites represent 79 percent of all those who completed the first

28

 

‘5" ‘
‘ .-_'

. {3

question, however, this does not differ substantially from their on-campus proportion of
76.5 percent. Completion rates for the first question suggest that findings of this survey
should not be altered in a meaningful way.

Using the embedded information to consider gender, women were only slightly
more likely to complete the survey than men (see Table 2.3). While women are
overrepresented in total respondents, it appears that men and women were equally likely

to complete the survey once they opened it (Table 2.4).

Table 2.3 Gender based on MSU Records * Skipped Q1 Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Skipped Q1
Gender based on MSU Records Completed Q1 Skipped Q1 Total
F Count 2702 243 2945
% within gender 91.7% 8.3% 100.0%
% within incomplete 64.9% 58.8% 64.4%
M Count 1461 170 1631
% within gender 89.6% 10.4% 100.0%
m % within incomplete 35.1% 41.2% 35.6%
Total Count 4163 413 4576
% within gender 91.0% 9.0% 100.0%
x % within incogiplete 100.0% 100.0% 100.0%
Table 2.4 Chi-Square Tests
Asymp. Sig.
Value df (2-sided)
Pearson Chi-Square 6.030a 1 .014
Likelihood Ratio 5.9201 1 .015
of Valid Cases 4576

 

 

 

 

 

The embedded university records allow us to see that students who opened the
SI-lrvey do not differ substantially, at least in terms of race and gender, from those who
cl‘lose to participate. The possibility that some students—White students in particular—
may have perceived this survey as “a race thing,” appears to be of minimum concern

then. Furthermore, we can be reasonably certain that the results of this survey are

29

representative of the entire undergraduate on-campus population at Michigan State
University.
Measures: Dependent Variables
There are five dependent variables that this survey addresses: racial tension,
racial climate, broadened diversity of friends, more aware of prejudices, and interaction.
Racial tension refers to a single-item question that measures students’ perceived tension
on campus at the time of the survey. Response choices for the first dependent variable
ranged from 1 (strongly disagree) to 4 (strongly agree):
There is racial tension on campus.
Racial climate refers to a single-item measure of students’ perceptions of the
current racial climate on campus. Response choices ranged from 1 (poor) to 4 (excellent).

In your opinion, would you say that the current racial climate on campus is poor,
fair, good, or excellent?

Broadened diversity of friends refers to a single-item question that asked students
the degree to which they felt they had broadened the diversity of their immediate circle of
fi‘iends (e. g. race, sexual orientation, background, religious, culture, etc.) since coming to
Michigan State University. This question was modeled off of the Michigan State
University Department of Residence Life survey of F irst-Time F irst- Year Students which

i S distributed during the first week of every fall and spring semester. Response choices
ratlged from 1 (strongly agree) to 4 (strongly disagree).
”inking about your experiences both at MSU and within your residence hall, please

indicate how strongly you agree or disagree with each statement: I have broadened
the diversity (e. g., race, sexual orientation, background, religious, culture, etc.) of my

immediate circle of friends.

30

More aware of prejudices refers to a single-item question that asked students the
degree to which they felt they had become more aware of their personal prejudices and
stereotypes towards others since coming to Michigan State University. This question was
also adopted fi'om the Michigan State University Department of Residence Life survey of
First-Time F irst— Year Students which is distributed during the first week of every fall and

spring semester. Responses ranged from 1 (strongly agree) to 4(strongly disagree).
Thinking about your experiences both at MS U and within your residence hall, please
indicate how strongly you agree or disagree with each statement: I have become
more aware of my personal prejudices and stereotypes toward others.
'I‘he limitations of single-item variables are discussed toward the end of this chapter.
Interactional Diversity refers to a 10-item scale (alpha=.916) that measures the
extent to which students engage in meaningful interactions across race lines. This
measure is an attempt to operationalize the concept of interactional diversity. The
Particular items used in this scale were chosen because they often appear in the
interactional diversity literature. Of the ten items, eight come from the Preparing
Students for a Diverse Democracy Project, a longitudinal multi-campus study designed to
address the educational benefits of racial diversity on college campuses. Saenz, Ngai, and
Hunado (2007) chose these eight items to develop the Frequency of Positive Interactions
across Race. Because this survey is primarily interested in the relationship between
il'lteractional diversity and the residence halls, the researcher chose to add two additional
items relating to interactions within the dorm rooms themselves. Drawing from the social
di Stance literature (Bogardus 1933; Park 1924), interactions that occur in the more

mti mate setting of dorm rooms suggest an increased comfort with interactions across race

llneS, particularly when the interaction occurs within the respondent’s own dorm room.

31

For each item of the Interactional Diversity scale, responses ranged from 1(never) to

4(very ofien).

Please think about the experiences you have had in your current residence hall.
For each of the following examples, please indicate how often each occurs in your
current residence hall—never, occasionally, often, or very often.

I.

2.

3.

I dine or share meals with students of a diflerent racial background than
myself (Saenz, Ngai, and Hurtado 2007)

I visit with residents of a diflerent racial background than myself in their
room.

I visit with residents of a different racial background than myself in my room.

Please think about your experiences both at MS U and within your current
residence hall. Please indicate how often each of the following occurs—never,
occasionally, often, or very often.

4.

8.

9.

I have meaningful and honest discussions about race and ethnic relations with
students from racial backgrounds different than my own outside of class.
(Saenz, Ngai, and Hurtado 2007)

I share personal feelings and problems with students of a diflerent racial
background than myself (Saenz, Ngai, and Hurtado 2007)

I study or prepare for class with students of a different racial background
than myself (Saenz, Ngai, and Hurtado 2007)

I socialize or party with students of a diflerent racial background than myself
(Saenz, Ngai, and Hurtado 2007)

I have intellectual discussions with students of a different racial background
than myself outside of class. (Saenz, Ngai, and Hurtado 2007)

I attend events with students of a diflerent racial background than myself
(modified from Saenz, Ngai, and Hurtado 2007)

10. I attend events sponsored by other racial/ethnic groups that difler from my

own racial/ethnic background. (Saenz, Ngai, and Hurtado 2007)

Measures: Independent Variables

Several demographics were controlled for. There were multiple race dummy

Variables created from respondents’ self-identified racial/ethnic backgrounds: Black,

ClTliczano/Latino, American Indian, Asian, and Mixed Race were each compared to

Mite/Caucasian students. Research on racial attitudes, perceptions of campus climate,

mid interactional diversity consistently find race to be a significant variable (Bonilla-

32

I
1...,
3.. ' nil

12.11

‘ ..
lw' I

|r‘-
I inns

I‘“

l
qul~

H55

WV

..
J‘s”

(7‘4:
...- ~
”:1 .

‘1“:—

L4““

'11 - '1

.., l‘

511'
r I)
V

Silva 2003; Chang 1996; Chang, Astin, and Kim 2004; Chavous 2005; Hu and Kuh 2003;
Joyner and Kao 2000; Koehler 2001; Saenz, Ngai, and Hurtado 2007; Sigelman, Bledsoe,
Welch, and Combs 1996; Suarez—Balcazar, Orellana-Damacela, Portillo, Rowan, and

Andrews—Guillen 2003; Tyson 2004). As discussed earlier, students were not asked to

identify their domestic status (i.e. International students).

A dummy variable was also created for gender (1 =female). Research on the
relationship between gender and interactional diversity has been inconsistent and varies
across racial groups (Antonio 2001; Bonilla-Silva 2003; Cowan 2005; Koehler 2001;
McClelland and Linnander 2006). For example, in his research on college students,
Bonilla-Silva (2003) found that working-class, White women were more likely to have
meaningful relationships across race lines than other groups. Other scholars have

suggested that boys and men are more likely to have interracial friendships than girls and

Women because of their higher rates of participation in sports (Tatum 1997).

Mother’s education, father’s education, and family income were also controlled
fO'r- While these variables do not appear in the interactional diversity literature, they have
ken found to play a critical role in racial attitudes (Bobo and Hutchings 1996; Bonilla-

S i lVa 1997; Milem, Umbach, and Liang 2004; Powers and Ellison 1995; Safron and

Broman 1997).
Finally, class standing is considered for two reasons. First, students who have

liV'ed on campus for a longer time frame (i.e. upperclassmen) are likely to have a different
Impression of the campus climate and any issues of racial tension. For example, during

the 2005-06 school year, there were several incidents of racial harassment as well as

SeVeral highly publicized crimes that were committed on campus—allegedly by African

33

 

13a

’31:'.‘.‘
iii-«h‘

‘r'vfi

"it.

Hm

I x
7.9"?

Americans. Freshmen that were new to campus in 2006 would not have been aware of
these issues, at least not to the same extent as returning students, and there were very few
race-related incidents during the 2006-07 academic year. Second, the likelihood of
developing meaningful fiiendships with other students on campus should be directly
related to time spent on campus. Based on this assumption, meaningful interactions
across racial lines should also be related to this variable.
Participants were also asked to estimate the average number of hours they spend
socializing on a weekly basis. Response choices ranged from 0(0 hours) to 5(more than
20 hours). The purpose of this variable is to control for those students who are more or
less inclined to socialize with anyone, regardless of race (Saenz, Ngai, and Hurtado
2 007).
The racial composition of the respondents’ hometown, elementary school, and
middle/high school was also considered. For each question, response choices ranged fi'om
1 (all or nearly all students of color) to 5(all or nearly all White students). These
categories were then collapsed to create bivariate variables where all or nearly all
Students of color and all or nearly all White students represented a segregated population
and the remaining categories represented an integrated community. Research indicates
t111-alt experience with segregation early in life, particularly in educational settings, tends to
be perpetuated in later life (Braddock 1985; Braddock, Crain, and McPartland 1984;
Gurin 1999; Milem, Umbach, and Liang 2004).
Measures: Other Control Variables
In addition to the demographic variables, several other variables were considered

as POtential explanatory variables. Current hall percent White is a scale variable that

34

ranks the residence halls on the campus of Michigan State University by the percentage
of residents in each hall that are non-Hispanic White. In the current study, data was
originally collected from all 22 residence halls. Upon further consideration, however, one
of the halls has been eliminated from some of the analysis. Van Hoosen Hall consists of
64 residents; all of the residents are female and more than 95 percent of them are
upperclassmen. Additionally, this hall is set up as apartment-style living such that each
resident has their own room, bathroom, and kitchen. While the results from these
residents will be used in most of the analyses, they will not be included for those issues
that pertain specifically to the residence hall experience. The remaining 21 residence
halls were recoded into an ascending list based on the racial composition of each hall. For
example, the hall with the lowest percentage of White students, Rather (49.4% non-
Hispanic White), was given a value of l. The hall with the highest percentage of White
Students, Wonders (89.9% non-Hispanic White), was given a value of 21. By creating
Such a variable, the effects of the racial composition of the residence halls can be
Considered. The opportunity hypothesis tells us that as the racial diversity of an institution
increases, so does the likelihood for interacting across racial lines (Blau 1994; Joyner and
K210 2000). Given the exceptional size of Michigan State University, structural diversity
can be assessed at both the institutional and residence hall levels. Research finds that the
prObability of a social tie between individuals increases as their proximity increases
(Feldman 1981; McPherson, Smith-Lovin, and Cook 2001; Moody 2001). Related to this,
people expend minimal effort in establishing ties outside their local areas (Zipf 1949).

Taken together, these findings suggest that students living in racially diverse residence

35

 

. 4 u
A“ d
1.. .

.55

”fr,
1

.0. ...1

‘1 [‘1’ t.

‘3“
ell .1

"Mi

rill-it;

t1,“

halls will be more likely to interact cross-racially than students in predominantly White
residence halls.
Students were also asked how they came to live in their current residence hall.
Like many schools, Michigan State University allows students to request their residence
halls. Hall assignments are based on a number of variables including the application date,
the request of a specific roommate, and requests for a smoke-free or alcohol-free hall.
Returning students are given first preference in requests for a specific hall. All other
students are assigned to halls based on room availability and additional requests (e. g.
roommate, smoke-free hall, etc.). Students were first asked: Are you required to live in
your current residence hall due to your participation in a Living-Learning program such
as Lyman Briggs, RISE, ROSES, or James Madison? (1=Yes, 5=No). For those students
Who responded that they were not required to live in their current hall, they were then
asked: Did you request to live in your current residence hall? (1=Yes, 5=No). 14 percent
of the respondents in this study report that their current hall is the result of their
PartiCipation in a living-learning program. For the students who were not required to live
in a particular hall, 73 percent requested their current hall. It is expected that the
emI’el’iences of students within the residence halls will differ depending on whether they
ale Participating in a living-leaming program, where they have regular and ongoing
c>Qntact With other members of the programs; students who requested their current hall,
Where they have made a conscious choice to move; and students who did not request their
current hall (see Tables 2.5 and 2.6). When these variables are included in the analysis,

they Will not include the 28 participants from Van Hoosen Hall—all of whom requested

to live in that hall.

36

 

#

Table 2.5 Current Residence Hall Required,

 

 

 

N I Valid Percent I
Yes 560' 14.0
No 3453 86.0
Total 401 3 100.0
Missing 563
Total 45 76 I

 

 

 

 

Table 2.6 Current Residence Hall Requested

 

 

 

Frequency I Valid Percent I

Yes 2514I 73.0

o 930 27.0

Total 3444 100.0
Missing 1 132

Total 4576 I

 

 

 

An important factor expected to affect respondents’ experiences and perceptions
on campus is the atmosphere of their current residence hall. A nine-item scale was
developed, the Overall Racial Environment of the Current Residence Hall, to help
Capture the context in which each respondent is living (alpha=.778). Research suggests
“lat perceived racial conflict in the residence hall leads to a decrease in interaction across
ram lines as well as a decrease in students’ comfort (Johnson-Durgans 1994).
FUIthermore, research has shown the White and Black students are more likely to have

interracial friendships if they perceive such friendships as normative on campus or, in this

case, Within the residence hall (Chavous 2005). It is expected that students residing in
halls With minimal racial conflict, who would recommend their current hall to friends and
peers 0f their same racial background, will hold more positive views of the campus racial
cl imate, Perceive less racial tension on campus, and report higher levels of Interaction
tl'lan other students. Many of the questions in this scale were modeled afier those that

appear in the Michigan State University Department of Residence Life Floor Community
37

 

Survey which is conducted every year. As they appear in this survey, these questions have

been modified to address specific issues of race in the residence halls. The remaining

questions in this scale were developed by the researcher and are intended to reveal the

level of satisfaction or comfort that residents feel about their current residence hall.

The following items comprise the Overall Racial Environment of the Current

Residence Hall scale:

We would like you to consider the experiences you have had in your current
residence hall. F or the first set of examples, please tell us how often each of the following
examples occurs: never, occasionally, often, or very often.

I.

2.

3.

There are conflicts between groups on your floor and/or in your residence
hall related to racial/ethnic diflerences. (modified Res Life)

There are conflicts between individuals on your floor and/or in your residence
hall related to racial/ethnic diflkrences. (modified Res Life)

You have conflicts with someone on your floor and/or in your residence hall

that you believe are based on the difference in your race/ethnicity. (modified
Res Life)

Please tell us how accurately each statement reflects your experience in your
current residence hall. For each statement, tell us whether you strongly disagree,
dzsagree, agree, or strongly agree.

991.“

N

8.
9

It

I feel emotionally safe in my residence hall. (Res Life)

I feel physically safe in my residence hall. (Res Life)

I would recommend this residence hall to my friends of the same racial
background as myself

If I were to live on campus next year, I would want to live in the same
residence hall.

It has been easy to make fiiends in my current residence hall.

I am comfortable with the racial composition of my residence hall.

should be noted that residents living in Van Hoosen Hall are not included in this scale.

Several other variables were considered as potential control variables. First year

at MSU is a bivariate variable (1=yes) that allows the researcher to differentiate transfer

8 . . . .
tIJdents and freshmen from other upperclassmen 1n the event that there IS a relatronshrp

E: New time spent on campus and respondents’ perceptions and experiences on campus.

 

38

Three additional variables consider the effect of respondents’ roommates.
Research suggests that first-year roommates have a positive effect on interracial
friendships; however, the strength of this correlation has been inconsistent. In his
research at Duke University, Tyson (2004) found that freshmen were more likely to have

interracial fiiendships if their roommates were of a different racial background than
themselves. This relationship was found to be stronger among students of color than
White students. Among White students with an interracial roommate, Tyson found that
most were close friends with that roommate but had few other non-White friends. In their
longitudinal study conducted on the UCLA campus, Van Laar et al. (2005) found that
although interracial roommates showed lower levels of prejudice and higher levels of
affect toward other racial groups, these pairings did not increase the racial diversity of
Students’ friendships. In their longitudinal research on White college students,
McClelland and Linnander (2006) found that the effects of interracial contact were more
Significant when it occurred between friends—who choose each other on a voluntary

basis—than between close contacts who had been assigned to one another, as in the case

Of randomly assigned roommates.

In the current survey, respondents were asked if the racial background of their

first-year roommate differed from their own (1=yes). Respondents who have attended
MiChigan State University for more than one year were also asked about the race of their

cllrrent roommate (1=yes). Respondents were also asked whether they had requested their

first-year and current roommates (1=yes).

The Housing Department at Michigan State University tries to accommodate all

oommate requests. Because the majority of students at Michigan State Umversrty come

39

from within the state of Michigan—the most segregated state in the country—students

who request their first-year roommate are most likely to request a roommate of their same

racial background. Additionally, the level of familiarity would most likely differ

substantially between students who have a pre-established relationship with their

roommate and those that went in “blind”. The data indicate that the majority of students

(72.1%) who participated in the survey did not request their first-year roommate. Of

those students who did request their first-year roommate (32.7%), only 5 percent

involved a roommate of a different racial background than themselves (see Table 2.7).

Table 2.7 First-year roommate

 

 

 

   
 
 
 
 
 

 

 

N Valid Percent]
Requested 1st yr roommate/same race 862 22.3
Requested lst yr roommate/different race 214 5.5
Blind lst yr roommate/same race 1733 44.9
Blind lst yr roommate/different race l048 27.2
Total 3857 100.0
Missing 71 9
Total 4576'

 

 

 

 

Similar findings were reported by upperclassmen. For those students who

indicated that this was not their first year at Michigan State University (41.5% of

respondents), more than half requested their roommate (57%) and of those students, 11

percent requested a roommate of a different racial background than themselves (see Table

2 - 3). It is expected that the roommates of the respondents will be predictive of which

St‘lrdents would branch out of that comfort zone in an effort to build a new social network

Ql'lce coming to campus. For analysis pertaining to first-year and current roommates,

I‘Gasidents in Van Hoosen Hall will be included.

40

Table 2.8 Current Roommate Upperclass Only

 

 

 

 

 

N Percent Valid Percent
Did not request current roommate/ Same race 649 14.2 34.2
Requested current roommate/ Same race 865 18.9 45.5
Did not request current roommate/Different race 167 3.6 8.8
Requested current roommate/Different race 219 4.8 11.5
Total 1900 41.5 100.0]
Missing 2676 58.5
Total 4576 100.0I I

 

 

 

 

Research on interactional diversity has shown that the classroom plays an
important role in facilitating informal student engagement across race lines. The literature
related to the classroom can be divided into two components. First, enrollment in courses
that utilize a diversified curriculum (e. g. racial/ethnic studies; women’s studies) results in

students’ increased civic participation and cultural awareness, as well as critical drinking
Skills, and satisfaction with college (Astin 1993; Gurin 1999; Hurtado 2005; Milem
1 994). Second, when faculty employ an active pedagogy (e. g. providing students with
oIDPOrtunities for intergroup dialogue, collaborative learning methods, and group
projects), students show improved academic achievement, critical thinking and problem-
Solving skills, increased civic participation and cultural awareness, and most importantly
for the purposes of this study, increased interactional diversity outside of the classroom
(Chang, Astin, and Kim 2004; Cohen et al. 1997; Hurtado 2001; Slavin 1995; Slavin and
Cooper 1997; Wolfe and Spencer 1996).
A five-item scale was developed to assess the classroom environment
(alpha=.692). The Classroom Experience measures the extent to which students are
eXposed t0 issues related to diversity in the classroom and to what degree they are

afieaed by these opportunities. This measure is an attempt to operationalize the role that

41

 

the classroom plays in students’ interactional diversity. The particular items within the
scale were chosen because they ofien appear in the interactional diversity literature. F our
of the five questions were modified from similar studies. The fifih question, related to
who students sit next to in class, is based on the social distance literature. Michigan State
University frequently holds classes that exceed 500 students. In fact, many students do
not attend classes with less than 150 students until they reach their senior year and are
well into their programs. Under these circumstances, many students are able to find peers
of their own racial background to sit beside during lecture, if they so choose.

For each item of the Classroom Experience scale, responses ranged from 1(never)

to 4(very often).

For this section, please think about your daily experiences that occur outside of
your residence hall. For each of the following, please indicate how often you have
experienced each in your classes at MSU -never, occasionally, often, or very often.

I. I have experiences in my classes that challenge my values/beliefs. (Gurin, Dey,

Hurtado, and Gurin 2002)

2. I have experiences in my classes which broaden my understanding of students
difierent from myself.

3. My views have been changed because of conversations with students of
dijferent racial/ethnic backgrounds than myself in my classes. (Gurin, Dey,
Hurtado, and Gurin 2002)

4. I have conversations with students of a different racial background than myself
in my classes. (Chang, Astin, and Kim 2004)

5. I sit next to students of a dijferent racial background than myself in my classes.

Focus Groups
In addition to the survey, multiple focus groups were conducted. The intent of the
focus STOUps was twofold. First, to help provide further insight into particular patterns
that revealed themselves in the quantitative data. Second, to give voice to the students
and allow them to validate the findings of the survey. Focus group discussions centered

on the following topics: the perceived racial tension on campus, the perceived racial

42

ha-

.A‘

all!

hi1.

 

climate on campus, and the perceived relationship between housing and interactional
diversity. Focus groups were conducted after preliminary analysis of the data had been
conducted and lasted between sixty and ninety minutes.

Upon the completion of the online survey, respondents were invited to participate
in the focus groups that would be race-specific to the greatest extent possible. A link was
provided at the end of the survey that opened a separate window where volunteers
entered their race and gender. A list of available meeting times was posted and
participants were asked to rank the meeting times that best fit their schedules.

A total of five focus groups were conducted: two for Black students, one for

White students, one for Asian American students, and one for all other students: Latino,
Native American, Arab American and mixed-race students. As discussed earlier, students
were asked to self-identify their race when they volunteered for the focus groups. For the
most part, the racial demographics that were based on university records and students’
Self-identification were highly consistent. However, for the purpose of the focus groups,
whenever there was a discrepancy for a student, the student’s self-identification was used
(See Table 2.9). Of the 2961 White students who participated in the survey, 18.7%
(11:54) volunteered to participate in the focus groups. Of the 291 Black students who
participated in the survey, 46.0% (n=134) volunteered to participate in the focus groups.
For b0th White and Black students, volunteers were selected based on their
aVailability, such that those students who were available on the most popular days were
cOntacted. In the end, 100 self-identified White students and 99 self-identified Black
Students were contacted regarding the focus groups. Of the 105 Latino students who

partiCiPated in the study, 35.2% (n=37) volunteered to participate in the focus groups. Of

43

 

the 24 American Indian students who completed the survey, 33.3% (n=8) volunteered to
participate in the focus groups. Of the 226 Asian American/Pacific Islander students who
completed the survey, 25.2% (n=49) volunteered to participate in the focus groups. All of
the students who self-identified as Chicano/Latino, American Indian, and Asian
American/Pacific Islander volunteers were contacted and invited to attend focus groups
on the day that was most popular for each group. An additional 29 students that
volunteered for the focus groups were contacted. These students self-identified as either
mixed-race, Arab American, or whose race was unknown. Although 20.9% (n=3 7) of all
International respondents volunteered to participate in the focus groups, none responded

when they were contacted directly. As a result, there were no focus groups held for this

group.

44

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mew N nm N. ON av w mm .32 emm :88.
ca o Nm o N m o N - _ o 3:39:25
N_ o _ u o o o o o v coda/x 82
mm _ N o e 3 o o o o §Mm<
N. o o o a o e o o o .863
:dutug
mm o o o N o o mm o _ 2E3
2: o o o o o o o I“ a o :85:
goE<Ro£m
0mm x N o o o N o N 9% BE?
50% Due HGOUMEQEH voxm< 00mm 53¢. 55:: 05020 Quote: 833 main-com Dm—Z
«oz ~5me Swerve/w §o£<b~oflm no woman— and“
3§ eogaoezfim

 

 

ill

 

the 355325 .. ass: a: .5 BE ea: 3 as:

 

45

Participation rates for the focus groups were very low. In the end, only 3 African
Americans participated in the first meeting and 5 participated in the second meeting. Five
White students and 1 Asian American student participated in their respective focus group
meetings. Two students participated in the group designated for “Others”; both identify
as bi-racial (Native American and White). The atmosphere for the focus groups was
relaxed and participants appeared to be at ease discussing the topics with the facilitators.
It is possible that the dynamics for each discussion varied to a certain extent due to the
fact that the facilitators were not consistent across meetings. However, because the
participation rates were so low, it is unlikely that the outcomes of these discussions
would be generalizable to the campus population even if the facilitators had been
Consistent. Because the participation in focus groups was extremely low, making
Students’ remarks ungeneralizable to the campus population, results from these
discussions have not been included here. In order to provide a qualitative component to
this topic in future studies, participation in focus groups must be considerably higher.
Analytic Strategy

Data collected from the online survey is considered from multiple perspectives.
USing SPSS, a variety of analyses were conducted using crosstabulations, chi square, and
mmtiple regressions, including race, gender, socioeconomic status of the respondent’s
fEunily, racial composition of the respondent’s neighborhood and schools growing up,
percentage of White students in the current residence hall, and the Interaction scale. As
discussed previously, the effects of discrimination and the atmosphere of respondents’
reSidence halls were analyzed through the use of the Experience with Discrimination and

the Overall Racial Environment of the Current Residence Hall scales.

46

Gaining Access

The focus and intent of this dissertation was approved and supported by the
Oflice of Student Affairs, the Department of Residence Life, the Department of Housing,
and the Office for Inclusion and Intercultural Initiatives (formerly the Office of
Afiirmative Action Compliance and Monitoring) at Michigan State University. In
addition, the Residence Halls Association passed a bill specific to this study requiring all
of its members to help promote the significance of the survey and encourage student
participation.
Limitations: Data Collection

Data collection was conducted only once. Pettigrew (1998; 2000) warns against
measuring the effects of contact with cross-sectional analysis. Compared to longitudinal
Studies that measure racial attitudes and experiences in real time, cross-sectional studies
ask respondents to draw upon their own memories and increase the likelihood of social
desirability bias. Although questions in the present survey (e. g. racial composition of the
respondent’s elementary school and the diversity of the respondent’s friends prior to
attending Michigan State University) require participants to reflect on their personal
history, there is only one question that asks them about past attitudes (You have become
more aware of your personal prejudices and stereotypes towards others since attending
Michigan State University).

Cross-sectional analysis will also limit the analysis of the role of the classroom in
interactional diversity. Respondents were asked about both issues simultaneously so that
if a relationship is determined, causality cannot be determined. Despite this limitation,

though, finding a significant relationship is meaningful in that it suggests that at the very

47

least what occurs inside of the classroom affects what occurs outside of the classroom
and vice versa. Such a finding would indicate to researchers that this relationship
deserves further, longitudinal consideration.

There are a number of questions in the survey related to the respondent’s behavior
and level of interaction across racial lines. These questions are certainly susceptible to
social desirability effects. Items related to racial attitudes and behaviors are typically
vulnerable to social desirability bias, especially from White and highly educated
respondents (Krysan 2000; Sears 1997). Moreover, research on interracial friendships is
fraught with reporting bias (Bonilla-Silva 2003; Smith 2002). While many steps were
taken to include a variety of student and campus organizations when promoting this
Survey, it is possible that some participants continued to perceive the survey in a
Particular light. As a result, those students who participated in the study may have over-
stated their levels of interaction across race.

In his review of interracial friendship studies, Smith (2002) finds that one-item
Questions regarding interracial friendships, such as “Do you have any Black/White
ffiends,” are most likely to elicit response biases for both White and Black respondents.
According to Smith, these types of questions encourage participants to exaggerate
“friendly” relations with a particular group into “friendships” within that group. In

addition to this, without defining “fiiend” for a respondent, this concept can be
interpreted in any variety of ways. In the current survey, I have broadened the diversity
(e. g., race, sexual orientation, background, religious, culture, etc.) of my immediate
Circle of friends, is an example of the one-step question that Smith warns against,

Particularly when testing the contact hypothesis. The terms “friends,” despite being

48

‘

fill

D ."
.113

hi

ICU”
.
but

I __
Jim

233.}

narrowed to the immediate circle, does lend itself to open interpretation as Smith
suggests. Friendship questions related to race are especially difficult to operationalize
because of the tendency among Whites to inflate the number of non-White friends they
have (Bonilla-Silva 2003).

Despite the limitations of the diversity variable, there are two reasons why it
should not be discounted. First, this question appears regularly in the Residence Life
surveys that are distributed to on-campus residents. Returning students are therefore
farniliar with the question and have answered it in the past. In addition, there are previous
studies to compare the results from this survey to and confirm the reliability of this
measure. Second, although single-item friendship questions are prone to bias effects,
questions related to specific activities across race lines are not (Bonilla—Silva 2003). The
Interaction scale then provides us with a greater understanding regarding the validity of
this measure.

One-item questions in the Interactional Diversity scale are expected to yield a
minimum response bias. Some items taken individually, such as those relating to
s0cializing and partying, might be over-stated because like “friends” these activities can
be interpreted more loosely. However, other items in the scale, such as those pertaining to
having meaningful discussions and sharing personal feelings and problems, are typically

answered honestly (Bonilla-Silva 2003). By using a scale to measure the behaviors of
I’EBSpondents, the effects of social desirability should be minimized.

Other one-item measures within the study such as perceived racial tension on
campus and perceptions of the campus racial climate may be limited. Without providing

respondents with a range of specific questions that can be collapsed into scale

49

 

measurements for these variables, it cannot be assumed that participants interpreted these
questions similarly. In this case, the validity of these measures is threatened. If a
significant relationship is found between these variables and the independent variables,
this contributes to our understanding of race relations on a university campus. However,
the extent to which we can interpret these findings will be restricted.

Response error may also come as the result of self-selection. All research is
vulnerable to the effects of self-selection. This likelihood tends to increase in populations
that are already over-surveyed. While 33 percent of the on-campus population
participated in the study, it is important to consider how these respondents may differ
from their peers. Given the subject of this study, it is likely that for at least some of the
participants, this is a salient issue (Groves 2006; Heberlein and Bamngartner 1978; Van
Kenhove, Wijnen, and De Wulf 2002). Additionally, it is possible that some those
students who did not participate chose not to because of the specific issue at hand. In a
recent review of the literature on nonresponse rates and nonresponse bias, Groves (2006)
explains that nonresponse rates alone are a weak predictor of bias. A significant
advantage for this study is that it employed a nonprobability sample design, surveying the
entire population. Characteristics of those students who participated in this survey (race,
gender, class standing, and current hall) are very similar to the on-campus population at
large. An examination of those students who “clicked” the survey but did not participate
in it shows that these students differed somewhat by gender but the ratio of those who
completed the survey more accurately represented the campus population of men and
women. Non-Hispanic White students (per university records) were slightly

overrepresented among students who participated in the survey while students of color

50

were slightly underrepresented which is frequently found in survey research (Curtin,
Presser, and Singer 2000; Porter and Umbach 2006; Sax, Gilmartin, and Bryant 2003).
These differences were not substantial, however. This information suggests that any
effects of nonresponse are nominal, at least in terms of basic demographics.

Finally, there is the matter of missing data. Of the 4,576 participants, there are at
least 400 missing cases for each question. This number rises in a near linear fashion to
just over 700 missing cases by the end of the survey. It is unlikely that this increase is
related to a particular issue. The format of the survey is such that questions specifically
related to race on campus appear from the beginning; demographic data is not addressed
until the end. Compared to those that skipped the first question, the racial/ethnic
background of respondents who skipped the last issue-specific question (prior to the
questions relating to demographics) are fairly similar (see Table 2.10). lntemational
students and Afiican Americans exhibited the largest increase in incompletion rates (both
above 9%). Both of these groups exhibited higher rates of incompletion on the first
question as well. In contrast, American Indian/Native American were the only group to
improve their completion rates (a difference of one student). This suggests that for these
students, once they began the survey, they all completed it. It may be that the particular
topic of this survey was more salient to them than for students from other populations.
The overall proportion of those who completed the last question, however, is very similar
to the proportion that completed the first question. These proportions are also similar to
the overall on-campus population, with non-Hispanic Whites and International students
somewhat overrepresented and African Americans and Chicano/Latinos slightly

underrepresented.

51

Men were somewhat less likely to complete the last question than women (see
Table 2.11). This is consistent with survey response rates more generally (Curtin, Presser,
and Singer 2000; Porter and Umbach 2006; Sax, Gilmartin, and Bryant 2003). Compared
to those who skipped the first question, the rate of dropout is higher for men. When we
consider the completion rates for the last issue-specific question, women are
overrepresented.

Table 2.10 Race based on MSU Records * Skipped last question Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

Skipped last question
Answered Skipped last
Race based on MSU Records last question question Total
Non-Hispanic Count 3011 518 3529
White % within MSU race 85.3% 14.7% 100.0%
0 . . .
A’ “9““ ““91”“ last 79.2% 73.8% 78.3%
questron
Non-Hispanic Count 301 76 377
Black % within MSU race 79.8% 20.2% 100.0%
0 . . .
/" “”lh‘“ $1“de la“ 7.9% 10.8% 8.4%
questron
Count 108 17 125
Chicano/Latino % within MSU race 86.4% 13.6% 100.0%
0 . . .
A “"9“” Sk‘Pped la“ 2.8% 2.4% 2.8%
questron
American Count 23 6 29
Indian, Alaskan % within MSU race 79.3% 20.7% 100.0%
Native ' ' '
% wrthrn skipped last . 6% 9% . 6%
questron
Asian Count 228 41 269
American, % within MSU race 84.8% 15.2% 100.0%
Pacific Islander 0 ' ' '
A’ “”9““ “upped la“ 6.0% 5.8% 6.0%
questron
lntemational Count 133 44 177
% within MSU race 75.1% 24.9% 100.0%
% w‘lhm Sklpped la“ 3.5% 6.3% 3.9%
questron
Total Count 3804 702 4506
% within MSU race 84.4% 15.6% 100.0%
% “"9““ “‘1’?“ 1a“ 100.0% 100.0% 100.0%
questron

 

 

 

52

 

Table 2.11 Gender * Skipped Last Question Crosstabulation

 

Skipped last question

 

 

 

 

 

 

 

 

Answered last Skipped last

Gender question question Total

F Count 2537 408 2945
% within gender 86.1% 13.9% 100.0%
% within skipped last question 65.7% 57.1% 64.4%

M Count 1325 306 1631
% within gender 81.2% 18.8% 100.0%
% within skipped last question 34.3% 42.9% 35.6%

Total Count 3862 714 4576
% within gender 84.4% 15.6% 100.0%
% within skipped last question 100.0% 100.0% 100.0%

 

 

Although the survey completion time was estimated at 10 minutes, it is likely that

the increase in missing cases based on race is due, in large part, to participants dropping

out of the survey because of the time it required to complete. It is likely that the increase

in missing cases based on gender is partially the result of survey fatigue and the time it

took to complete the survey. However, it is also likely that gender itself played a role

with women being more likely to participate in surveys, having higher rates of

completion on surveys, and possibly also placing a higher value on the topic of this

survey than men. In the next chapter, we will review the results from the bivariate

analyses.

53

 

CHAPTER 3: PRELIMINARY RESULTS

Racial Tension on Campus

Students were asked whether they strongly disagreed, disagreed, agreed or
strongly agreed with the following statement: “There is racial tension on campus.” Of the
students who responded (n=3 876), 7.5 percent strongly agreed, 34.2 percent agreed, 47.3
percent disagreed, and 11.1 percent strongly disagreed (see Table 3.1). This indicates that
the majority of students on campus do not perceive racial tension, but that a solid
minority of students, 41.7 percent, does perceive such tension. The chi-square statistic on
this variable is large and significant (.000), indicating that we can reject the null
hypothesis that students on campus are equally likely to agree and disagree with this
statement (Table 3.2). When we examine this variable more closely, it becomes clear that
some students are more likely than others to perceive racial tension on campus.

Table 3.1 Racial Tension on Campus: Frequencies

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Strongly Agree 290 6.3 7.5 7.5
Agree 1325 29.0 34.2 41.7
Disagree 1832 40.0 47 .3 88.9
Strongly Disagree 429 9.4 11.1 100.0
Total 3876 84.7 100.0
Missing 700 l 5.3
Total 4576 100.0
Table 3.2 Racial Tension on Cam Jus: Chi-Square Test Frequencies

Observed N Expected N Residual

Strongly Agree 290 969.0 -679.0
Agree 1325 969.0 356.0
Disagree 1832 969.0 863.0
Strongly Disagree 429 969.0 -540.0
Total 3876

 

 

 

 

 

54

Racial Tension on Campus: Chi-Square Test Statistics

 

 

 

 

 

 

Racial Tension on Campus
Chi-Square(a) 1676.105
(if 3
Asymp. Sig. .000
Table 3.2 (cont’d)

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 969.0.

Racial Tension: Race/Ethnicig

Crosstabulations for racial tension by race suggest clear differences based on the
race/ethnicity of respondents (see Table 3.3). Black students were most likely to strongly
agree (35.1%) or agree (46.4%) that there is tension on campus, with nearly three-
quarters of Black students feeling this way. White/Caucasian students were the least
likely to perceive tension on campus. Nearly two-thirds of White students disagreed
(50.7%) or strongly disagreed (12.0%) with this statement. After White students, Asian
students, followed by Native American students, were the least likely to perceive racial
tension on campus (61.1% and 60.0% respectively). Mixed Race students appeared the
most conflicted on this issue with half of the students agreeing and half of the students
disagreeing that there is racial tension on campus. Chicano/Latino students were
somewhat divided in their perceptions; however, the majority report strongly agreeing
(14.5%) or agreeing (41.9%) with this statement.

The chi-square value on this crosstabulation is less than .0005, which leads us to
reject the null hypothesis that perceived racial tension on campus and the racial/ethnic
background of the participants are independent (see Table 3.4). This indicates that the

discrepancies seen in the frequencies between racial/ethnic groups are meaningful.

55

Table 3.3 Racial Tension on Campus*Self-Identified Race Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Race/Ethnicity
Racial Tension White/ Chicano/ Native Mixed
on Campus Caucasian Black Latino American Asian Race Total
Strongly Count 136 93 17 l 19 10 276
Agree % 4.6% 35.1% 14.5% 14.5% 6.6% 12.8% 7.4%
Agree Count 965 123 49 3 93 29 1262
% 32.7% 46.4% 41.9% 30.0% 32.3% 37.2% 34.0%
Disagree Count 1495 40 40 5 154 26 1760
% 50.7% 15.1% 34.2% 50.0% 53.5% 33.3% 47.5%
Strongly Count 354 9 11 1 22 13 410
Disagree % 12.0% 3.4% 9.4% 10.0% 7.6% 16.7% 11.1%
Total Count 2950 265 1 17 10 288 78 3708
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 3.4 Chi-Square Tests
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 428.156(a) 15 .000
Likelihood Ratio 326.594 15 .000
Linear-by-Linear Association 15.423 1 .020
N of Valid Cases 3708

 

 

 

 

 

 

a 4 cells (16.7%) have expected count less than 5. The minimum expected count is .74.

Racial Tension: Gender

Crosstabulations for racial tension by gender indicates that men and women share

fairly similar perceptions on this issue (see Table 3.5). Only 6.5 percent of men and 8.0

percent of women strongly agreed that there is racial tension on campus, while 31.2

percent of men and 35.8 percent of women agreed with this statement. Both groups were

most likely to report disagreeing with this statement: 48.0 percent of men and 46.9

percent of women disagreed that there is racial tension on campus. Men were somewhat

more likely to strongly disagree (14.4%) that there is racial tension on campus compared

with women (9.3%).

The chi-square value on this crosstabulation is less than .0005, which leads us to

reject the null hypothesis that perceived racial tension on campus and the gender of the

56

 

participants are independent (Table 3.6). This indicates that the discrepancies in
frequencies between men and women on this variable are meaningful.

Table 3.5 Racial Tension on Campus * Self-Identified Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gender
Racial Tension on Campus Male Female Total
Strongly Agree Count 85 201 286
% 6.5% 8.0% 7.5%
Agree Count 409 899 1 308
% 31.2% 35.8% 34.2%
Disagree Count 630 l 177 l 807
% 48.0% 46.9% 47.3%
Strongly Disagree Count 189 233 422
% 14.4% 9.3% 11.0%
Total Count 1313 2510 3823
% 100.0% 100.0% 100.0%
Table 3.6 Chi-Square Tests
Value df Asymp. Sig. (2-sided
Pearson Chi-Square 28.822(a) 3 .000
Likelihood Ratio 28.191 3 .000
Linear-by-Linear Association 23.459 1 .000
N of Valid Cases 3823

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 98.23.

Racial Tension: Class Standing

Crosstabulations for racial tension by class standing indicate that perceptions of
tension on campus increase with class standing in a linear fashion (see Table 3.7). Only
5 .2 percent of freshmen strongly agreed that there is racial tension on campus, compared
to 7.1 percent of sophomores, 12.0 percent of juniors, and 14.1 percent of seniors. Seniors
were most likely to agree that there is racial tension on campus (41.3%), compared to
28.7 percent of freshmen, 39.3 percent of sophomores, and 45.9 percent of juniors.
Freshmen were most likely to disagree that there is racial tension on campus (52.5%),

compared to 44.8 percent of sophomores, 42.6 percent of juniors and 36.9 percent of

57

seniors. Only 7.7 percent of seniors strongly disagreed that there is racial tension on

campus compared with 9.4 percent of juniors, 8.8 percent of sophomores, and 13.6

 

 

 

 

 

 

 

 

percent of freshmen.
Table 3.7 Racial Tension on Campus * Self-Identified Class Standing
Crosstabulation
Class Standing
Racial Tension on Campus Freshman Sophomore Junior Senior Total
Strongly Agree Count 91 86 63 44 284
% 5.2% 7.1% 12.0% 14.1% 7.5%
Agree Count 504 475 1 88 129 1296
28.7% 39.3% 35.9% 41.3% 34.1%
Disagree Count 924 541 223 1 1 5 1 803
52.5% 44.8% 42.6% 36.9% 47.4%
Strongly Disagree Count 240 106 49 24 419
% 13.6% 8.8% 9.4% 7.7% 11.0%
Total Count 1759 1208 523 312 3802
% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

The chi-square value on this crosstabulation is less than .0005, which leads us to

reject the null hypothesis that perceived racial tension on campus and the class standing

of the participants are independent (Table 3.8). This indicates that the discrepancies in

frequencies between freshmen, sophomores, juniors and seniors on this variable are

meaningful.

 

Table 3.8 Chi-Square Tests

 

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 1 18.214(a) 9 .000
Likelihood Ratio 114.164 9 .000
Linear-by—Linear Association 92.151 1 .000
N of Valid Cases 3802

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 23.31.
Racial Tension: First-Year Status
Students in their first year at Michigan State University were also examined to

account for possible transfer students among upper-classmen. Crosstabulations for racial

58

tension by first-year status were largely consistent with results for freshmen students (see
Table 3.9). Among first-year students 5.0 percent strongly agreed that there is racial

tension on campus, 28.5 percent agreed, 52.5 percent disagreed and 14.0 strongly

 

 

 

 

 

 

 

 

 

 

 

 

disagreed.
Table 3.9 Racial Tension on Campus * First Year at MSU Crosstabulation
First year at MSU

Racial Tension on Campus No Yes Total

Strongly Agree Count 187 97 284
% 10.0% 5.0% 7.5%

Agree Count 752 549 1301
% 40.1% 28.5% 34.2%

Disagree Count 786 1013 1799
% 41 .9% 52.5% 47.3%

Strongly Disagree Count 151 270 421
% 8.0% 14.0% 11.1%

Total Count 1 876 1929 3805
% 100.0% 100.0% 100.0%

 

 

The chi-square value on this crosstabulation is less than .0005, which leads us to
reject the null hypothesis that perceived racial tension on campus and the status of the
participants (first years versus retumers) are independent (see Table 3.10). This indicates
that the discrepancies in frequencies between first-year and returning students on this
variable are meaningful.

Table 3.10 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 121.761(a) 3 .000
Likelihood Ratio 122.904 3 .000
Linear-by-Linear Association 1 18.182 1 .000
N of Valid Cases 3805

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 140.02.

59

 

 

Racial Tension: Current Hall Percent White

Crosstabulations for racial tension by current hall were conducted. Results
suggest that perceptions of racial tension increase as the percentage of students of color in
the hall increases (see Table 3.11). Similarly, perceptions of racial tension tend to
decrease as the percentage of White students in the hall increase. For example, Rather,
Hubbard and Butterfield halls house the greatest percentage of students of color (50.6%,
48.0%, and 39.0% of their residents respectively). They also house the greatest
percentage of Black students (28.4%, 31.1%, 30.8%) of any halls on campus. On the
issue of racial tension, students in these halls are considerably more likely to agree or
strongly agree than students from other, less diverse halls. More than half of the students
living in these halls agree or strongly agree that there is racial tension on campus. In
contrast, more than 60 percent of students residing in Wonders, Mayo, and Wilson
halls—which have the greatest percentage of non-Hispanic White students (89.9%,
88.3%, and 88.0% respectively)—disagree or strongly disagree that there is racial tension

on campus.

Table 3.11 Current Hall Percent White" Racial Tension Bivariate Crosstabulation

 

 

 

 

 

 

 

 

 

Current Hall Percent Racial Tension Bivariate
White Strongly Agree or Agree Strongly Disagree or Disagree Total
Rather Count 57 46 103
% 55.3% 44.7% 100.0%
Hubbard Count 147 134 281
% 52.3% 47.7% 100.0%
Butterfield Count 47 35 82
% 57.3% 42.7% 100.0%
McDonel Count 92 177 269
% 34.2% 65.8% 100.0%
Bryan Count 54 78 132
% 40.9% 59.1% 100.0%

 

 

 

 

 

 

60

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Akers Count 133 162 295
% 45.1% 54.9% 100.0%

Shaw Count 90 1291 219
% 41.1% 58.9% 100.1;

Holden ICoum 105 145 250
% 42.0% 58.0% 100.0%
EBailey Count 34 64 98'
% 34.7% 65.3% 100.0%

Holmes Count 149 264 413
% 36.1% 63.9% 100.0%

Williams Count 35 3 1 66
% 53.0% 47.0% 100.0%

Emmons Count 40 66 106
% 37.7% 62.3% 100.0%

lMason-Abbot Count 124 139 263
% 47.1% 52.9% 100.0%

Armstrong Count 38 77 1 15
% 33.0% 67.0% 100.0%

Case Count 104 131 235
% 44.3% 55.7% 100.0%

Landon [Count 50 55 105
% 47.6% 52.4% 100.0%

Yakeley- Count 60 108 168
Eh‘hfi“ % 35.7% 64.3% 100.0%
lCampbell Count 32 53 85
% 37.6% 62.4% 100.0%
Wilson Count 74 145 219'
% 33.8% 66.2% 100.0%

Mayo lCount 19 36 55
% 34.5% 65.5% 100.0%

Wonders Count 81 129 210
% 38.6% 61 .4%| 100.0%

Total Count 1565 2204 3769
% 41.5% 58.5% 100.0%

Table 3.11 (cont’d) 7

61

 

 

The chi-square value on this crosstabulation is less than .0005, which leads us to
reject the null hypothesis that perceived racial tension on campus and the current halls of
the participants, based on the percentage of White students in those halls, are independent
(see Table 3.12). This indicates that the discrepancies in frequencies between students
living in predominantly White residence halls and those living in more diverse halls on

this variable are meaningful.

Table 3.12 Chi-Square Tests

 

 

Value ldf I Asymp. Sig. (2-sided) I
Pearson Chi-Square 67.961(a) 20 .000
Likelihood Ratio 67.819 20 .000
Linear-by-Linear Association 13.773 1 .000
N of Valid Cases 3769

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 22.84.

Further proof on the importance of the racial composition of the residence halls
can be found when we look at the one-way analysis of variance for the test of the null
hypothesis that perceptions of racial tension on campus will be equal across residence
halls (Table 3.13). The observed significance level for the comparison of means between

halls is less than .0005, suggesting that there is a relationship between these variables.

Table 3.13 Racial Tension Bivariate AN OVA f
Sum of Squares df Mean Squarel F Sig.

 

 

 

 

 

 

Between Groups 16.502 20 .825 3.441 .000
Within Groups 898.664 3748 .240
Total 915.166 3768

 

Racial Tension: Current Hall Assignment

Michigan State University has several living-learning programs on campus that
require its students to live in a particular hall. Additionally, student athletes are assigned

to specific residence halls. Consideration was given to whether perceptions of racial

62

tension varied depending on if students were required to live in the current hall. Students
who are not required to live in their current hall have the option of requesting their
current hall. The likelihood of this request being met is strongly related to whether or not
the students are “returning” students compared with first-year students. Crosstabulation
results show little difference between student opinions based on how their living
arrangements were made (see Table 3.14). The majority of students disagree or strongly
disagree that there is racial tension on campus, regardless of the nature of their housing

assignment.

Table 3.14 Hall Assignment without Van Hoosen * Racial Tension on Campus
Crosstabulation

 

 

 

 

 

 

Hall Assignment without Van Racial Tension on Campus
H003“ Strongly Strongly
Agree Agree Disagree Disagree Total
Hall Required Count 37 169 ml 56 532
% 7.0% 31.8% 50.8% 10.5% 100.0%
Requested Current Hall lCount 163 809 1126 257 2355
% 6.9% 34.4% 47.8% 10.9% 100.0%
Did not request current Count 82 304 384 105 875
Ha“ % 9.4% 34.7% 43.9% 12.0% 100.0%
Total Count 282 1282 1780 418 3762
% 7.5% 34.1% 47.3% 11.1% 100.0%

 

 

 

 

 

 

 

 

 

The chi-square value on this crosstabulation is not significant; therefore we
cannot reject the null hypothesis that perceived racial tension on campus and living in a
required residence hall are independent (Table 3.15). This indicates that the frequencies
on this variable are near enough to their expected count that, at least on the issue of

perceived racial tension, students living in required residence halls, residence halls they

63

requested, or halls in which they were randomly assigned do not differ substantially from

other students.

Table 3.15 Chi-Square Tests

 

 

Value | df Asymp. Sig. (2-sided)
Pearson Chi-Square 11.807(a) 6 .066
Likelihood Ratio 11.504 6 .074
Linear-by-Linear Association 2.528 1 .112
N of Valid Cases 3869

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 40.78.

Racial Tension: First-Year Roommate

Michigan State University allows its students to request other students as their
roommates and makes every attempt to fill these requests. We can assume that the vast
majority of students who requested each other for roommates were already friends. It
follows, then, that incoming freshmen and transfer students who requested a roommate,
most likely already knew those students prior to attending Michigan State University.
Students were asked whether or not they had requested their first-year roommate.
Respondents were also asked whether the racial background of their first-year roommate
was different from their own, regardless of whether or not they had requested that
roommate. Crosstabulations for racial tension by the status of first-roommates were
conducted (see Table 3.16).

Chi-square tabulations are less than .0005; therefore, we reject the null hypothesis
that perceived racial tension is independent from first-year roommate assignments (see
Table 3.17). Results suggest that students who requested their first-year roommate, and
whose roommate shared their same racial background, were the least likely to agree that

there is racial tension on campus. In contrast, students who requested their first-year

64

roommate, and whose roommate is of a different racial background than themselves,

were most likely to perceive racial tension on campus.

Table 3.16 First-Year Roommate * Racial Tension on Campus Crosstabulation

 

First-Year Roommate

Racial Tension on Campus

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Strongly Strongly
Agree Agree Disagree Disagree Total

Requested first-year Count 57 252 434 105 848
roommate/ same race % 6.7% 29.7% 51.2% 12.4% 100.0%
Requested first-year Count 24 77 85 25 211
mmate/diffmnt race % 11.4% 36.5% 40.3% 11.8% 100.0%
Blind first-year Count 80 608 828 199 1715
roommate/ same race % 4.7% 35.5% 48.3% 11.6% 100.0%
Blind first-year Count 122 367 454 92 1035
j’oommate/differem race % 11.8% 35.5% 43.9% 8.9% 100.0%
Total Count 283 1304 1801 421 3809
% 7.4% 34.2% 47.3% 11.1% 100.0%

Table 3.17 Chi-Square Tests
Value dfl Asymp. Sig. (2-sided) |
Pearson Chi-Square 69.816(a) 9 .000
Likelihood Ratio 68.469 9 .000
Linear-by-Linear Association 6.557 1 .010

of Valid Cases 3809

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 15.68.

Racial Tension: Current Roommate (Uggerclassmen)

Consideration was also given to the role that the current roommate plays on the

perception of racial tension on campus among returning students. Crosstabulation results

suggest that the current roommate of upperclassmen is related to perceived racial tension

(see Table 3.18). Students who requested their current roommate, regardless of the

roommate’s racial background, were the least likely to agree with this statement.

65

Upperclassmen who requested their current roommate, and whose roommate shares their

same racial background, were the most likely to disagree or strongly disagree that there is

racial tension on campus (46.1%). Upperclassmen who requested their current roommate,

and whose roommate is of a different racial background were somewhat more likely to

agree with this statement. Upperclassmen who did not request their current roommate,

and whose roommate is of a different racial background were the most likely to agree or

strongly agree that there is racial tension on campus.

Table 3.18 Current Roommate (upperclassmen only) * Racial Tension on Campus

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Current Roommate Racial Tension on Campus
(upperclassmen only) Strongly Strongly
Agree Agree Disagree Disagree Total
Blind/Same race Count 80 264 245 50' 639
% 12.5% 41.3% 38.3% 7.8% 100.0%
Requested/Same race Count 66 324 389I 76 855
% 7.7% 37.9% 45.5% 8.9% 100.0%
Blind/Different race Count 26 70I 57 13 166
% 15.7% 42.2% 34.3% 7.8% 100.0%
Requested/Different race Count 15 94 95 12 216
% 6.9% 43.5% 44.0% 5.6% 100.0%
Total Count 187 752 786 151 1876
% 10.0% 40.1% 41 .9% 8.0% 100.0%

 

 

 

 

 

 

 

 

 

The chi—square value on this crosstabulation is .less than .005, indicating that we

cannot reject the null hypothesis that perceived racial tension on campus and the current

roommate of upperclassmen are independent (Table 3.19)

66

 

 

Table 3.19 Chi-Square Tests

 

 

Value Idf I Asymp. Sig. (2-sided)
Pearson Chi-Square 27.501(a) 9 .001
Likelihood Ratio 27.291 9 .001
Linear-by-Linear Association .381 1 .537
N of Valid Cases 1876

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 13.36.

Racial Tension: Overall Racial Environment

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like to perceive racial tension on campus regardless of the overall
racial environment of their current residence hall. The observed significance level for this
test is less than .0005 indicating that we should reject the null hypothesis (see Table
3.20). The relationship between these two variables seems only natural: students who
observe racial tension and conflict within their residence halls should be more likely than

other students to perceive such tension on campus.

Table 3.20 Racial Tension on Campus ANOVA
Sum of Squaresl df Mean Square F Sig.|

 

 

Between Groups 170.321 47 3.624 6.365 .000
Within Groups 2118.638 3721 .569
Total 2288.959 3768

 

 

 

 

 

 

 

Racial Tension: Classroom Experience

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like to perceive racial tension on campus regardless of their
classroom experience. The observed significance level for this test is less than .0005

indicating that we should reject the null hypothesis (see Table 3.21).

67

Table 3.21 Racial Tension on Campus ANOVA

 

Sum of Squares df Mean Square F Sig.

 

 

 

 

 

 

 

 

Between Groups 57.321 22 2.606 4.375 .000
Within Groups 2294.611 3853 .596
Total 2351.932 3875

 

Current Racial Climate

Students were asked to rate the Current racial climate on campus as poor, fair,
good, or excellent. Of the students who responded (n=3899), 5.8 percent viewed the
current racial climate as poor, 28.7 percent perceived the climate as fair, 55.6 percent
perceived the climate as good, and 9.9 percent perceived the current racial climate as
excellent (see Table 3.22). This indicates that the majority of students on campus view
the current racial climate positively, while a sizeable minority (34.5%) of students view
the current racial climate negatively. When we examine this variable more closely, it
becomes clear that some students are more likely than others to view the current racial
climate as fair or poor. The chi-square statistic on this variable is large and significant
(.000), indicating that we can reject the null hypothesis that students on campus are
equally likely to perceive the current racial climate excellent, good, fair, and poor (see
Table 3.23).

Table 3.22 Current Racial Climate

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Poor 228 5.0 5.8 5.8
Fair 1119 24.5 28.7 34.5
Good 2167 47.4 55.6 90.1
Excellent 385 8.4 9.9 100.0
Total 3899 85.2 100.0
Missing 677 14.8
Total 4576 100.0

 

 

 

 

 

 

 

68

 

Table 3.23 Current Racial Climate: Chi-Square Tests Frequencies

 

 

Observed N Expected N Residual
Poor 228 974.8 -746.8
Fair 1119 974.8 144.3
Good 2167 974.8 1192.3
Excellent 385 974.8 -589.8
Total 3899

 

 

 

 

 

 

Current Racial Climate: Chi-Square Test Statistics

 

 

Current Racial Climate
Chi-Square(a) 2408.524
(If 3
Asymp. Sig. .000

 

 

 

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 974.8.

Racial Climate: Race/Ethnicig

Crosstabulations for racial climate by race indicate that students perceptions do
vary by racial/ethnic background (see Table 3.24). Native American students view the
current racial climate in the most favorable light. None of these students described the
current climate as poor, while 30.0 percent reported the current climate is fair. The
majority of Native American students (70.0%) viewed the current racial climate as good,
although none described it as excellent. Following Native Americans, White/Caucasian
students were the least likely group to view the climate negatively. Only 3.1 percent of
White students perceive the racial climate on campus as poor. In contrast, 10.6 percent
viewed the climate as excellent. A solid majority of White students, 60.5 percent, report
that the campus racial climate is good. Finally, 25.7 percent of White students view the
racial climate as fair. Black students were the most likely to describe the current racial
climate on campus negatively with 32.1 percent of all Black students reporting that the
current climate is poor and an additional 47.0 percent reporting the climate as fair. While

18.3 percent of Black students view the current racial climate as good, only 2.6 percent

69

view it as excellent. As was the case for perceived racial tension on campus, Mixed Race

students were divided in their perceptions of the current racial climate with half

describing it as poor or fair and half describing it as good or excellent. Chicano/Latino

students hold a somewhat more positive perception of the current racial climate: 12.8

percent reported the climate as poor, 35.9 percent as fair, 44.4 percent as good, and 6.8

percent as excellent. Asian students were more likely to view the racial climate on

campus favorably than other non-White groups but less likely than Whites. Only 4.8

percent reported that the climate is poor and an additional 37.2 percent described the

climate as fair. Approximately half of all Asian students described the current racial

climate as good (50.2%) but only 7.8 percent perceived the climate as excellent.

Table 3.24 Current Racial Climate * Self-Identified Race Crosstabulation

 

 

 

 

 

 

 

 

Race/Ethnicity
Current Racial White/ Chicano/ Native Mixed
Climate Caucasian Black Latino American Asian Race T0131
Poor Count 93 86 1 5 0 14 12 220
% 3.1% 32.1% 12.8% .0% 4.8% 15.4% 5.9%
Fair Count 762 126 42 3 109 27 1068
% 25.7% 47.0% 35.9% 30.0% 37.2% 34.6% 28.7%
Good Count 1794 49 52 7 147 26 2075
% 60.5% 18.3% 44.4% 70.0% 50.2% 33.3% 55.6%
Excellent Count 3 15 7 8 0 23 13 366
% 10.6% 2.6% 6.8% .0% 7.8% 16.7% 9.8%
Total Count 2964 268 1 17 10 293 78 3730
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

 

 

 

The chi-square value on this crosstabulation is less than .0005, which leads us to

reject the null hypothesis that perceived racial climate on campus and the racial/ethnic

background of the participants are independent (see Table 3.25). This indicates that the

discrepancies seen in the frequencies between racial/ethnic groups on this variable are

meaningful.

70

Table 3.25 Chi-Square Tests

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 538.296(a) 15 .000
Likelihood Ratio 405.713 15 .000
Linear-by-Linear
Association 44.774 1 .000
N of Valid Cases 3730

 

 

 

 

 

 

a 4 cells (16.7%) have expected count less than 5. The minimum expected count is .59.

Racial Climate: Gender

Crosstabulations for racial climate by gender suggests that women are somewhat
more likely to view the climate negatively (see Table 3.26). Only 5.3 percent of men and
6.2 percent of women described the current racial climate on campus as poor. However,
31.0 percent of women compared to 24.2 percent of men viewed the climate as fair. Both
groups were most likely to report that the current climate is good: 55.9 percent of men
and 55.4 percent of women. Men were more likely to rate the current racial climate as
excellent (14.6%) than women (7.4%).

Table 3.26 Current Racial Climate * Self-Identified Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

Gender
Current Racial Climate Male Female Total
Poor Count 70 156 226
% 5.3% 6.2% 5.9%
Fair Count 319 784 1 103
% 24.2% 31.0% 28.7%
Good Count 736 1400 21 36
% 55.9% 55.4% 55.6%
Excellent Count 192 188 380
% 14.6% 7.4% 9.9%
Total Count 1317 2528 3845
% 100.0% 100.0% 100.0%

 

 

 

The chi-square value on this crosstabulation is less than .0005, which leads us to

reject the null hypothesis that perceived racial climate on campus and the gender of the

71

participants are independent (see Table 3.27). This indicates that the discrepancies seen in
the frequencies between men and women on this variable are meaningful.

Table 3.27 Chi-Square Tests

 

 

Value df Asymp. Sig (2-sided)
Pearson Chi-Square 55.728(a) 3 .000
Likelihood Ratio 57.797 3 .000
Linear-by-Linear Association 40.171 1 .000
N of Valid Cases 3845

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 77.41.

Racial Climate: Class Standing

Crosstabulations for racial climate by class standing indicate that perceptions of
campus climate are more negative for upperclassmen (see Table 3.28). Only 4.0 percent
of freshmen rate the current racial climate on campus as poor, compared to 6.0 percent of
sophomores, 10.7 percent of juniors, and 8.0 percent of seniors. A similar pattern can be
found among those students who described the racial climate as fair: 24.8 percent of
freshmen, 31.2 percent of sophomores, and 29.8 percent of juniors, and 38.7 percent of
seniors. Among all groups, respondents were most likely to perceive the current racial
climate as good: 59.7 percent of freshmen, 53.6 percent of sophomores, 51.7 percent of
juniors, and 47.0 percent of seniors. Only 6.4 percent of seniors described the racial
climate on campus as excellent compared with 7.8 percent of juniors, 9.2 percent of
sophomores, and 11.5 percent of freshmen.

The chi-square value on this crosstabulation is less than .0005, which leads us to
reject the null hypothesis that perceived racial climate on campus and the class standing
of the participants are independent (see Table 3.29). This indicates that the discrepancies
seen in the frequencies between freshmen, sophomores, juniors, and seniors on this

variable are meaningful.

72

Table 3.28 Current Racial Climate * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Class Standing
Current Racial Climate Freshman Sophomore Junior Senior Total
Poor Count 71 73 56 25 225
% 4.0% 6.0% 10.7% 8.0% 5.9%
Fair Count 439 380 156 121 1096
% 24.8% 31.2% 29.8% 38.7% 28.7%
Good Count 1056 652 271 147 2126
% 59.7% 53.6% 51.7% 47.0% 55.6%
Excellent Count 204 l 12 41 20 377
% 11.5% 9.2% 7.8% 6.4% 9.9%
Total Count 1770 1217 524 313 3824
% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 3.29 Chi-Square Tests
Value df Asymg Sig. (2-sided)
Pearson Chi-Square 79.905(a) 9 .000
Likelihood Ratio 76.870 9 .000
Linear-by-Linear Association 64.21 1 1 .000
N of Valid Cases 3824

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 18.42.

Racial Climate: First-Year Status

Crosstabulations for racial climate by first-year status were largely consistent with
results for freshmen students (see Table 3.30). Among first-year students only 3.7 percent

described the racial climate on campus as poor, 25.0 percent as fair, 59.4 percent as good,

and 11.9 percent as excellent.

The chi-square value on this crosstabulation is less than .0005, which leads us to
reject the null hypothesis that perceived racial climate on campus and the status of
participants (first-year versus returning) are independent (see Table 3.31). This indicates

that the discrepancies seen in the frequencies between first-year and returning students on

this variable are meaningful.

73

 

Table 3.30 Current Racial Climate * First Year at MSU Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

First Year at MSU

Current Racial Climate No Yes Total
Poor Count 153 72 225

% 8.1% 3.7% 5.9%
Fair Count 613 485 1098

% 32.5% 25.0% 28.7%
Good Count 974 1 152 2126

% 51.6% 59.4% 55.5%
Excellent Count 148 23 1 379

% 7.8% 11.9% 9.9%
Total Count 1 888 1940 3828

% 100.0% 100.0% 100.0%

Table 3.31 Chi-Stulare Tests
Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 76.469(a) 3 .000
Likelihood Ratio 77.320 3 .000
Linear-by-Linear Association 74.859 1 .000
N of Valid Cases 3828

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 110.97.

Racial Climate: Current Hall Percent White

Crosstabulations for racial climate by current hall indicate that the racial climate
is typically viewed more negatively as the percentage of students of color in the hall
increases (see Table 3.32). Similarly, perceptions of the racial climate tend to improve as
the percentage of White students in the hall increase. Students in the most racially
diverse halls are more likely to describe the current climate as poor or fair when
compared with students from other, less diverse halls. For example, 44.1 percent of
students living in Rather reported that the climate on campus is poor or fair compared

with only 34.6 percent of students living in Wonders. Similarly, nearly two-thirds of the

74

students residing in Wonders described the racial climate on campus as good or excellent

compared to 55.9 percent of students in Rather.

Table 3.32 Current Hall Percent White * Current Racial Climate Bivariate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Current Hall Percent White Current Racial Climate Bivariate
Poor or Fair Good or Excellent Total
Rather Count 45 57 102
% 44.1% 55.9% 100.0%
Hubbard Count 121 161 282
% 42.9% 57.1% 100.0%
Butterfield Count 42 41 83
% 50.6% 49.4% 100.0%
McDonel Count 86 186 272
% 31.6% 68.4% 100.0%
Bryan Count 44 91 135
% 32.6% 67.4% 100.0%
Akers Count 110 188 298
% 36.9% 63.1% 100.0%
Shaw Count 8 1 140 22 1
% 36.7% 63.3% 100.0%
Holden Count 80 172 252
% 31.7% 68.3% 100.0%
Bailey Count 22 76 98
% 22.4% 77.6% 100.0%
Holmes Count 113 304 417
% 27.1% 72.9% 100.0%
Williams Count 30 36 66
% 45.5% 54.5% 100.0%
Emmons Count 36 70 106
% 34.0% 66.0% 100.0%
Mason-Abbot Count 102 163 265
% 38.5% 61.5% 100.0%
Amstrong Count 28 87 1 15
% 24.3% 75.7% 100.0%
Case Count 82 153 235
% 34.9% 65.1% 100.0%
Landon Count 35 70 105
% 33.3% 66.7% 100.0%

 

 

 

 

 

 

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Yakeley- Count 59 108 167
6mm“ % 35.3% 64.7% 100.0%
Campbell Count 33 52 85
% 38.8% 61.2% 100.0%
Wilson Count 67 157 224
% 29.9% 70.1% 100.0%
Mayo Count 16 39 55
% 29.1% 70.9% 100.0%
Wonders Count 72 l 36 208
% 34.6% 65.4% 100.0%
Total Count 1304 2487 3791
% 34.4% 65.6% 100.0%
Table 3.32 (cont’d)

The chi-square value on this crosstabulation is less than .05, which leads us to
reject the null hypothesis that perceived racial climate on campus and the current halls of
the participants, based on the percentage of White students in those halls, are independent
(see Table 3.33). This indicates that the discrepancies in frequencies between students
living in predominantly White residence halls and those living in more diverse halls on

this variable are meaningful.

Table 3.33 Chi-Square Tests

 

 

Value I df I Asymp. Sig. (2-sided) I
Pearson Chi-Square 56.540(a) 20 .000
Likelihood Ratio 56.634 20 .000
Linear-by-Linear Association 6.111 1 .013
N of Valid Cases 3791

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 18.92.
Further proof on the importance of the racial composition of the residence halls

can be found when we look at the one-way analysis of variance for the test of the null

hypothesis that perceptions of the racial climate on campus will be equal across residence

halls (see Table 3.34). The observed significance level for the comparison of means

76

between halls is less than .0005, suggesting that there is a relationship between these

variables.

Table 3.34 Current Racial Climate Bivariate ANOVA

 

 

 

 

 

 

Sum of Squares df IMean Square F Sig.I
Between Groups 12.758 20 .638 2.854 .000
Within Groups 842.701 3770 .224
Total 855.460I3790

 

 

Racial Climate: Current Hall Assignment

Crosstabulations of racial climate and residence hall assignment indicate that
those students who were required to live in their current hall due to participation in a
living-learning program or athletics hold a more favorable view of the current racial
climate on campus (see Table 3.35). Among students living in a required hall, 58.4
percent describe the racial climate as good and an additional 12.5 percent as excellent.
Students not required to live in their current hall—whether they requested the hall or

not—reported very similar perceptions of the campus racial climate.

Table 3.35 Hall Assignment without Van Hoosen * Current Racial Climate
Crosstabulation

 

 

 

 

 

 

 

 

 

 

Hall Assignment without Van Current Racial Climate 7
Hoosen Poor Fair Good Excellent I Total
Hall Required Count 22 133 312 67' 534
% 4.1% 24.9% 58.4% 12.5% 100.0%.
Requested Current Hall'Count 145 690- 1317 217 2369I
% 6.1% 29.1% 55.6% 9.2% 100.0%
Did not request Current Count 56 257 475 93 881
Ha“ % 6.4% 29.2% 53.9% 10.6% 100.0%
Total Count 223 1080 2104 377 3784
% 5.9% 28.5% 55.6% 10.0% 100.0%

 

 

 

 

 

 

77

Chi-square values for this crosstabulation are less than .05 which leads us to reject
the null hypothesis that perceived racial climate is independent from the campus housing

assignments (see Table 3.36).

Table 3.36 Chi-Square Tests

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 12.897(a) 6 .045
Likelihood Ratio 13.088 6 .042
Linear-by-Linear Association 6.027 1 .014
N of Valid Cases 3784

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 31.47.

Racial Climate: First-Year Roommate

Crosstabulations of the current racial climate and the first-year roommates suggest
there is some difference in the perceptions of current racial climate (see Table 3.37).
Two-thirds of all students described the campus racial climate as good or excellent.
However, students whose first-year roommate is of a different racial background from
themselves were more likely to describe the campus climate as poor or fair. Students who
requested their first-year roommate, and whose roommate is of a different racial
background were the most likely to describe the climate as poor (9.4%). Students who did
not request their first-year roommate, regardless of that roommates racial background,
were the most likely to describe the racial climate as excellent (10.1%). Students who
share the same racial background as their first-year roommate were the most likely to
describe the campus racial climate favorably.

Chi-square values for this crosstabulation are less than .0005 which leads us to
reject the null hypothesis that perceived racial climate is independent from the first-year

roommate (see Table 3.38).

78

Table 3.37 First-year Roommate * Current Racial Climate Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

First-year Roommate Current Racial Climate
Poor Fair IGood Excellent Total
Requested/Same race ICount 46 230| 496 83 855
% 5.4% 26.9% 58.0% 9.7% 100.0%
Requested/Different race Count 20l 66 109 18 213
% 9.4% 31.0% 51.2% 8.5% 100.0%
Blind/Same race Count 67 466 1015 173 1721
% 3.9% 27.1% 59.0% 10.1% 100.0%
Blind/Different race Count 92 336 510 105 1043
% 8.8% 32.2% 48.9% 10.1% 100.0%
Total Count 225 1098 2130 379 3832
% 5.9% 28.7% 55.6% 9.9% 100.0%
Table 3.38 Chi-Square Tests
Value Idf I Asymp. Sig. (2-sided)
Pearson Chi-Square 53.355“l 9
Likelihood Ratio 52.322 9
Linear-by-Linear Association .274 l
N of Valid Cases 3832

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 12.51.

Racial Climate: Current Roommate (Qpperclassmen)

Consideration was also given to the role that the current roommate plays on the

perception of campus racial climate among returning students. Crosstabulation results

suggest that the current roommate of upperclassmen is related to perceptions of the

current racial climate (see Table 3.39). Upperclassmen who requested their current
roommate, and who shared the same racial background as their roommates, were the
most likely to describe the campus racial climate as good or excellent. Upperclassmen

who requested their current roommates—regardless of racial background—were less

likely than those who went in blind—to perceive the racial climate negatively. The

79

students who were least likely to describe the campus climate positively are

upperclassmen with a current roommate of a different racial background from their own

that they did not request to live with.

Table 3.39 Current Roommate (U pperclassmen only) * Current Racial Climate

 

 

 

 

 

 

 

 

Crosstabulation

Fluent Roommate (U pperclassmen Current Racial Climate

only) Poor I Fair Good Excellent Total

Blind/Same race Count 60 223 306 54 643
% 9.3% 34.7% 47.6% 8.4% 100.0%

Requested/Same race Count 60' 2501 485 65 8601
% 7.0% 29.1% 56.4% 7.6% 100.0%

Blind/Different race Count 21 64 68 14 167
% 12.6% 38.3% 40.7% 8.4% 100.0%

Requested/Different race Count 12 76 115 15 218
% 5.5% 34.9% 52.8% 6.9% 100.0%

Total Count 153 613 974 148 1888
% 8.1% 32.5% 51 .6% 7.8% 100.0%

 

 

 

 

 

 

 

 

Chi-square values for this crosstabulation are less than .005 which leads us to

reject the null hypothesis that perceived racial climate is independent from the first-year

roommate (see Table 3.40).

Table 3.40 Chi-Square Tests

 

 

 

 

 

Value de Asymp. Sig. (2-sided)
Pearson Chi-Square 25.046(a) 9 .003
Likelihood Ratio 24.921 9 .003
Linear-by-Linear Association .121 1 .728
N of Valid Cases 1888

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 13.09.

80

Racial Climate: Overall Racial Environment

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like to perceive the racial climate on campus positively or
negatively regardless of the overall racial environment of their current residence hall (see
Table 3.41). The observed significance level for this test is less than .0005 indicating that
we should reject the null hypothesis. The relationship between these two variables seems
only natural: students who observe racial tension and conflict within their residence halls

should perceive the racial climate in a different light than students who do not.

Table 3.41 Current Racial Climate ANOVA -
Sum of Squares df Mean Square F Sig.I

 

 

Between Groups 213.792 47 4.549 9.513 .000
Within Groups 1788.293 3740 .478
Total 2002.084 3787

 

 

 

 

 

 

Current Racial Climate: Classroom Experience

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like to perceive the racial climate on campus positively or
negatively regardless of their experiences within the classroom (see Table 3.42). The
observed significance level for this test is less than .0005 indicating that we should reject
the null hypothesis.

Table 3.42 Current Racial Climate ANOVA

 

Sum of Squares df Mean Square F Sig.

 

 

 

 

 

 

 

 

Between Groups 18.649 22 .848 1.615 .035
Within Groups 2033.580 3874 .525
Total 2052.228 3896

 

81

Broadened Diversity of Friends

Students were asked whether they strongly disagreed, disagreed, agreed, or
strongly agreed that they had Broadened the diversity (race, sexual orientation,
background, religious, culture, etc.) of their immediate circle of friends since coming to
Michigan State University. Of the students who responded (n=3 871), 5.5 percent strongly
disagreed, 25.3 percent disagreed, 54.8 percent agreed, and 14.4 strongly agreed (Table
3.43). This indicates that the majority of students on campus feel they have diversified
their fiiendships since coming to Michigan State University, but a third of students have
done so only minimally, if at all.

Table 3.43 Broaden Diversity of Friends

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly Disagree 212 4.6 5.5 5.5
Disagree 980 21.4 25.3 30.8
Agree 2122 46.4 54.8 85.6
Strongly Agree 557 12.2 14.4 100.0
Total 3871 84.6 100.0
Missing System 705 15.4
Total 4576 100.0

 

 

 

 

 

 

 

The chi-square statistic on this variable is large and significant (.000), indicating
that we can reject the null hypothesis that students on campus are equally likely to agree
/and disagree with this statement (Table 3.44). When we examine this variable more
closely, it becomes clear that some students are more likely to have diversified their
circle of friends than others.

Table 3.44 Broaden Diversity Friends: Chi-Square Test Frequencies

 

 

Observed N Expected N Residual
Strongly Disagree 212 967.8 -755.8
Disagree 980 967.8 12.3
Agree 2122 967.8 1 154.3
Strongly Agree 557 967.8 -410.8
Total 3871

 

 

 

 

 

 

82

Broaden Diversity Friends: Chi-Square Test Statistics

 

 

 

 

 

 

Broaden Diversity Friends

Chi-Square(a) 2141 .376

df 3

. Asmp. Sig. .000
Table 3.44 (cont’d)

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 967.8.

Broadened Diversig: Race/Ethnicity

Crosstabulations for broadened diversity by race indicate that the majority of
students agree or strongly agree that they have diversified their friendships since coming
to Michigan State University (see Table 3.45). More than two-thirds of
Whites/Caucasian and Black students agree or strongly agree with this statement. More
than three-quarters of Latino and Mixed Race students agree or strongly agree with this
statement. Native American students, followed by Asian students, were more likely than
other racial/ethnic groups to strongly agree with having diversified their friendships.
Compared with all other groups, Black and White students were most likely to disagree
or strongly disagree with this statement.

Table 3.45 Broaden Diversity of Friends * Self-Identified Race Crosstabulation

 

 

 

 

 

 

 

 

 

Race/Ethnicity
Broaden Diversity of White/ Native Mixed
Friends Caucasian Black Latino American Asian Race Total
Strongly Count 172 22 4 0 6 0 204
Disagree % 5.8% 8.3% 3.4% .0% 2.1% .0% 5.5%
Disagree Count 778 66 16 1 60 16 937
% 26.4% 24.9% 13.6% 10.0% 20.9% 20.8% 25.3%
Agree Count 1620 125 71 6 167 40 2029
% 54.9% 47.2% 60.2% 60.0% 58.2% 51.9% 54.7%
Strongly Count 380 52 27 3 54 21 537
A8"“ % 12.9% 19.6% 22.9% 30.0% 18.8% 27.3% 14.5%
Total Count 2950 265 l 18 10 287 77 3707
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

 

83

 

 

 

The chi-square value on this crosstabulation is less than .0005, which leads us to
reject the null hypothesis that broadening the diversity of friends and the racial/ethnic
background of the participants are independent (Table 3.46). This indicates that the
discrepancies seen in the frequencies between racial/ethnic groups on this variable should
be examined more closely.

Table 3.46 Chi-Square Tests

 

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 61 .788(a) 15 .000
Likelihood Ratio 66.279 15 .000
Linear-by-Linear Association 34.657 1 .000
N of Valid Cases 3707

 

a 4 cells (16.7%) have expected count less than 5. The minimum expected count is .5

Broadened Diversity: Gender

Crosstabulations for broadened diversity by gender suggests that men and women
share similar experiences (see Table 3.47). Only 6.0 percent of men and 5.2 percent of
women strongly disagreed that with having diversified their friendships, while 25.1
percent of men and 25.5 percent of women disagreed with this statement. Both groups
were most likely to agree with this statement: 54.9 percent of men and 54.7 percent of
women agreed that they had diversified their friendships since coming to Michigan State
University. Finally, 13.9 percent of men and 14.6 percent of women strongly agreed with
this statement.

The chi-square value on this crosstabulation is greater than .05 (see Table 3.48).
Therefore, we cannot reject the null hypothesis that broadening the diversity of friends
and the gender of the participants are independent. This indicates that the discrepancy

seen in the frequencies between men and women on this variable is not significant.

84

Table 3.47 Broaden Diversity of Friends * Self-Identified Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gender

Broaden Diversity of Friends Male Female Total
Strongly Disagree Count 79 130 209

% 6.0% 5.2% 5.5%
Disagree Count 328 640 968

% 25.1% 25.5% 25.4%
Agree Count 717 1374 2091

% 54.9% 54.7% 54.8%
Strongly Agree Count 182 368 550

% 13.9% 14.6% 14.4%
Total Count 1306 2512 3818

% 100.0% 100.0% 100.0%

Table 3.48 Chi-Square Tests
Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 1.554(3) 3 .670
Likelihood Ratio 1.537 3 .674
Linear-by-Linear Association .666 1 .415
N of Valid Cases 3818

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 71.49.

Broadened Diversity: Class Standing

Crosstabulations for broadened diversity by class standing indicate a less
consistent pattern than was found with racial tension and the current racial climate (Table
3.49). Only 6.2 percent of freshmen strongly disagree with having diversified their
fi'iendships, compared to 4.6 percent of sophomores, 5.0 percent of juniors, and 6.1
percent of seniors. Freshmen were more likely to disagree with this statement (26.9%)
than other groups: 25.0 percent of sophomores, 24.9 percent of juniors and 17.9 percent
of seniors disagreed. Among all groups, respondents were most likely to agree that they
have diversified their circle of friends since coming to Michigan State University: 54.9
percent of freshmen, 56.1 percent of sophomores, 54.2 percent of juniors, and 50.5

percent of seniors. Finally, seniors were more likely to strongly agree (25.6%) with

85

having diversified their circle of friends than other groups: 15.9 percent of juniors, 14.2

percent of sophomores, and 12.0 percent of freshmen strongly agreed with this statement.

Table 3.49 Broaden Diversity of Friends * Class Standinerosstabulation

 

 

 

 

 

 

 

 

Class Standig
Broaden Diversity 01' Friends Freshman Sophomore Junior Senior Total
Strongly Disagree Count 108 56 26 19 209
% 6.2% 4.6% 5.0% 6.1% 5.5%
Disagree Count 472 303 1 30 56 961
% 26.9% 25.0% 24.9% 17.9% 25.3%
Agree Count 961 679 283 158 2081
% 54.9% 56.1% 54.2% 50.5% 54.8%
Strongly Agree Count 211 172 83 80 546
% 12.0% 14.2% 15.9% 25.6% 14.4%
Total Count 1752 1210 522 313 3797
% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

 

The chi—square value on this crosstabulation is less than .0005 (see Table 3.50).
Therefore, we can reject the null hypothesis that broadening the diversity of friends and
class standing of the participants are independent. This indicates that the discrepancy
seen in the frequencies between freshmen, sophomores, juniors, and seniors on this
variable are significant need to be explored further. lntuitively, this should make sense.
The longer students are on a college campus, the greater their opportunities to meet and
interact with students of diverse backgrounds. Results from the chi-square tests suggest
that this relationship holds true.

Table 3.50 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 48.430(a) 9 .000
Likelihood Ratio 44.456 9 .000
Linear-by-Linear Association 24.234 1 .000
N of Valid Cases 3797

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 17.23.

86

Broadened Diversity: First-Year Status

Crosstabulations for broadened diversity by first-year status were largely
consistent with results for fieshmen students (Table 3.51). Among first-year students 6.0
percent strongly disagreed with having diversified their friendships, 27.6 percent
disagreed, 54.5 percent agreed, and 11.9 percent strongly agreed.

Table 3.51 Broaden Diversity of Friends * First Year at MSU Crosstabulation

 

 

 

 

 

 

 

First Year at MSU
Broaden Diversity of Friends Yes No Total
Strongly Disagree Count 116 94 210
% 6.0% 5.0% 5.5%
Disagree Count 531 433 964
% 27.6% 23.1% 25.4%
Agree Count 1048 1030 2078
% 54.5% 54.9% 54.7%
Strongly Agree Count 228 320 548
% 11.9% 17.0% 14.4%
Total Count 1923 1877 3800
% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

Chi-square tabulations for first-year students are similar to those of class standing
with a value less than .0005 (Table 3.52). This provides further evidence that time spent
on campus is an important predictor of diversifying friendships.

Table 3.52 Chi-Square Tests

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 27.316(a) 3 .000
Likelihood Ratio 27.407 3 .000
Linear-by-Linear Association 23,] 16 .000
N of Valid Cases 3800

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 103.73.

Broadened Diversity: Current Hall Percent White

Crosstabulations for broadened diversity by current hall does not reveal any

obvious patterns (Table 3.53). There are clear differences between residence halls, but it

87

is not immediately apparent what these can be attributed to, even when the variable is

collapsed into bivariate categories. Williams Hall has the greatest number of residents of

agree or strongly agree that they have diversified their friendships since attending

Michigan State University (84.6%). Of the 21 halls included in this analysis, Williams

has the highest percentage of upperclassmen with freshmen constituting just 7 percent of

their residents. Holmes Hall had the second highest rate of agreement with this statement

(77.1%). As discussed earlier, Holmes Hall is the largest living-learning program on

campus.

Table 3.53 Current Hall Percent White * Broaden Diversity of Friends

Bivariate Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Broaden Diversity of Broaden Diversity of Friends Bivariate
Friends Strongly Disagree or Disagree I Strongly Agree or Agree Total
Rather Count 29 74 103
I % 28.2% 71.8% 100.0%
Hubbard Count 80| 201 281
% 28.5% 71.5% 100.0%
Butterfield Count 26 56 82
% 31.7% 68.3% 100.0%
IMcDonel Count 81 191 272
% 29.8% 70.2% 100.0%
Bryan ICount 42 90F 132
% 31.8% 68.2% 100.0%
Akers Count 78 218 296
% 26.4% 73.6% 100.0%
Shaw Count 68 l 5 1 219
% 31.1% 68.9% 100.0%
Holden Count 82 165 247
I % 33.2% 66.8% 100.0%
Bailey Count 23 76 99
% 23.2% 76.8% 100.0%I
Holmes Count 94 316 410
% 22.9% 77.1% 100.0%
Williams Count 10 55 65
I % 15.4% 84.6% 100.0%I

 

 

 

 

88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Emmons Count 37 69 106
I% 34.9% 65.1% 100.0%

Mason-Abbot Count 76 1 89 265
% 28.7% 71.3% 100.0%

Armstrong Count 49 65 1 l4
% 43.0% 57.0% 100.0%

Case Count 77 158 235
i I% 32.8% 67.2% 100.0%
Landon 1Count 33 72 105
% 3 1.4% 68.6% 100.0%

Yakeley- Count 74 93 167
[Gilchrist % 44.3% 55.7% 100.0%.
[Campbell Count 34 51 85
% 40.0% 60.0% 100.0%I
Wilson |Count 76 143 219I
% 34.7% 65.3% 100.0%

Mayo Count 22 3 1 53
% 41.5% 58.5% 100.0%

Wonders Count 7 O 1 39 209
% 33.5% 66.5% 100.0%I

Total Count 1 161 2603 3764
% 30.8% 69.2% 100.0%
Table 3.53 (cont’d)

The chi-square value on this crosstabulation is less than .0005 (see Table 3.54).
Therefore, we can reject the null hypothesis that broadening the diversity of friends and

current residence hall are independent.

Table 3.54 Chi-Square Tests

 

 

Value I (if Asymp. Sig. (2-sided) I
Pearson Chi-Square 59.074“ 20 .000
Likelihood Ratio 59.337 20 .000
Linear-by-Linear Association 13.657 1 .000
N of Valid Cases 3764

 

 

 

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 16.35.
Results from the one-way analysis of variance indicate that the racial composition

of the residence hall is an important determinant for diversifying friendships (see Table

89

3.55). When we test the null hypothesis that the likelihood of broadening the diversity of
friends will be equal across residence halls, the observed significance level for the
comparison of means between balls is less than .0005, suggesting that there is a

significant relationship between these two variables.

Table 3.55 Broaden Diversity of Friends Bivariate ANOVA

 

Sum of SquaresI df IMean Square F Sig.

 

 

Between Groups 12.601 20 .630 2.984 .000
Within Groups 790.290 3743 .211
Total 802.891 3763

 

 

 

 

Broadened Diversity: Current Hall Assiggment

Crosstabulations of broadened diversity and required hall suggest that those
students who were required to live in their current hall are more likely to report having
diversified their circle of friends (Table 3.56). Among students participating in a living-
learning program, 75.5 percent agreed or strongly agreed with this statement. In contrast,
68.5 percent of students who requested their current hall, and 67.4 percent of those
randomly assigned to their current hall agreed or strongly agreed.

The chi-square values for this crosstabulation are less than .05 indicating that we
should reject the null hypothesis that broadening diversity of friends is independent from

campus housing assignments (Table 3.57).

90

Table 3.56 Hall Assignment without Van Hoosen * Broaden Diversity of Friends
Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

Hall Assignment without Broaden Diversity Friends
Van Hoosen Strongly Strongly
Disagree Disagree Agree Agree Total
Hall required Count 24 106 305 96 531
% 4.5% 20.0% 57.4% 18.1% 100.0%
Requested current hall Count 127 614 1291 316 2348
% 5.4% 26.1% 55.0% 13.5% 100.0%
Did not request ICount 57 229 464 128 878
[current ha“ % 6.5% 26.1%I 52.8% 14.6%I 100.0%
Total Count 208 949 2060 540 3757
% 5.5% 25.3%I 54.8% 14.4%I 100.0%

 

 

 

 

 

Table 3.57 Chi-Square Tests

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square l7.145(a) 6 .009
Likelihood Ratio 17.227 6 .008
Linear-by-Linear Association 9.433 1 .002
N of Valid Cases 3757

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 29.40.

Broadened Diversity: First-Year Roommate

Crosstabulations of broadened diversity of fiiends and students’ first-year
roommates indicate similar experiences for students on this variable (see Table 3.58).
Students whose first-year roommates is of their same racial background, whether they
had requested that roommate or not, were equally likely to disagree or strongly disagree
that they had diversified their friendships since attending Michigan State University.
Students who had requested their first-year roommate, and whose roommate is of a

different racial background, were the most likely to strongly agree with this statement.

91

Table 3.58 First-year Roommate * Broaden Diversity of Friends Crosstabulation

 

First-year Roommate

Broaden Diversity of Friends

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

|
Strongly Strongly I
Disagree Disagree Agree Agree Total

Requested/Same race Count 51 236 4801 83 850
% 6.0% 27.8% 56.5% 9.8% 100. 0%I
Requested/Different Count 7 41 115 47 210
race % 3.3% 19.5% 54.8% 22.4% 100. 0%I
Blind/Same race Count 103 464 906 236 1709
% 6.0%I 27.2% 53.0% 13.8% 100. 0%I
Blind/Different race Count 49 223 582 181 1035
% 4.7%I 21.5% 56.2% 17.5% 100. 0%I

Total Count 210 964 2083 547 3804
% 5.5% 25.3% 54.8% 14.4% 100. 0%

 

The chi-square value on this crosstabulation is less than .0005 indicating that we

should reject the null hypothesis that the likelihood of students’ broadening the diversity

of their friends and their first-year roommates are independent (Table 3.59).

Table 3.59 Chi-Square Tests

 

 

 

 

 

Value df I Asymp. Sig. (2-sided) I
Pearson Chi-Square 48.087(a) 9 .000
Likelihood Ratio 48.667 9 .000
Linear-by-Linear Association 3.935 1 .047
N of Valid Cases 3804

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 11.59.

Broadened Diversi

: Current Roommate

erclassmen

Consideration was also given to the role that the current roommate plays on the

likelihood that returning students feel they have broadened the diversity of their

friendships. Crosstabulation results suggest that the current roommate of upperclassmen

is related to broadening diversity (see Table 3.60). The majority of upperclassmen agree

92

or strongly agree that they have broadened the diversity of their friendships; however,

those students who requested their current roommate, and whose roommate is of their

same racial background, are somewhat less likely to agree or strongly agree with this

statement than other upperclassmen. While 68.4 percent of these students agree or

strongly agree that they have diversified their friendships, three-quarters of all other

upperclassmen felt this way. Upperclassmen currently living with a roommate of a

different racial background, that did not request this roommate, are the most likely to

strongly agree with this statement (26.2%).

Table 3.60 Current Roommate (U pperclassmen Only) * Broaden Diversity of
Friends Crosstabulation

 

Current Roommate

Broaden Diversity Friends

 

 

 

 

 

 

 

 

 

 

 

 

 

(Upperclassmen Only) Strongly Strongly
Disagree Disagree Agree Agree Total

Blind/Same race Count 38 127 347 1291 641

% 5.9% 19.8% 54.1% 20.1% 100.0%
Requested/Same Count 41 230! 475 111 857
race % 4.8% 26.8% 55.4% 13.0% 100.0%
Blind/Different race Count 7 33 81 43 164

% 4.3% 20.1% 49.4% 26.2% 100.0%
Requested/Different Count 8 43 127 37 215
race % 3.7% 20.0% 59.1% 17.2% 100.0%
Total Count 94 433 1030 320 1877

% 5.0% 23.1% 54.9% 17.0% 100.0%

 

 

The chi-square value on this crosstabulation is less than .0005 indicating that we

should reject the null hypothesis that the likelihood of upperclassmen broadening the

diversity of their friends and their current roommates are independent (Table 3.61).

93

Table 3.61 Chi-Square Tests

 

 

 

 

Value Idf Asymp. Sig. (2-sided) I
Pearson Chi-Square 33.566(a) 9 .000
Likelihood Ratio 32.974 9 .000
Linear-by-Linear Association .356 1 .551
N of Valid Cases 1877

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 8.21.

Broadened Diversig: Overall Racial Environment

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like to report broadening the diversity of their friends, regardless
of the overall racial environment of their current residence hall. The observed
significance level for this test is less than .0005 indicating that we should reject the null

hypothesis (see Table 3.62).

Table 3.62 Broaden Diversity of Friends ANOVA
Sum of Squaresl df Mean Square F Sig.I

 

 

Between Groups 73.485 47 1.564 2.804 .000
Within Groups 2071.812 3716 .558
Total 2145.297 3763

 

 

 

 

 

 

 

Broadened Diversity: Classroom Exgerience

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally likely to report broadening the diversity of their friends,
regardless of their experiences within the classroom. The observed significance level for
this test is less than .0005 indicating that we should reject the null hypothesis (see Table

3.63).

94

Table 3.63 Broadened Diversity of Friends ANOVA

 

Sum of Squares df Mean Square F Sig.

 

 

 

 

 

 

 

 

Between Groups 284.840 22 12.947 26.019 .000
Within Groups 1914.831 3848 .498
Total 2199.671 3870

 

Aware of Personal Prejudices

Students were asked whether they strongly disagreed, disagreed, agreed, or
strongly agreed that they had become more aware of their personal prejudices and
stereotypes towards others since coming to Michigan State University. Of the students
who responded (n=3 874), 7.8 percent strongly disagreed, 22.6 percent disagreed, 58.5
percent agreed, and 11.1 strongly agreed (see Table 3.64). This indicates that the majority
of students on campus feel they have become more aware of their prejudices since
coming to Michigan State University, but nearly a third of students do not feel this way.
When we examine this variable more closely, it becomes clear that some students are

more likely to report having become more aware of their personal prejudices and

 

 

 

stereotypes.

Table 3.64 Aware Personal Prejudices

Aware of Preflldices Frequency Percent Valid Percent Cumulative Percent
Strongly Disagree 301 6.6 7.8 7.8
Disagree 875 19.1 22.6 30.4
Agree 2268 49.6 58.5 88.9
Strongly Agree 430 9.4 11.1 100.0
Total 3874 84.7 100.0

Missing System 702 15.3

Total 4576 100.0

 

 

 

 

 

 

The chi-square statistic on this variable is large and significant (.000), indicating
that we can reject the null hypothesis that students on campus are equally likely to agree

or disagree with this statement (Table 3.65).

95

Table 3.65 Aware Personal Prejudices: Chi-Square Test Frequencies

 

 

 

Observed N Expgcted N Residual
Strongly Disagree 301 968.5 -667.5
Disagree 875 968.5 -93.5
Agree 2268 968.5 1299.5
Strongly Agree 430 968.5 -538.5
Total 3874

 

 

 

 

 

Aware Personal Prejudices: Chi-Square Test Statistics

 

Aware Personal Prejudices

 

 

 

 

Chi-Square(a) 25 12. 1 1 3
df 3
Asymp. Sig. .000

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 968.5.

Aware of Prejudices: Race/Ethnicity

Crosstabulations for more aware of prejudices by race/ethnicity indicate that
Black and Native American students are the most likely to agree and strongly agree with
this experience (see Table 3.66). More than one-fifth of Black respondents strongly
agreed that they have become more aware of their personal prejudices and stereotypes
since coming to Michigan State University and an additional 59.2 percent agreed. In
contrast, just 13.1 percent of Black students disagreed with this statement and only 6.4
percent strongly disagreed. Ten percent of Native American students strongly agreed with
this statement; however 70 percent agreed. After African Americans, Chicano/Latino
students are the most likely to strongly agree with having become more aware of their
personal prejudices (13.9%). An additional 60 percent agreed with this statement.
Responses from Asian and White students were comparable with a few exceptions.
Compared with Whites, Asian students were more likely to strongly agree with having

becoming more aware of personal prejudices and stereotypes: 10.8 percent of Asian

96

students strongly agreed compared with 10.3 percent of non-Hispanic Whites. Similarly,

Whites students were more to strongly disagree with having become more aware of

personal prejudices (8.0%) than Asian students (4.9%). Mixed Race students were the

most likely to strongly disagree (9.0%) and the least likely to strongly agree (7.7%) with

this statement than any of the other racial/ethnic groups.

Table 3.66 Aware Personal Prejudices * Self-Identified Race Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Race/Ethnicifl
Aware Personal White/ Native Mixed
ml udices Caucasian Black Latino American Asian Race Total
511mg" CW“ 237 17 7 0 14 7 282
Disagree
% 8.0% 6.4% 6.1% .0% 4.9% 9.0% 7.6%
Disagree Count 680 35 23 2 69 24 833
% 23.1% 13.1% 20.0% 20.0% 24.0% 30.8% 22.5%
Agree Count 1727 158 69 7 174 41 2 176
% 58.6% 59.2% 60.0% 70.0% 60.4% 52.6% 58.7%
Strongly Agree Count 304 57 16 l 3 1 6 415
% 10.3% 21.3% 13.9% 10.0% 10.8% 7.7% 11.2%
Total Count 2948 267 1 15 10 288 78 3706
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

 

Chi-square values for this crosstabulation are less than .0005 indicating that we

can reject the null hypothesis that becoming aware of personal prejudices and stereotypes

is independent from the racial/ethnic background of the participants (Table 3.67).

Table 3.67 Chi-Square Tests: Racial/Ethnic background

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 57.583(a) 15 .000
Likelihood Ratio 58.143 15 .000
Linear-by-Linear Association .167 1 .683
N of Valid Cases 3813

 

 

a 2 cells (8.3%) have expected count less than 5. The minimum expected count is 1.81.

Aware of Prejudices: Gender

Crosstabulations for aware of prejudices by gender indicates that women are

somewhat more likely than men to believe they become more aware of their personal

97

 

prejudices and stereotypes toward others since coming to Michigan State University
(Table 3.68). Men are more likely to strongly disagree with this statement than women
(9.9% and 6.6% respectively). While women were more likely to agree with this
statement than men (60.4% and 55.1% respectively), both groups were most likely to
choose this response.

Table 3.68 Aware Personal Prejudices * Self-Identified Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

Gender
Aware Personal Prejudices Male Female Total
Strongly Disagree Count 129 167 296
% 9.9% 6.6% 7.7%
Disagree Count 3 17 546 863
% 24.3% 21.7% 22.6%
Agree Count 719 1519 2238
% 55.1% 60.4% 58.6%
Strongly Agree Count 140 284 424
% 10.7% 11.3% 11.1%
Total Count 1305 2516 3821
% 100.0% 100.0% 100.0%

 

 

The chi-square value for this crosstabulation is less than .0005 which means that
we can reject the null hypothesis that gender of the participant is independent from
becoming aware of personal prejudices since coming to Michigan State University (Table
3.69). This suggests that the difference in reporting between men and women on this
variable is significant and this relationship should be explored further.

Table 3.69 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided
Pearson Chi-Square 18.580(a) 3 .000
Likelihood Ratio 18.174 3 .000
Linear-by-Linear Association 13.925 1 .000
N of Valid Cases 3821

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 101.09.

98

Aware of Prejudices: Class Standing

Crosstabulations for aware of prejudice by class standing suggest that as class
standing increases, so does the likelihood of becoming aware of personal prejudices (see
Table 3.70). For all groups, the majority of students agreed that they have become more
aware of their prejudices. Freshmen were the most likely to strongly disagree with this

experience (8.8%) and the least likely to strongly agree (9.1%). Seniors were the most

likely to strongly agree (14.1%).

Table 3.70 Aware Personal Prejudices * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

Class Standing
Aware Personal Prejudices Freshman Sophomore Junior Senior Total
Strongly Disagree Count 155 76 42 22 295
% 8.8% 6.3% 8.1% 7.1% 7.8%
Disagree Count 467 230 99 60 856
% 26.6% 19.0% 19.0% 19.2% 22.5%
Agree Count 975 754 312 186 2227
% 55.5% 62.3% 59.9% 59.6% 58.6%
Strongly Agree Count 160 150 68 44 422
% 9.1% 12.4% 13.1% 14.1% 11.1%
Total Count 1757 1210 521 312 3800
% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

The chi-square value for this crosstabulation is less than .0005 so we should reject
the null hypothesis that becoming more aware of personal prejudice and class standing
are independent (Table 3.71). As was the case for broadening the diversity of friends, the
relationship seen here also makes sense intuitively. As time on campus increases for
students, the opportunities for exposure to new information, both socially and in the

classroom, increases.

99

Table 3.71 Chi-Square Tests

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 48.447(a) 9 .000
Likelihood Ratio 48.547 9 .000
Linear-by-Linear Association 23.938 1 .000
N of Valid Cases 3800

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 24.22.

Aware of Prejudices: First-Year Status

Crosstabulations for aware of prejudices by first-year status were largely
consistent with results for freshmen students (see Table 3.72). Among first-year students
8.9 percent strongly disagreed with having become more aware of personal prejudices
since coming to Michigan State University, 26.6 percent disagreed, 55.6 percent agreed,
and 8.9 percent strongly agreed.

Table 3.72 Aware Personal Prejudices * First Year at MSU Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

First year at MSU
Aware Personal Prejudices No Yes Total
Strongly Disagree Count 125 171 296
% 6.7% 8.9% 7 .8%
Disagree Count 346 512 858
% 18.4% 26.6% 22.6%
Agree Count 1 154 1072 2226
% 61.5% 55.6% 58.5%
Strongly Agree Count 251 172 423
% 13.4% 8.9% 11.1%
Total Count 1876 1927 3803
% 100.0% 100.0% 100.0%

 

 

Similar to class standing, the chi-square value for this crosstabulation is less than
.0005 indicating that we can reject the null hypothesis that the status of students, as first-
year compared with returning students, and becoming aware of personal prejudices are

independent (Table 3.73).

100

Table 3.73 Chi-Square Tests

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 56.366(a) 3 .000
Likelihood Ratio 56.676 3 .000
Linear-by-Linear Association 47.753 1 .000
N of Valid Cases 3803

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 146.02.

Aware of Prejudices: Current Hall Percent White

Crosstabulations for more aware of personal prejudices by current hall percent
White does not reveal any clear patterns. However, when the variable is collapsed into
bivariate categories (agree or disagree) differences between residence halls begin to
emerge (Table 3.74). Those residence halls with the greatest diversity, and the greatest
percentage of Black residents, tend to show higher rates of agreement than those balls
with fewer students of color. For example, 80.4 percent of students residing in Rather
Hall and 71.8 percent of students living in Hubbard Hall strongly agree or agree that they
have become more aware of their personal prejudices and stereotypes towards other since
coming to Michigan State University. In contrast, 65.1 percent of students in Wonders
Hall and 61.8 percent of students in Mayo Hall strongly agree or agree with this
statement. Regardless of current hall, however, most students agree or strongly agree

with this statement.

Table 3.74 Current Hall Percent White * Aware of Personal Prejudices Bivariate
Crosstabulation

 

 

 

 

 

 

 

Current Hall Percent White Aware of Personal Prejudices Bivariate
Strongly Disagree or Disagree Strongly Agree or Agree Total
Rather Count 20 82 102
% 19.6% 80.4% 100.0%
Hubbard Count 79 201 280
% 28.2% 71.8% 100.0%

 

 

 

 

 

101

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Butterfield ICount 20I 60I 80
% 25.0% 75.0% 100.0%

McDonel Count 86 184 270|
% 31.9% 68.1% 100.0%

Bryan Count 43 901 133
% 32.3% 67.7% 100.0%

Akers Count 69 227 296
% 23.3% 76.7% 100.0%

Shaw Count 66 152 218
% 30.3% 69.7% 100.0%

IHolden Count 67 l 82 249
% 26.9% 73.1% 100.0%
Bailey Count 33 66 99'
% 33.3% 66.7% 100.0%

Holmes Count 149 265 414
% 36.0% 64.0% 100.0%

Williams Count 14 52 66
% 21.2% 78.8% 100.0%

Emmons Count 38 66 104
% 36.5% 63.5% 100.0%

Mason-Abbot Count 63 200 263
% 24.0% 76.0% 100.0%

Armstrong Count 37 78 1 15
% 32.2% 67.8% 100.0%

ICase Count 78 158 236
% 33.1% 66.9% 100.0%

Landon Count 34 70I 104
% 32.7% 67.3% 100.0%

Yakeley-Gilchrist Count 61 107 168
% 36.3% 63.7% 100.0%

ICampbell Count 27 58 85
% 3 1.8% 68.2% 100.0%

Wilson Count 68 153 221
% 30.8% 69.2% 100.0%

Mayo Count 2 1 34 55
% 38.2% 61.8% 100.0%

Wonders Count 73 1 36 209
% 34.9% 65.1% 100.0%I

Total Count 1 146 2621 3767
% 30.4% 69.6% 100.0%

Table 3.74 (cont’d) i

102

The chi-square value for this crosstabulation is less than .05 indicating that we can
reject the null hypothesis that students are equally likely to report becoming aware of

personal prejudices regardless of their current hall (Table 3.75).

Table 3.75 Chi-Square Tests

 

 

Value I df I Asymp. Sig. (2-sided)
Pearson Chi-Square 40. 145(a)1 20 .005
Likelihood Ratio 41.035 20 .004
Linear-by-Linear Association 8.693 1 .003
N of Valid Cases 3767

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 16.73.

Results from the one-way analysis of variance indicate that the residence hall is
an important determinant for becoming aware of personal prejudices (Table 3.76). When
we test the null hypothesis that the likelihood of agreeing with this statement will be
equal across halls, the observed significance level for the comparison of means between
halls is less than .05, suggesting that there is a significant relationship between these two

variables.

Table 3.76 Aware Prejudices Bivariate ANOVA
Sum of Squares df Mean Square F Sig.

 

 

Between Groups 8.498 20 .425 2.018 .005
Within Groups 788.865 3746 .211
Total 797.363 3766

 

 

 

 

 

 

 

 

Aware of Prejudices: Current Hall Assiggment

Crosstabulations of aware of prejudice and current hall assignment indicate
similar reports from students (Table 3.77). Students who did not request their current hall
were somewhat more likely to agree or strongly agree that they have become more aware

of their personal prejudices since attending Michigan State University (70.8%). Students

103

participating in living-learning programs were the least likely to agree or strongly agree

with this statement (65.0%). However, the majority of all students report becoming more

aware of their personal prejudices.

Table 3.77 Hall Assignment without Van Hoosen * Aware Personal Prejudices

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Hall Assignment without Van Aware Personal Prejudices
Hoosen Strongly Strongly
Disagree Disagree Agree Agree Total
Hall required Count 44 143 286 60 533
% 8.3% 26.8% 53.7% 11.3% 100.0%
Requested current hall Count 183 519 1403 2491 2354
% 7.8% 22.0% 59.6% 10.6% 100.0%
Did not request current Count 66 189' 507 111 873
|hall % 7.6% 21.6% 58.1% 12.7% 100.0%
Total Count 293 851 2196 420 3760
% 7.8% 22.6% 58.4% 11.2% 100.0%I

 

 

 

 

 

 

 

 

The chi—square value for this crosstabulation is greater than .05 indicating that we

cannot reject the null hypothesis that becoming more aware of personal prejudices and

campus housing assignments are independent (see Table 3.78).

Table 3.78 Chi-Square Tests

 

 

 

 

 

Value Idf Asymp. Sig. (2-sided)
Pearson Chi-Square 10.341 (a) 6 .11 l
Likelihood Ratio 10.099 6 .121
Linear-by-Linear Association 3.644 1 .056
N of Valid Cases 3760

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 41.53.

Aware of Prejudices: First-Year Roommate

Crosstabulations of more aware of personal prejudices and students’ first-year

roommates suggest that requesting the first-year roommate may influence the likelihood

of agreeing with this statement (see Table 3.79). Students who share the same racial

104

background as their first-year roommate, but who did not request this roommate, were
most likely to strongly disagree that they have become more aware of their personal
prejudices and stereotypes since attending Michigan State University (8.4%). Students
who requested their first-year roommate, and who share their roommate’s racial
background, were most likely to disagree with this statement. Students who requested
their first-year roommate, and whose roommate is of a different racial background, were
most likely to agree with this statement (62.5%). Students whose first-year roommate is
of a different racial background, but did not request that roommate, were most likely to
strongly agree that they had become more aware of their personal prejudices since

attending Michigan State University (12.8%).

Table 3.79 First-year Roommate * Aware Personal Prejudices Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

First-year Roommate Aware Personal Prejudices
Strongly I Strongly
Disagree Disagree Agree Agree Total
Requested/Same race Count 64 225 4801 83 852
% 7.5% 26.4% 56.3% 9.7% 100.0%
Requested/Different Count 12 42 130 24 208
trace % 5.8% 20.2% 62.5% 11.5% 100.0%
Blind/Same race Count 144 354 1031 182 1711
% 8.4% 20.7% 60.3% 10.6% 100.0%
Blind/Different race Count 76 237 590 133 1036
% 7.3% 22.9% 56.9% 12.8% 100.0%
Total Count 296 858 2231 422 3807
% 7.8% 22.5% 58.6% 11.1% 100.0%

 

 

 

The chi-square value for this model is less than .05 indicating that we can reject

the null hypothesis that the first-year roommate and becoming more aware of personal

prejudices are independent (see Table 3.80).

105

Table 3.80 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 18.304(a) 9 .032
Likelihood Ratio 18.105 9 .034
Linear-by-Linear Association 1.899 1 .168
N of Valid Cases 3807

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 16.17.

Aware of Prejudices: Current Roommate (Upperclassmen)

Crosstabulations with the current roommate of upperclassmen and aware of

personal prejudices were conducted and results suggest slight differences (see Table

3.81). Upperclassmen currently living with roommates of their same racial background,

that they did not request to live with, were the most likely to strongly disagree with this

statement (8.1%). Upperclassmen living with a roommate of a different racial

background, that they did not request to live with, were the most likely to strongly agree

(15.8%) and agree (66.1%) with this statement.

Table 3. 81 Current Roommate (Upperclassmen Only)* Aware Personal Prejudices

recoded Crosstabulation

 

 

 

 

 

 

 

 

 

 

Current Roommate Aware Personal Prejudices recoded
(Upperclassmen Only) Strongly Strongly
Disagree Disagree Agree Agree Total
Blind/Same race Count 52 116 379 92 6391
% 8.1% 18.2% 59.3% 14.4%I 100.0%
Requested/Same race Count 57 164 530i 106 857
% 6.7% 19.1% 61.8% 12.4%I 100.0%
Blind/Different race Count 9 21 109 26 165
% 5.5% 12.7% 66.1% 15.8%I 100.0%I
Requested/Different race Count 7 45 136 27 215
% 3.3% 20.9% 63.3% 12.6%I 100.0%I
Total Count 125 346 1154 251 1876
% 6.7% 18.4% 61.5% 13.4% 100.0%I

 

 

 

 

 

 

 

106

The chi-square value for this model is greater than .05, however, indicating that
we cannot reject the null hypothesis that the current roommate of upperclassmen and

becoming more aware of personal prejudices are independent (see Table 3.82).

Table 3.82 Chi-Square Tests

 

 

Value df I Asymp. Sig. (2-sided)
Pearson Chi-Square 13.225(a) 9 .153
Likelihood Ratio 14.277 9 .1 13
Linear-by-Linear Association 2.008 1 .156
N of Valid Cases 1876

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 10.99.

Aware of Prejudices: Overall Racial Environment

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like to report becoming more aware of their personal prejudices
regardless of the overall racial environment of their current residence hall (Table 3.83).
The observed significance level for this test is greater than .05 indicating that we cannot
reject the null hypothesis. It appears that the overall racial environment is not a

significant predictor for this variable.

Table 3.83 Aware Personal Prejudices AN OVA
Sum of Squares df Mean Square F Sig.

 

 

 

 

 

 

 

 

 

Between Groups 34.735 47 .739 1.286 .092
Within Groups 2137.077 3719 .575
Total 2171.812 3766

 

Aware of Prejudices: Classroom Experience

One-way analysis of variance was conducted to test the null hypothesis that

participants are equally like to report becoming more aware of their personal prejudices

107

regardless of their classroom experiences (Table 3.84). The observed significance level
for this test is less than .005 indicating that we should reject the null hypothesis.

Table 3.84 Aware Personal Prejudices AN OVA

 

 

 

 

 

 

 

 

 

 

Sum of Squares df Mean Square F Sig.
Between Groups 150.070 22 6.821 12.654 .000
Within Groups 2075.964 3851 .539
Total 2226.034 3873
Summary

Many of the crosstabulations analyzed here have proven to be significant. As
discussed in the previous chapter, however, there are several limitations when using one-
item questions. Because respondents were not provided with a range of specific questions
that could not be collapsed into scale measurements for these variables, it cannot be
assumed that participants interpreted these questions similarly. As a result, the validity of
these measures is threatened. While the findings in this chapter contribute to our
understanding of race relations on the campus of Michigan State University, they cannot
be used to draw any meaningful conclusions at this time. Therefore, the dependent

variables considered in this chapter will not be included in future steps.

108

Clll'

 

 

CHAPTER 4: Interactional Diversity Scale

As discussed in chapter 2, a 10-item Interactional Diversity scale was created to
measure the extent to which students engage in meaningfiil interactions across race lines
(alpha=.916). This measure is an attempt to operationalize the concept of interactional
diversity. The items of this scale include dining with, having meaningful discussions
about race, sharing feelings, studying, socializing or partying, having intellectual
discussions outside of class, and attending events with students of a different racial
background. Also included is the extent to which students attend events sponsored by
racial/ethnic groups different from their own racial/ethnic background, and the frequency
with which students visit in the rooms of students of a different racial background as well
as in their own rooms with students of a different racial background. For each item,
responses ranged from 1(never) to 4(very often). The scales scores range from a
minimum of 10 points (students who report never engaging in any of the ten activities) to
40 points (students who report engaging in all ten activities very often). Of the students
who responded to these items (n=3534), the mean was 22.3, indicating that on average
students participate in these cross-racial activities occasionally (see Table 4.1). When we
consider this scale more closely, however, results indicate that some students are more
likely to engage in interactional diversity more often than others. The chi-square value for
this scale is less than .0005 (Table 4.2). Therefore, we can reject the null hypothesis that
participants are equally likely to earn any given score. This suggests that the differences
in students’ scores on the interactional scale are meaningful.

Table 4.1 Interaction Scale Statistics

 

 

 

* N Valid 4055
Missing 521
I Mean | 22.2653 |

 

109

I231

L111-
11

153'

 

a [it

1165]

E91151

in:

1:55

1101.

an:

0i it

Slide

Table 4.2 Test Statistics

 

Interaction scale

 

 

 

 

Chi-Square(a) 5992.012
df 61
Asymp. Sig. .000

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 65.4.

Before looking at the results for the Interactional Diversity Scale, we will first
consider how each item within the scale differed by some of the key independent
variables: racial/ethnic background, gender, class standing, and current hall. Following
this discussion, we will examine the Interactional Diversity Scale to determine how the
scale results differ on a number of important variables: racial/ethnic background, gender,
class standing, first-year status, current hall, hall assignment, first-year roommate, current

roommate, the Overall Racial Environment scale, and the Classroom Experience scale.

Interactional Diversity Scale Item Analysis

I dine or share meals with students of a different racial background than myself.

For the first item of the Interactional Diversity scale, the majority of students
reported dining with students of a different racial background occasionally (see Table
4.3). Of the students who responded to this question (n=4,03 5), 16.5 percent reported
never engaging in this activity, compared with 43.7 percent who described occasionally
engaging in this activity, 21.2 percent who described engaging in this activity often, and
18.6 percent who described engaging in this activity very often. The chi-square statistic
on this variable is less than .0005 indicating that we can reject the null hypothesis that

students are equally likely to dine or share a meal with peers of a different racial

110

background (Table 4.4). In the following section, we will explore how this variable

differs on some of the key independent variables.

Table 4.3 Dine/Eat Different Racial Background

 

 

 

 

 

 

 

 

 

 

 

 

 

I Frequency Percent Valid Percent
ever 664 14.5 16.5
Occasionally 1765 38.6 43.7
Often 857 18.7 21.2
Very Often 749 16.4 18.6
Total 4035 88.2 100.0
Missing 541 1 1.8 I
Total 4576 100.01 |
Table 4.4 Dine/Eat Different Racial Background: Chi-Square Test Frequencies
I Observed N Expected N Residual
ever 664 1008.8 -344.8
Occasionally 1765 1008.8 756.2
Often 857 1008.8 -151.8
Very Often 749 1008.8 -259.8
Total 4035

 

 

 

 

Dine/Eat Different Racial Background: Chi-Square Test Statistics

 

 

I Dine/Eat Different Racial Background I
Chi-Square 774.488(a)
df 3
Asymp. Sig. .000

 

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 1008.8.

Dine/Eat: Race/Ethnicig

Crosstabulations for this item indicate clear racial differences (Table 4.5). White,
Black, and Native Americans students are the least likely to dine with students of a
different racial background. Nearly two-thirds of the White respondents and at least half

of the Black and Native American students reported that they never, or only occasionally,

lll

engage in this activity. In contrast, more than 60 percent of Latino and Mixed Race

students report dining with students of a different racial background often or very often.

Asian students were the most likely to engage in this activity with more than 40 percent

reporting that they eat meals with peers of a different racial background very often and an

additional 25 percent reporting that they engage in this activity ofien. The chi-square for

this variable is significant at .000; therefore, we reject the null hypothesis that students

are equally likely to eat meals with peers of a different racial background than themselves

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(see Table 4.6).
Table 4.5 Dine/Eat Different Racial Background * Race/Ethnicity Crosstabulation
Dine/Eat Different Racial Race/Ethnicity
Background , ,
Native Mixed
White/Caucasian Black Latino American Asian Race Total
Never Count 538 44 8 l 12 7 610'
% 18.1% 16.6% 6.8% 10.0% 4.1%. 9.0% 16.4%
Occasionally Count 1393 100 37 4 81 23 1638
I % 47.0% 37.7% 31.4% 40.0% 27.8% 29.5% 43.9%
lOtten Count 605 55 31 3 72 17 783
% 20.4% 20.8% 26.3% 30.0% 24.7% 21.8% 21.0%
Very Often Count 430 66 42 2 126 31 697
% 14.5% 24.9% 35.6% 20.0% 43.3% 39.7% 18.7%
Total Count 2966 265 118 10 291 78 3728
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.6 Chi-Square Tests
Value J df Asymp. Sig. (2-sided)
Pearson Chi-Square 2.441E2 15 .000
Likelihood Ratio 228.849 15 .000
Linear-by-Linear Association 202.352 1 .000
N of Valid Cases 3728

 

 

 

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.64.

112

Dine/Eat: Gender

Crosstabulations on this variable suggest that men are somewhat more likely than
women to dine or eat with students of a different racial background (see Table 4.7).
Women were more likely to report never engaging in this activity than men (18.2% and
13.1% respectively). In contrast, men were more likely than women to report eating with
students of a different racial background very often (23.3% and 16.3% respectively). For

both groups, the most frequent response was “occasionally”.

Table 4.7 Dine/Eat Different Racial Background * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

Dine/Eat Different Racial Backgroundl Gender
Male Female Total
Never Count 1 73 459 632
% 13.1% 18.2% 16.4%
IOccasionally ICount 545 1144 1689i
% 41.2% 45.4% 43.9%
Ioaen Count 298 507 805
% 22.5% 20.1% 20.9%
Very Often Count 308 412 720!
% 23.3% 16.3% 18.7%
Total Count 1324 2522 3846
% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

The chi-square tabulation on this variable is less than .0005 indicating that we can
reject the null hypothesis that men and women are equally likely to dine with students of

a different racial background (see Table 4.8).

113

 

Table 4.8 Chi-Square Tests

 

 

 

 

Value Idf Asymp. Sig. (2-sided) I
Pearson Chi-Square 42.055(a) 3 .000
Likelihood Ratio 41.890 3 .000
Linear-by-Linear Association 41.959 1 .000
N of Valid Cases 3846

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 217.57.

Dine/Eat: Class Standing

Crosstabulations for this variable reveal that freshman are the least likely to dine

across racial lines with 17.6 percent reporting they never engage in this activity (Table

4.9). The likelihood of dining or eating with students of a different racial background

tends to increase with class standing although freshmen and sophomores have very

similar responses. Seniors were the most likely to dine with students of a different racial

background: 49.6 percent of seniors describe engaging in this activity often or very often

compared with 40.6 percent of juniors, 38.2 percent of sophomores, and 38.5 percent of

freshmen.

Table 4.9 Dine/Eat Different Racial Background * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

Dine/Eat Different Racial Class Standing |
Background Freshman Sophomore Junior I Senior Total I
Never Count 312 194 79I 45 6301
% 17.6% 15.9% 15.1% 14.3% 16.5%
Occasionally Count 775 5601 232 114 1681
% 43.8% 45.9% 44.3% 36.2% 43.9%
Often |Count 372 251 97 84 804
% 21.0% 20.6% 18.5% 26.7% 21.0%
Very Often Count 3091 215 116 72 712
% 17.5% 17.6% 22.1% 22.9% 18.6%
Total Count 1768 1220 524 315 3827
I% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

114

The chi-square value for this variable is less than .05; therefore, we can reject the null
hypothesis that students are equally likely to dine across racial lines, regardless of their
academic standing (Table 4.10).

Table 4.10 Chi-Square Tests

 

 

 

Value Idf I Asymp. Sig. (2-Sided)
Pearson Chi-Square 23.534(a) 9 .005
Likelihood Ratio 23.132 9 .006
Linear-by-Linear Association 1 1.580 1 .001
N of Valid Cases 3827

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 51.86.
Dine/Eat: Current Hall Percent White 9

When we consider the bivariate responses of this variable, crosstabulations
suggest that differences do exist between halls (Table 4.11). In most of the halls, students
were most likely to report never or only occasionally eating across racial lines. However,
there were a few exceptions. Holmes Hall, had the greatest number of residents reporting
that they engage in this activity often or very often (54.4%). As discussed in previous
chapters, Holmes Hall has a number of residents who participate in a living-learning
program. The only other hall with more than half of its residents eating with students of a
different racial background often or very often was Bryan Hall (52.2%). Bryan Hall is the
filth most diverse hall on campus with a high percentage of Asian/Pacific Islander and
International students. Students in Armstrong Hall were least likely to dine or eat with
someone of a different racial background (71.3%), followed by Holden (70.9%), Wilson
(70.4%), and Emmons (70.1%) halls. The chi-square value on this crosstabulation is less
than .0005 (see Table 4.12). Therefore, we can reject the null hypothesis that dining or

sharing a meal across racial lines and current residence hall are independent.

115

Table 4.11 Current Hall Percent White without Van Hoosen * Dine/Eat Bivariate

Crosstabulation

 

Current Hall Percent White without

Dine/Eat Bivariate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Van Hoosen Never or Occasionally Often or Very Often Total

Rather Count 5 l 51 102
I% 50.0% 50.0% 100.0%

Hubbard Count 1 5 1 130 281
% 53.7% 46.3% 100.0%

Butterfield Count 54 29 83
% 65. 1% 34.9% 100.0%

McDonel Count 137 137 274
% 50.0% 50.0% 100.0%

Bryan Count 64 70' 134
% 47.8% 52.2% 100.0%

Akers Count 166 126 292
% 56.8% 43.2% 100.0%

Shaw Count 144 78 222
% 64.9% 35.1% 100.0%

Holden |Count 178 73 251
% 70.9% 29.1% 100.0%
Bailey Count 57 43 100}
% 57.0% 43.0% 100.0%

Holmes Count 190 227 417
I% 45.6% 54.4% 100.0%

Williams Count 42 24 66
% 63.6% 36.4% 100.0%

Emmons Count 75 32 107
% 70.1% 29.9% 100.0%

Mason-Abbot Count 178 88 266
% 66.9% 33.1% 100.0%

Armstrong Count 82 33 l 15
% 71.3% 28.7% 100.0%

Case Count 147 89 236
% 62.3% 37.7% 100.0%

Landon Count 73 32 105
% 69.5% 30.5% 100.0%

Yakeley-Gilchrist ICount 1 12 52 164
% 68.3% 31.7% 100.0%

Campbell Count 54 3 1 85
% 63.5% 36.5% 100.0%

 

116

111513?

Slate

|
(1,};
r ‘J

 

 

11
1.

i

 

 

 

 

 

 

 

 

 

 

 

 

Wilson ICount 157 66 223
% 70.4% 29.6% 100.0%
IMayo Count 3 7 1 8 55
% 67 .3% 32.7% 100.0%
Wonders Count 137 74 21 1
% 64.9% 35.1% 100.0%
Total Count 2286 1503 3789
% 60.3% 39.7% 100.0%
Table 4.11 (cont’d) *

Table 4.12 Chi-Square Tests

 

 

 

 

 

 

Value df I Asymp. Sig. (2-sided) I
Pearson Chi-Square 1.215132 201 .000
Likelihood Ratio 121.514 20 .000
Linear-by-Linear Association 39.429 1 .000
N of Valid Cases 3789

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 21.82.

Results from the one-way analysis of variance filrther indicate the importance of

racial composition within the residence hall for predicting dining patterns of respondents

(see Table 4.13). When we test the null hypothesis that the likelihood of dining or sharing

a meal with students of a different racial background will be equal across residence halls,

the observed significance level for the comparison of means between halls is less than

.0005, suggesting that there is a significant relationship between these two variables.

Table 4.13 Dine/Eat Bivariate AN OVA

 

 

 

 

 

 

Sum of Squares df Mean Square F Sig.
Between Groups 29.074 20 1.454 6.241 .000
Within Groups 877.724 3768 .233
Total 906.798 3788

 

 

117

 

 

I visit with residents of a drfl'erent racial background than myself in they room.

For the second item of the Interactional Diversity scale, the majority of students
reported visiting with students of a different racial background in their room occasionally
(see Table 4.14). Of the students who responded to this question (n=4,039), 23.5 percent
reported never engaging in this activity, compared with 39.8 percent who described
occasionally engaging in this activity, 18.4 percent who described engaging in this
activity often, and 18.3 percent who described engaging in this activity very often.

Table 4.14 Visit Different Racial Background Their Room

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent

Never 948 20.7 23.5
Occasionally 1607 35.1 39.8
Often 744 16.3 18.4
Very Often 740] 16.2 18.3
Total 4039 88.3 100.0
Missing 537 1 1.7

Total 4576| 100.0.

 

The chi-square statistic on this variable is less than .0005 indicating that we can
reject the null hypothesis that students are equally likely to Visit the rooms of peers of a
different racial background (see Table 4.15). In the following section, we will explore

how this variable differs on some of the key independent variables.

Table 4.15 Visit Different Racial Background Their Room: Chi-Square Test
Frequencies

 

Observed N Expected N Residual

 

Never 948 1009.8 -61 .8
Occasionally 1607 1009.8 597.2
Often 744 1009.8 -265.8
Very Often 740 1009.8 -269.8

Total 4039

 

 

 

 

 

 

118

 

INA M

|‘-I

in

f
n 4.
it" t

tht

lites.

isms

 

Visit Different Racial Background Their Room: Chi-Square Test Statistics

 

 

 

 

I Visit Difference Racial Background Room I
Chi-Square 499.043(a)

f 3
Asymp. Sig. .000
Table 4.15 (cont’d)

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 1009.8.

Visit/Their Room: Race/Ethnicity

Crosstabulation results suggest that there are racial/ethnic differences for this
variable (see Table 4.16). White students were the least likely to visit the rooms of
students of a different racial background than themselves. More than two-thirds of White
students reported never, or only occasionally, participating in this activity. Black students
were also more likely than other non-White students to report never (21.3%) or
occasionally (41.6%) visiting the rooms of students of a different racial background than
themselves. Mixed Race students were the most likely to participate in this activity with
41 percent reporting that they visit the rooms of students of a different racial background
very often. Approximately 60 percent Latino, Native American, and Asian students
describe participating in this activity often or very often. The chi-square for this variable
is significant at .000; therefore, we reject the null hypothesis that students are equally

likely to visit the rooms of peers of a different racial background than themselves (see

Table 4.17).

119

Table 4.16 Visit Different Racial Background Their Room * Race/Ethnicity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Visit Different Racial Race/Ethnicity
Background Their Native Mixed
Room White/Caucasian Black Latino American Asian Race Total
ever Count 79 1 57 9 l 22 7 887
IN % 26.7% 21.3% 7.6% 10.0% 7.5% 9.0% 23.8%
Occasionally Count 1216 111 35 3 99 21 1485
I % 41.0% 41.6% 29.7% 30.0% 33.9% 26.9% 39.8%
Ioaen Count 506 39 37 5 77 18 682
% 17.0% 14.6% 31.4% 50.0% 26.4% 23. 1% 18.3%
Very Often Count 455 60 37 l 94 32 679
% 15.3% 22.5% 31.4% 10.0% 32.2% 41.0% 18.2%
Total Count 2968 267 l l 8 10 292 78 3733
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.17 Chi-Square Tests
Value df Asymp. Sig. (2-sided) I
Pearson Chi—Square 1.886E2 15 .000
Likelihood Ratio 189.966 15 .000
Linear-by-Linear Association 153.796 1 .000
N of Valid Cases 3733

 

 

 

 

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.82.

Visit/T heir Room: Gender

Crosstabulations on this variable reveal that men are more likely than women to

visit the rooms of students of a different racial background (Table 4.18). Twenty-two

percent of men describe engaging in this activity very often compared with 16.3 percent

of women. In contrast, 26.6 percent of women report never Visiting the room of students

of a different racial background compared with 18.3 percent of men. Men and women

were equally likely to describe engaging in this activity occasionally (39.8% and 39.7%

respectivelY); this was also the most popular response for both groups. The chi-square

value for this tabulation is less than .0005 indicating that we can reject the null hypothesis
120

that men and women are equally likely to visit in the rooms of students of a different

racial background (see Table 4.19).

Table 4.18 Visit Different Racial Background Their Room * Gender

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Visit Different Racial Gender
Background Their Room Male Female Total
Never Count 243 671 914
% 18.3% 26.6% 23.7%
Occasionally ICount 528 1004 1532
% 39.8% 39.7% 39-8%l
lonen Count 261 438 699
% 19.7% 17.3% 18.2%
Very Often Count 293 413 706
I% 22.1% 16.3% 18.3%
Total Count 1325 2526 3851
% 100.0% 100.0% 100°0%l
Table 4.19 Chi-Square Tests
Value I df Asymp. Sig. (2-sided) I
Pearson Chi-Square 43.179(a)1 3 .000
Likelihood Ratio 43.781 3 .000
Linear-by-Linear Association 40.237 1 .000
N of Valid Cases 3851

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 240.50.

Visit/T heir Room: Class Standing

Crosstabulation results for this variable reveal very similar results across classes
(Table 4.20). The majority of students report visiting the rooms of students of a different
racial background occasionally. In fact, at least two-thirds of respondents never or only
occasionally engage in this activity. The chi-square value for this tabulation is greater

than .05 (see Table 4.21). Therefore, we cannot reject the null hypothesis that the

121

likelihood for visiting the rooms of students of a different racial background is equal

regardless of academic standing.

Table 4.20 Visit Different Racial Background Their Room * Class Standing

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Visit Different Racial Academic Standing
Background Their Room FreshmanI Sophomore I Junior Senior Total
Never Count 4201 3001 124 67 91 1
% 23.7% 24.5% 23.8% 21.2% 23.8%]
Occasionally |Count 707 498 205 116 1526
% 39.9% 40.8% 39.3% 36'7%l 39.8%I
lOfien Count 324 21 1 87 74 6961
% 18.3% 17.3% 16.7% 23.4% 18.2%
Very Often Count 321 213 105 59 698
I% 18.1% 17.4% 20.2% 18.7%I 18.2%
Total Count 1772 1222 521 3 16 3831
% 100.0% 100.0% 100.0% 100.0%I 100.0%
Table 4.21 Chi-Square Tests f
Value I df I Asymp. Sig. (2—sided) I
Pearson Chi-Square 9.76431 9 .370
Likelihood Ratio 9.406 9 .401
Linear-by-Linear Association 1.263 1 1 .261
N of Valid Cases 383 1

 

 

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 57.41.

Visit/T heir Room: Current Hall Percent White

Crosstabulation results suggest that the current hall of the resident plays a role in

predicting the likelihood that they will visit the rooms of students of a different racial

background (see Table 4.22). Residents of Rather Hall are the most likely to describe

engaging in this activity often or very often (50%), followed by Holmes Hall (48.4%). Of

all the balls on the campus of Michigan State University, Rather has the highest

percentage of students of color. Holmes Hall houses the largest living-learning program

122

on campus. In contrast, residents were the least likely to Visit the rooms of students of a
different racial background in Mayo and Wilson Halls: only 18.2 percent of residents in
Mayo Hall and 22.0 percent of residents in Wilson Hall describe engaging in this activity
often or very often. Mayo Hall has the second highest proportion, and Wilson has the
third highest proportion of White students on campus. The chi-square value on this
tabulation is less than .0005 (see Table 4.23). Therefore, we can reject the null hypothesis
that respondents are equally likely to visit the rooms of students of a different racial

background regardless of which residence hall they live in.

Table 4.22 Current Hall Percent White without Van Hoosen * Visit/1' heir Room
Bivariate Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Current Hall without Van Hoosen Visit/Their Room Bivariate
Never or Occasionally Often or Very Ofien Total

Rather Count 5 1 5 1 102
% 50.0% 50.0% 100.0%
Hubbard Count 151 130F 281
% 53.7% 46.3% 100.0%
Butterfield Count 5 l 32 83
% 61.4% 3 8.6% 100.0%
McDonel Count 144 1 30 274
% 52.6% 47.4% 100.0%
Bryan Count 72 62 134
% 53.7% 46.3% 100.0%
Akers Count 159 138 297
% 53.5% 46.5% 100.0%
Shaw Count 146 76 222
% 65.8% 34.2% 100.0%
Holden Count 170' 8 1 25 1
% 67.7% 32.3% 100.0%
Bailey Count 58 42 100
% 58.0% 42.0% 100.0%
Holmes Count 2 1 5 202 417
% 51.6% 48.4% 100.0%
Williams Count 50 16 66
% 75.8% 24.2% 100.0%

 

123

 

 

We

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Emmons ICount 74 33I 107
% 69.2% 30.8% 100.0%
Mason-Abbot Count 195 701 265
% 73.6% 26.4% 100.0%
Armstrong Count 74 41 1 15
% 64.3% 35.7% 100.0%
Case Count 163 72 235
% 69.4% 30.6% 100.0%
Landon Count 79 26 105
% 75.2% 24.8% 100.0%
Yakeley-Gilchrist Count 123 44 167
% 73.7% 26.3% 100.0%
{Campbell Count 64 21 85
% 75.3% 24.7% 100.0%
Wilson |Count 174 49 223
% 78.0% 22.0% 100.0%
Mayo Count 45 101 55
% 81.8% 18.2% 100.0%
Wonders Count 146 65 21 1
% 69.2% 30.8% 100.0%
Total Count 2404 1391 3795
% 63.3% 36.7% 100.0%
Table 4.22 (cont’d)
Table 4.23 Chi-Square Tests
Value de Asymp. Sig. (2-sided) |
Pearson Chi-Square 1.521E2 20I .000
Likelihood Ratio 154.556 20 .000
Linear-by-Linear Association 92.938 1 .000
N of Valid Cases 3795

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 20.16.

Results from the one-way analysis of variance indicate that the racial composition
of the residence hall is an important determinant predictor of students’ likelihood to visit
the rooms of peers of a different racial background (see Table 4.24). The Observed

significance level for the comparison of means between balls is less than .0005,

124

indicating that we can reject the null hypothesis that the likelihood of engaging in this
activity will be equal across residence halls.

Table 4.24 Visit/T heir Room Bivariate AN OVA

 

 

Sum of Squares df IMean Square F Sig.I
Between Groups 35.317 20 1.766 7.879 .000
Within Groups 845.833 3774 .224
Total 881.150 3794

 

 

 

 

 

 

I visit with residents of a different racial background in r_n_p room.

For the third item on the Interactional Diversity scale, students were most likely to
report occasionally visiting with students of a different racial background in their own
room (see Table 4.25). Of the respondents who answered this question (n= 4,029) 41
percent of students describe engaging in this activity occasionally and an additional 25.8
percent report never doing so. The chi-square value for this measure is Significant at less
than .0005 (Table 4.26). Therefore, we can reject the null hypothesis that all students are
equally likely to visit with students of a different racial background in their own room. In
the following section, we will consider how this variable differs on some of the key

independent variables.

Table 4.25 Visit Different Racial Background My Room

 

 

 

 

 

r Frequency Percent Valid Percent
ever 1038 22.7 25.8
Occasionally 1659 36.3 41.2
Often 664 14.5 16.5
Very Often 668 14.6 16.6
Total 4029 88.0 100.0]
Missing 547 12.0
Total 4576 100.0]

 

 

 

 

125

Table 4.26 Visit Different Racial Background My Room: Chi-Square Test
Frequencies

 

 

I Observed N Expected N Residual

ever 1038 1007.2 30.8
Occasionally 1659 1007.2 651.8
Often 664 1007.2 -343.2
Very Often 668 1007.2 -339.2
Total 4029

 

 

 

 

 

Visit Different Racial Background My Room: Chi-Square Test Statistics

 

 

I Visit Different Racial Background My Room I
Chi-Square 653.894(a)
df 3
Asymp. Sig. .000

 

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 1007.3.

Visit/My Room: Race/Ethnicity
There appear to be stark contrasts for this item based on race/ethnicity (see Table 4.27).

White students were far less likely than other respondents to invite students of a different
racial background into their rooms with more than 70 percent reporting that they never or
only occasionally engage in this activity. Black students were less likely than other non-
White students to visit with students of a different racial background in their own room.
Mixed Race students were the most likely to invite students of a different racial
background into their rooms with nearly 40 percent reporting that they engage in this
activity very often. Except for Mixed Race students, however, the most common
response for all racial/ethnic groups on this item was “occasionally”. The chi-square for
this variable is significant at .000; therefore, we reject the null hypothesis that students
are equally likely to invite peers of a different racial background than themselves to visit

their rooms (see Table 4.28).

126

 

Vet}

1811’

 

lab
,—

Peal
111:

 

..._
1:5 e.

P)
_“.'A

’:
5?‘

I

011:
net
“'0:

42.;

Table 4.27 Visit Different Racial Background My Room * Race/Ethnicity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Visit Different Racial Race/Ethnicity
Background My Room Native Mixed
White/Caucasian Black Latino American Asian Race Total
Never Count 846 58 15 1 29 101 959
% 28.6% 21.7% 12.7% 10.0% 9.9% 12.8% 25.8%
Occasionally Count 1277 97 40 4 102 22 1542
% 43 .2% 36.3% 33.9% 40.0% 34.9% 28.2% 41.4%
lOficn Count 459 47 24 4 61 15 610
% 15.5% 17.6% 20.3% 40.0% 20.9% 19.2% 16.4%
Very Often Count 377 65 39 l 1001 31 613
% 12.7% 24.3% 33.1% 10.0% 34.2% 39.7% 16.5%
Total Count 2959 267 1 18 10 292 78 3724
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.28 Chi-Square Tests y
Value df Asymp. Sig. (2-sided) I
Pearson Chi-Square 2.117E2 15 .000
Likelihood Ratio 198.413 15 .000
Linear-by-Linear Association 174.217 1 .000
N of Valid Cases 3724

 

 

 

 

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.64.

Visit/My Room: Gender

Crosstabulation results suggest that gender of the respondent plays a role in the
likelihood that they will visit with students of a different racial background in their own
room (Table 4.29). More than 20 percent of men report engaging in this activity very
often compared to 14.3 percent of women. In contrast, 28.4 percent of women report
never engaging in this activity compared to 20.9 percent of men. For both men and
women, the most popular response for this measure was “occasionally” (39.9% and

42.2% respectively). The chi-square value for this tabulation is less than .0005 indicating

127

that we can reject the null hypothesis that men and women are equally likely to visit with

students of a different racial background in their own rooms (Table 4.30).

Table 4.29 Visit Different Racial Background My Room * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Visit Different Racial Gender
Background My Room Male Female Total
Never Count 277 71 5 992
% 20.9% 28.4% 25.8%
Occasionally Count 528 1063 1591
% 39.9% 42.2% 41.4%
lonen Count 243 382 625
% 18.4% 15.2% 16.3%
Very Often Count 275 359 634
% 20.8% 14.3% 16.5%
Total Count 1323 2519 3842
% 100.0% 100.0% 100.0%
Table 4.30 Chi-Square Tests f
Value Idf Asymp. Sig. (2-sided) |
Pearson Chi-Square 47.643(a) 3 .000
Likelihood Ratio 47.446 3 .000
Linear-by-Linear Association 47.551 1 .000
N of Valid Cases 3842

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 215.22.

Visit/My Room: Class Standing

Crosstabulations on this variable suggest there are some differences in the

likelihood of visiting with students of a different racial background between students

based on class standing although there does not appear to be an obvious pattern (Table

4.31). Sophomores were the most likely to report never engaging in this activity (26.6%)

while juniors were the most likely to report engaging in this activity very ofien (20.0%).

Among all groups, students were most likely to describe visiting with students of a

128

different racial background in their own rooms occasionally: 42.4 percent of freshmen,

42.3 percent of sophomores, 40.4 percent of juniors, and 35.4 percent of seniors. The chi-

square value for this tabulation is less than .05 so we can reject the null hypothesis that

students are equally likely to engage in this activity regardless of academic standing

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(Table 4.32).
Table 4.31 Visit Different Racial Background My Room * Class Standing
Crosstabulation
Visit Different Racial Class Standing
Background My Room Freshman Sophomore I Junior Senior Total
Never Count 452 324 136 74 986
% 25.6% 26.6% 25.9% 23.6% 25.8%
Occasionally Count 748 515 212 111 1586
L % 42.4% 42.3% 40.4% 35.4% 41.5%
lOften Count 287 188 72 74 621
% 16.3% 15.4% 13.7% 23.6% 16.2%
Very Ofien Count 278 191 105 55 629
% 15.8% 15.7% 20.0% 17.5% 16.5%
Total Count 1765 1218 525 314 3822
% 100.0% 100.0% 100.0% 100.0%I 100.0%
Table 4.32 Chi-Square Tests
Value df I Asymp. Sig. (2-sided)
Pearson Chi-Square 22.586(a) 9 .007
Likelihood Ratio 21.351 9 .011
Linear-by-Linear Association 3.941 1 .047
N of Valid Cases 3822

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 51.02.

Visit/My Room: Current Hall Percent White

The crosstabulations for the bivariate results on this measure suggest that current

hall plays a role in the likelihood that students will visit with peers of a different racial

background in their own rooms (Table 4.33). As was the case for the previous variable,

129

01111 1

lab]
BM]

lam

‘-

Helm

“1111;

 

residents of Rather and Holmes Hall are the most likely to report engaging in this activity

often or very often (51.0% and 44.7% respectively). Students living in Wilson Hall were

the least likely to engage in this activity: 79.] percent report never or only occasionally

visiting with students of a different racial background in their own room. The chi-square

value for this tabulation is less than .0005. Therefore, we can reject the null hypothesis

that students are equally likely to visit with peers of a different racial background in their

own room, regardless of their current residence hall (Table 4.34).

Table 4.33 Current Hall Percent White without Van Hoosen * Visit/My Room

Bivariate Crosstabulation

 

Current Hall Percent White without
Van Hoosen

 

Visit/My Room Bivariate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Never or Occasionally I Ofien or Very Ofien Total

Rather Count 501 52 102
I % 49.0% 51.0% 100.0%
Hubbard |Count 165 1 15 2801
% 58.9% 41.1% 100.0%

'Butterfield Count 58 25 33
% 69.9% 30.1% 100.0%

McDonel Count 1 59 l 15 274
% 58.0% 42.0% 100.0%

Bryan Count 74 59 133
% 55.6% 44.4% 100.0%

Akers Count 191 105 296
% 64.5% 35.5% 100.0%

Shaw Count 165 56 221
% 74.7% 25.3% 100.0%

Holden Count 1 75 76 25 l
% 69.7% 30.3% 100.0%
Bailey iCount 62 37 99‘
% 62.6% 37.4% 100.0%

Holmes Count 230! 186 416
% 55.3% 44.7% 100.0%

Williams Count 5 l 15 66
% 77.3% 22.7% 100.0%

 

 

 

 

 

 

130

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Emmons Count 80 27 107
% 74.8% 25.2% 100.0%
Mason-Abbot |Count 192 75 267
% 71.9% 28.1% 100.0%
Armstrong Count 83 32 l 15
% 72.2% 27.8% 100.0%
HCase Count 167 68 235
% 71.1% 28.9% 100.0%
Landon Count 81 24 105
% 77.1% 22.9% 100.0%
Yakeley-Gilchrist Count 125 41 166
% 75.3% 24.7% 100.0%
{Campbell Count 6 l 24 85
% 71.8% 28.2% 100.0%
Wilson Count 174 46 220!
% 79.1% 20.9% 100.0%
IMayo Count 42 13 55
% 76.4% 23.6% 100.0%
Wonders Count 154 55 209'
% 73.7% 26.3% 100.0%
Total Count 2539 1246 3785
% 67.1% 32.9% 100.0%
Table 4.33 (cont’d)
Table 4.34 Chi-Square Tests ,
Value dfl Asymp. Sig. (2-sided) |
Pearson Chi-Square 1.195E2 20 .000
Likelihood Ratio 119.158 20 .000
Linear-by-Linear Association 59.371 1 .000
of Valid Cases 3785

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 18.11.

Results from the one-way analysis of variance indicate that the racial composition

of the residence hall is an important predictor for whether students will visit with students

of a different racial background in their own rooms (see Table 4.35). When we test the

null hypothesis, the observed significance level for the comparison of means between

131

halls is less than .0005, suggesting that there is a significant relationship between these

two variables.

Table 4.35 Visit/My Room Bivariate ANOVA ,
Sum of Squares df IMean SquareI F Sig.I

 

 

Between Groups 26.400 20 1.320 6.138 .000
Within Groups 809.424 3764 .215
Total 835.824 3784

 

 

 

 

 

I have meaningful and honest discussions about race and ethnic relations with
students from different racial backgrounds than myself outside at class.

For the fourth item on the Interactional Diversity scale, few respondents describe
having meaningful discussions about racial/ethnic relations with students of a different
racial background (see Table 4.36). Of the participants who answered this question (n=
3,916) 47.7 percent of students describe engaging in this activity occasionally and an

additional 30.5 percent report never doing so.

Table 4.36 Meaningful Honest Discussions About Race

 

 

 

 

I Frequency Percent Valid Percent 1
ever 1 196 26.1 30.5
Occasionally 1 869 40.8 47.7
Often 559 12.2 14.3
Very Often 292 6.4 7.5
Total 3916 85.6 100.0]
Missing 660 14.4
Total 4576 100.0

 

 

 

 

 

The chi-square value for this measure is significant at less than .0005 (Table
4.3 7). Therefore, we can reject the null hypothesis that all students are equally likely to

have meaningful and discussions about racial/ethnic relations across race lines.

132

Table 4.37 Meaningful Honest Discussions About Race: Chi-Square Test
Frequencies

 

 

I Observed N Expected N I Residual I

ever 1 196 979.0 217.0
Occasionally 1 869 979.0 890.0
Often 559 979.0 -420.0
Very Often 292 979.0 -687.0
Total 3916

 

 

 

Meaningful Honest Discussions About Race: Chi-Square Test Statistics

 

Meaningful Honest Discussions About Race
Chi-Square 1 5 19.467(a)
df 3
Asymp- .000
81g.

 

 

7a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 979.0.

Meaninfiul Discussions: Race/Ethnicity

The racial/ethnic background of respondents appears to play some role on this
variable (Table 4.38). White students were considerably less likely to engage in these
discussions. More than 80 percent of White students described never or only occasionally
participating in these types of discussions. It should be noted, though, that the majority of
students report low levels of participation in discussions about racial/ethnic issues with
students of a different racial background outside of class. Native American students were
the most likely to engage in these discussions, with half reporting that they do so ofien.
The chi-square for this variable is significant at .000; therefore, we reject the null
hypothesis that students are equally likely to have meaningful discussions about
racial/ethnic issues with peers of a different racial background outside of class (Table

4.39).

133

Table 4.38 Meaningful Honest Discussions About Race * Race/Ethnicity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
IbTeaningful Honest Race/Ethnicity
Discussions About Native Mixed
Race White/Caucasian Black Latino American Asian Race I Total
ever Count 982 61 21 1 64 101 1139
IN I% 33 .2% 22.8% 17.9% 10.0% 21.9% 13.0% 30.6%
Occasionally Count 1420! 129 50 4 139 37 1779
I % 48.0% 48.3% 42.7% 40.0% 47.6% 48.1% 47.8%
lonen Count 375 41 30F 5 63 17 531
% 12.7% 15.4% 25.6% 50.0% 21 .6% 22.1% 14.3%
Very Often Count 184 36 16 0 26 13 275
% 6.2% 13.5% 13.7% .0% 8.9% 16.9% 7.4%
Total Count 2961 267 l 1 7 10 292 77 3724
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.39 Chi-Square Tests
Value df Asymp. Sig. (2-sided) I
Pearson Chi-Square 1.105E2 15 .000
Likelihood Ratio 102.574 15 .000
Linear-by-Linear Association 59.225 1 .000
N of Valid Cases 3724

 

 

 

 

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is .74. 7

Meaningful Discussions: Gender

Crosstabulations suggest that men and women are similar in their likelihood of

engaging in meaningful and honest discussions about racial/ethnic issues across race lines

(Table 4.40). More than three-quarters of participants from both groups describe never or

only occasionally engaging in these discussions. The chi-square value for this tabulation

is greater than .05, indicating that we cannot reject the null hypothesis that men and

women are equally likely—or in this case, equally unlikely—to engage in meaningful

discussions about racial/ethnic issues with students of a different racial background

(Table 4.41 ).

134

Table 4.40 Meaningful Honest Discussions About Race * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Meaningful Honest Discussions About Gender I
Race Male Female Total I
Never |Count 395 784 l 1791
% 29.9% 31.1% 30.7%
Occasionally Count 610 1217 1827
% 46.2%I 48.3% 47.6%
lOften Count 2101 341 551
% 15.9% 13.5% 14.4%
Very Often lCount 105 177 282
% 8.0% 7.0% 7.3%
Total Count 1320 2519 3839
% 100.0%I 100.0%I 100.0%I
4.41 Chi-Square Tests 7
Value Idf Asymp. Sig. (2-sided)
Pearson Chi-Square 5.619(a) 3 .132
Likelihood Ratio 5.559 3 .135
Linear-by-Linear Association 3.427 1 .064
N of Valid Cases 3839

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 96.96.

Meaningful Discussions: Class Standing

Crosstabulation results on this variable reveal that class standing plays a role in

predicting whether students engage in meaningful discussions about race with students of

a different racial background (Table 4.42). Participant responses suggest that as students

advance in their academic careers, they become more likely to engage in these

discussions. Freshmen were the least likely to participate in these discussions with 36.7

percent reporting that they never have meaningful discussions about racial/ethnic issues

across race lines and 45.2 percent reporting they only occasionally do so. In contrast,

seniors were the most likely to engage in these discussions: 10.9 percent describe

135

:1:
,7/

(

cf
U

'31

‘

es

 

engaging in meaningful discussions about racial/ethnic issues across race lines very often

and an additional 23.3 percent report engaging in these discussions often. The chi-square

value for this tabulation is less than .0005 indicating that we can reject the null hypothesis

that students are equally likely to discuss racial/ethnic issues with peers of a different

racial background regardless of their academic standing (Table 4.43).

Table 4.42 Meaningful Honest Discussions About Race * Class Standing

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Meaningful Honest Class Standing
Discussions About Race F reshmanI Sophomore Junior Senior I Total
Never Count 649 334 131 59l 1173
% 36.7% 27.5% 25.0% 18.8% 30.7%
Occasionally Count 7991 613 255 147 1814
% 45.2% 50.5% 48.7% 47.0% 47.5%
IOften 'Count 218 1701 87 73 548
% 12.3% 14.0% 16.6% 23.3% 14.4%
Very Often Count 1001 98 51 34 283
% 5.7% 8.1% 9.7% 10.9% 7.4%
Total Count 1766 1215 524 313 3818
% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.43 Chi-Square Tests
Value I (if I Asymp. Sig. (2-sided) I
Pearson Chi-Square 90.548at 9 .000
Likelihood Ratio 89.140 9 .000
Linear-by-Linear Association 76.458 1 .000
151 ofValid Cases 3818

 

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 23.20.

Meaningful Discussions: Current Hall Percent White

Crosstabulations on this variable suggest that the current residence hall may play

a role in the likelihood that students will engage in meaningful and honest discussions

about racial/ethnic issues across race lines (see Table 4.44). Respondents living in

136

Emmons and Armstrong halls were the least likely to report engaging in these activities:
92.5 percent of residents in Emmons Hall and 87.0 percent of residents in Armstrong Hall
report never or only occasionally discussing racial/ethnic issues with peers of a different
racial background. More than 90 percent of the residents in these halls are freshmen—the
highest ratio of any halls on campus. Respondents living in Williams Hall were the most
likely to describe engaging in this activity often or very often (30.3%), followed by Case
Hall (28.5%). Excluding Van Hoosen Hall (which is not included in the independent
variable) Williams Hall has the greatest percentage of seniors, and the lowest percentage
of freshmen, on campus. The majority of residents in Case Hall are freshmen and
sophomores; however, the second largest living-learning program on campus is located
here. The James Madison College offers students an interdisciplinary program in the
social sciences. The living-learning component and the focus of this specific program are
likely contributors to the findings on this variable. The chi-square value for this
tabulation is less than .005 indicating we can reject the null hypothesis that students are
equally likely to have meaningful and honest discussions about racial/ethnic issues across

race lines regardless of the hall in which they live (see Table 4.45).

Table 4.44 Current Hall Percent White without Van Hoosen * Meaningful
Discussions Bivariate Crosstabulation

 

 

 

 

 

 

 

 

Current Hall Percent White Meaningful Discussions Bivariate
Iwithout Van Hoosen Never or Occasionally Often or Very Often Total
lnatlter Count 76 27 103
I% 73.8% 26.2% 100.0%
Hubbard Count 204 78 282
% 72.3% 27.7% 100.0%
Butterfield Count 61 20 81
J% 75.3% 24.7%I 100.0%
McDonel Count 198 73 271
J% 73.1% 26.9% 100.0%

 

 

 

 

 

 

137

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hyatt Count 108 26 134
% 80.6% 19.4% 100.0%
Akers Count 234 63 297
1% 78.8% 21.2% 100.0%
Shaw Count 177 43 220
% 80.5% 19.5% 100.0%
Holden Count 198 54 252
% 78.6% 21.4% 100.0%
Bailey Count 76 23 99
% 76.8% 23.2% 100.0%
lllolmes Count 329 87 416
I% 79.1% 20.9% 100.0%
Williams Count 46 20F 66
% 69.7% 30.3% 100.0%
Emmons lCount 98 8 106
% 92.5% 7.5% 100.0%
Mason-Abbot Count 209 57 266
% 78.6% 21.4% 100.0%
Armstrong Count 1001 l 5 l 15
% 87.0% 13.0% 100.0%
Case Count 168 67 235
% 71.5% 28.5% 100.0%
lLandon Count 801 25 105
I% 76.2% 23.8% 100.0%
Yakeley-Gilchrist Count 135 3 1 166
% 81.3% 18.7% 100.0%
Campbell Count 68 l 7 85
I% 80.0% 20.0% 100.0%
Wilson Count 184 37 221
% 83.3% 16.7% 100.0%
Mayo Count 46 9 55
% 83.6% 16.4% 100.0%
Wonders [Count 167 42 209
% 79.9%. 20.1% 100.0%
Total Count 2962 822 37 84
% 78.3% 21 .7% 100.0%

Table 4.44 (cont’d)

138

 

Table 4.45 Chi-Square Tests

 

 

 

Value | de Asymp. Sig. (2-sided) |
Pearson Chi-Square 45.8723 20 .001
Likelihood Ratio 48.763 20 .000
Linear-by-Linear Association 8.460 1 .004
N of Valid Cases 3784

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 11.95.

Results from the one-way analysis of variance indicate that the racial composition
of the residence hall is an important predictor for whether students will discuss
racial/ethnic issues with peers of a different racial background (see Table 4.46). When we
test the null hypothesis, the observed significance level for the comparison of means
between halls is less than .005, suggesting that there is a significant relationship between
these two variables.

Table 4.46 Meaningful Discussions Bivariate ANOVA

 

 

 

 

 

 

Sum of SquaresI df IMean Square F ISig.
Between Groups 7.800 20 .390 2.309 .00]
Within Groups 635.637 3763 .169
Total 643.437 3783

 

I share personal feelings and problems with students of a different racial background
than myself.

For the fifth item on the Interactional Diversity scale, the majority of respondents
describe rarely sharing their feelings and problems with students of a different racial
background (see Table 4.47). Of the participants who answered this question (n= 3,908)
43.8 percent of students describe engaging in this activity occasionally and an additional
25.5 percent report never doing so. The chi-square value for this measure is significant at

less than .0005 (Table 4.48). Therefore, we can reject the null hypothesis that all students

139

gal/0”7fl'i
stafldiflg

«\i

 

 

are equally likely to share personal feelings and problems across race lines. In the
following section, we will examine how this variable differs by race, gender, class

standing, and current residence hall.

Table 4.47 Share Feelings

 

 

 

 

I Frequency Percent Valid Percent
[Never 996 21.8 25.5
Occasionally 1713 37.4 43.8
Often 801 17.5 20.5
Very Ofien 398 8.7 10.2
Total 3908 85.4 100.0]
Missing 668 14.6

Total 4576 100.0

 

 

 

 

 

Table 4.48 Share Feelings: Chi-Square Test Frequencies

 

 

I Observed N Expected N I Residual I

ever 996 977.0 19.0
Occasionally 1713 977.0 736.0
Often 801 977.0 -l76.0
Very Often 398 977.0 -579.0
Total 3908

 

 

 

Share Feelings: Chi-Square Test Statistics

 

 

 

[ Share Feeling

Chi-Square 929.656(a)
df 3
Asymp. Sig. 000'

 

 

la 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 977.0.

Share Feelingg: Race/Ethnicig

More than two-thirds of White and Black students reported never, or only
occasionally, sharing personal feelings and problems with student of a different racial

background than themselves (Table 4.49). Native American students were the most likely

140

 

 

to share their feelings with students of a different racial background, followed by Mixed

Race students. For students of all racial/ethnic backgrounds, other than Native

Americans, “occasionally” was the most frequent response for this item. The chi-square

for this variable is significant at .000; therefore, we reject the null hypothesis that

students are equally likely to share personal feelings and problems with students from a

different racial background than themselves (see Table 4.50).

Table 4.49 Share Feelings with Students of Different Racial Background *
Race/Ethnicity Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fhare Feeling with Race/Ethnicity
Students of Different White / Native
Races Caucasian Black Latino American Asian Mixed Race Total
Never Count 806 65 16 01 44 13 944
% 27.3% 24.4% 13.7% .0% 15. 1% 16.9% 25.4%
Occasionally Count 1325 119 45 3 116 25 1633
% 44.9% 44.7% 38.5% 30.0% 39.7% 32.5% 43.9%
IOften Count 558 55 41 4 81 24 763
% 18.9% 20.7% 35.0% 40.0% 27.7% 31.2% 20.5%
Very Often Count 265 27 15 3 51 15 376
% 9.0% 10.2% 12.8% 30.0% 17.5% 19.5% 10.1%
Total Count 2954 266 117 10 292 77 3716
% 100.0% 100.0% 100.0% 100.0%I 100.0% 100.0% 100.0%
Table 4.50 Chi-Square Tests
Value | df Asymp. Sig. (2-sided) |
Pearson Chi-Square 93.098(a) 15 .000
Likelihood Ratio 90.353 15 .000
Linear-by-Linear Association 72.959 1 .000
N of Valid Cases 3716

 

 

 

 

a 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.01.

Share Feelings: Gender

Crosstabulation results indicate that men and women are equally likely to share

their personal feelings and problems with peers of a different racial background (Table

4.51). For both groups, less than one-third of respondents describe engaging in this

141

 

, ' 27‘
’7”, ‘

MW

05

 

It.

 

 

activity often or very often. The majority of respondents report sharing their feelings and

problems across race lines occasionally. The chi-square value for this tabulation is greater

than .05 (see Table 4.52). Therefore, we cannot reject the null hypothesis that men and

women are equally likely to share their personal feelings and problems with students of a

different racial background.

4.51 Share Feelings * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Share Feelings Gender I
Male Female Total I
Never Count 352 628 980|
% 26.7% 25.0% 25.6%
Occasionally Count 594 1084 1678
If/o 45.1% 43.1% 43.8%
IOften Count 254 530 784
% 19.3% 21.1% 20.5%
Very Often Count 118 271 389
% 9.0% 10.8% 10.2%
Total Count 1318 2513 3831
% 100.0% 100.0% 100.0%
Table 4.52 Chi-Square Tests -
Value df Asymp. Sig. (2-sided) I
Pearson Chi-Square 5.985“I 3 .112
Likelihood Ratio 6.048 3 .109
Linear-by-Linear Association 5.308 1 .021
N of Valid Cases 3831

 

 

 

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 133.83.

Share Feelings: Class Standing

Crosstabulations reveal a fairly linear relationship between sharing feelings across

race lines and respondent’s academic standing (see Table 4.53). Freshmen were the most

likely to report never sharing personal feelings and problems with peers of a different

142

racial background (28.4%). In contrast, seniors were the most likely to report engaging in

this activity very often (18.6%). Respondents were most likely to describe occasionally

sharing their personal feelings and problems across race lines, however, regardless of

academic standing. The chi-square value for this tabulation is less than .0005 (Table

4.54). As a result, we can reject the null hypothesis that students are equally likely to

engage in this activity across classes.

Table 4.53 Share Feelings * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Share Feelings Class Standing
Freshman Sophomore Junior Senior Total

Never Count 502 293 129 48 972

% 28.4% 24.2% 24.6% 15.4% 25.5%
Occasionallleount 776 572 210- 113 1671
I % 44.0% 47.3% 40.1% 36.3% 43.9%
IOften Count 350 220 117 92 779

% 19.8% 18.2% 22.3% 29.6% 20.4%
Very Often lCount 137 125 68 58 388

% 7.8% 10.3% 13.0% 18.6% 10.2%
Total Count 1765 1210 524 311 3810

% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.54 Chi-Square Tests 7

Value Idf I Asymp. Sig. (2-sided) I

Pearson Chi-Square 81 .283a 9 .000
Likelihood Ratio 78.136 9 .000
Linear-by—Linear Association 59.962 1 .000
N of Valid Cases 3810

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 31.67.

Share Feelings: Current Hall Percent White

Crosstabulations on this variable suggest that students residing in less diverse

halls are less likely to share their personal feelings and problems across race lines (Table

143

4.55). Students living in Wilson and Mayo halls were the most likely to describe never or
only occasionally engaging in this activity (80.5% and 80.0% respectively). In contrast,
students living in McDonel and Hubbard halls were the most likely to report sharing their
personal feelings with peers of a different racial background often or very often (38.0%
and 36.9% respectively). The chi-square value for this tabulation is less than .0005
indicating that we can reject the null hypothesis that students are equally likely to share
their personal feelings and problems across racial lines regardless of the which hall they

reside in (Table 4.56).

Table 4.55 Current Hall Percent White without Van Hoosen * Share Feelings
Bivariate Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

Current Hall Percent White Share Feelings Bivariate
without Van Hoosen Never or Occasionally Often or Very Often Total
lRather Count 68 34 102
% 66.7% 33.3% 100.0%
Hubbard Count 178 104 282
% 63.1% 36.9% 100.0%
Butterfield Count 58 23 81
% 71.6% 28.4% 100.0%
lMcDonel Count 168 103 27 l
% 62.0% 38.0% 100.0%
Bryan Count 95 39 134
% 70.9% 29. 1% 100.0%
Akers Count 194 103 297
% 65.3% 34.7% 100.0%
Shaw Count 165 55 220
% 75.0% 25.0% 100.0%
Holden Count 182 66 248
% 73.4% 26.6% 100.0%
Bailey Count 68 3 l 99!
% 68.7% 31.3% 100.0%
Holmes Count 264 151 415
% 63.6% 36.4% 100.0%
Williams Count 42 24 66
% 63.6% 36.4% 100.0%

 

 

 

 

 

 

 

144

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Emmons Count 82 24 106
% 77.4% 22.6% 100.0%
Mason-Abbot Count 1 86 79 265
% 70.2% 29.8% 100.0%
Armstrong Count 88 27 l 15
% 76.5% 23.5% 100.0%
Case |Count 154 80 234
% 65.8% 34.2% 100.0%
Landon Count 74 30' 104
% 71.2% 28.8% 100.0%
Yakeley-Gilchrist Count 114 53 167
% 68.3% 31.7% 100.0%
Campbell Count 60! 25 85
% 70.6% 29.4% 100.0%
Wilson Count 1 78 43 221
% 80.5% 19.5% 100.0%
Mayo Count 44 ll 55
% 80.0% 20.0% 100.0%
Wonders Count 153 S6 209
% 73.2% 26.8% 100.0%
Total Count 261 5 1 161 3776
% 69.3% 30.7% 100.0%
Table 4.55 (cont’d)
Table 4.56 Chi-Square Tests
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 52.745(a) 20 .000
Likelihood Ratio 54.050 20 .000
Linear-by-Linear Association 15.244 1 .000
N of Valid Cases 3776

 

a 0 cells (.0%) have expected count less than 5. The minimmn expected count is 16.91 .

Results from the one-way analysis of variance indicate that the racial composition

of the residence hall is an important predictor for whether students will share personal

feelings and problems with students of a different racial background (see Table 4.57).

When we test the null hypothesis, the observed significance level for the comparison of

145

 

 

 

means between halls is less than .0005, suggesting that there is a significant relationship
between these two variables.

Table 4.57 Share Feelings Bivariate ANOVA

 

 

Sum of Squares df IMean Square F ISig.I
Between Groups 11.231 20 .562 2.660 .000
Within Groups 792.798 3755 .211
Total 804.029 3775

 

 

 

 

 

I study or prepare for class with students of a difi'erent racial background than myself.
For the sixth item on the Interactional Diversity scale, the majority of respondents
describe occasionally studying with peers of a different racial background (see Table
4.58). Of the participants who answered this question (n= 3,906) 42.5 percent of students
describe engaging in this activity occasionally compared with 28.5 percent who report
never studying across racial lines, 19.0 percent who report doing so often, and 9.8 percent
of participants who study with peers of a different racial background very often. The chi-
square value for this measure is significant at less than .0005 (Table 4.59). Therefore, we

can reject the null hypothesis that all students are equally likely to study across race lines.

 

 

 

 

 

 

Table 4.58 Study

r FrequencyIPercent Valid Percent
ever 1118I 24.4 28.6

Occasionally 1660 36.3 42.5

Often 744 16.3 19.0

Very Often 384 8.4 9.8

Total 3906 85.4 100.0

Missing 670 14.6

Total 4576| 100.0I I

 

 

 

146

 

 

_Table 4.59 Study: Chi-Square Test Frequencies

 

 

L Observed N IExpected N Residual

ever 1118I 976.5 141.5
Occasionally 1660 976.5 683.5
Often 744 976.5 -232.5
Very Often 384 976.5 -592.5
Total 3906

 

 

 

 

 

Study: Chi Square Test Statistics

 

 

| Study |
Chi-Square 91 3.781(a)
3

symp. Sig. .000

 

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 976.5.

Study: Race/Ethnicig

There are clear racial differences that exist on this variable (see Table 4.60).
Three-quarters of White respondents and two-thirds of Black respondents describe never,
or only occasionally studying with students of a different racial background than
themselves. Except for Black students, the students of color were most likely to describe
studying with students of a different racial background often. The chi-square for this
variable is significant at .000; therefore, we reject the null hypothesis that students are
equally likely to study or prepare for class with students fiom a different racial

background than themselves (see Table 4.61).

147

 

 

Table 4.60 Study with Students of Different Race * Race/Ethnicity Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IStudy with Students of Race/Ethnicity |
”mm" Rm Native Mixed
White/Caucasian Black Latino American Asian Race Total
Never Count 942 65 15 1 34 13 1070
% 31.9% 24.4% 12.8% 10.0% 1 1.6% 16.9% 28.8%
Occasionally Count 1309 104 38 3 93 22 15691
% 44.3% 39.1% 32.5% 30.0% 31.8% 28.6% 42.2%
lOften Count 484 57 391 4 98 26 708
I% 16.4% 21.4% 33.3% 40.0% 33.6% 33.8% 19.1%
Very Often Count 217 401 25 2 67 16 367
% 7.4% 15.0% 21.4% 20.0% 22.9% 20.8% 9.9%
Total Count 2952 266 l 17 10 292 77 3714
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.61 Chi-Square Tests f
Value df Asymp. Sig. (2-sided) I
Pearson Chi-Square 2.446E2 15 .000
Likelihood Ratio 229.008 15 .000
Linear-by-Linear Association 197.257 1 .000
N of Valid Cases 3714

 

 

 

 

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is .99.

Study: Gender

Crosstabulations suggest that men and women have very similar experiences

studying across racial lines (Table 4.62). Both groups were most likely to describe

occasionally studying with peers of a different racial background: 41.9 percent of men

and 42.9 percent of women. The chi-square value for this tabulation is greater than .05

(Table 4.63). Therefore, we cannot reject the null hypothesis that men and women are

equally likely to study with students Of a different racial background than themselves.

148

Table 4.62 Study * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Study Gender
Male Female Total

Never Count 376 721 1097

% 28.6% 28.7% 28.6%
Occasionally Count 552 1079 1631

I% 41.9% 42.9% 42.6%
lOften Count 252 474 726

% 19.1% 18.9% 19.0%
Very Often Count 136 239 375

I% 10.3% 9.5% 9.8%
Total Count 1316 2513 3829

I% 100.0% 100.0% 100.0%I
Table 4.63 Chi-Square Tests

Value Idf Asymp. Sig. (2-sided) I

Pearson Chi-Square .840(a) 3 .840
Likelihood Ratio .835 3 .841
Linear-by-Linear Association .425 1 .515
N of Valid Cases 3829

 

 

 

 

a 0 cells (.0%) have expected count less than S. The minimum expected count is 128.88.

Study: Class Standing

Crosstabulations suggest that students are increasingly likely to study across race

lines as they progress through their academic careers (see Table 4.64). More than 30

percent of freshmen describe never engaging in this activity compared with 26.9 percent

of sophomores, 26.0 percent Of juniors, and 19.6 percent Of seniors. In contrast, seniors

were the most likely to report studying with peers Of a different racial background: 14.8

percent of seniors describe engaging in this activity very Often compared to 12.2 percent

of juniors, 9.6 percent of sophomores, and 8.3 percent Of freshmen. For all groups,

however, students were most likely to report studying across race lines occasionally. The

149

 

 

chi-square value for this tabulation is less than .0005; therefore we can reject the null

hypothesis that students are equally likely to study with peers of a different racial

background regardless Of their class standing (Table 4.65).

Table 4.64 Study * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Study Class Standing
Freshman Sophomore Junior Senior Total
Never ICount 571 326 136 61 1094
% 32.4% 26.9% 26.0% 19.6% 28.7%
Occasionally Count 704 554 228 132 1618
% 40.0% 45.7% 43.6% 42.4% 42.5%
lOften Count 339 216 95 72 722
% 19.3% 17.8% 18.2% 23.2% 19.0%
Very Often Count 147 117 64 46 374
% 8.3% 9.6% 12.2% 14.8% 9.8%
Total Count 1761 1213 523 311 3808
% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.65 Chi-Square Tests , 7
Value Idf I Asymp. Sig. (2-sided)
Pearson Chi-Square 44.672(a) 9
Likelihood Ratio 44.283 9
Linear-by-Linear Association 28.876 1
N of Valid Cases 3808

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 30.54.

Study: Current Hall Percent White

Crosstabulations suggest that current residence hall and studying across racial
lines are related (Table 4.66). Students residing in Landon Hall, followed by Wilson Hall,
were the most likely to describe never or occasionally studying across race lines (85.7%
and 82.2%). Students residing in Rather Hall, followed by those in Holmes Hall were the

most likely to report engaging in this activity Often or very Often (43.7% and 42.0%

150

 

respectively). As discussed earlier, Rather Hall is the most racially diverse hall on

campus and Holmes Hall has the largest living-learning program on campus.

Table 4.66 Current Hall Percent White without Van Hoosen * Study Bivariate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Current Hall Percent White 7 Study Bivariate 7
Ilhout Van Hoosen Never or Occasionally IOften or Very OftenI Total
IRatber ICount 58I 45 103
% 56.3% 43.7% 100.0%
Hubbard Count 190 90| 280
% 67.9% 32.1% 100.0%
Butterfield Count 59I 22 81
% 72.8% 27.2% 100.0%
McDonel Count 1 82 89 27 1
I% 67.2% 32.8% 100.0%
Bryan Count 90 43 133
% 67.7% 32.3% 100.0%
Akers Count 215 82 297
% 72.4% 27.6% 100.0%
Shaw Count 1 59 6 l 220I
% 72.3% 27.7% 100.0%
Holden Count 1 96 55 25 1
% 78.1% 21.9% 100.0%
Bailey Count 71 27 98
% 72.4% 27.6% 100.0%
Holmes Count 240 174 414
% 58.0% 42.0% 100.0%
Williams Count 44 22 66
% 66.7% 33.3% 100.0%
Emmons Count 84 22 106
% 79.2% 20.8% 100.0%
Mason-Abbot Count 195 70 265
% 73.6% 26.4% 100.0%
Armstrong Count 96 191 1 15
% 83.5% 16.5% 100.0%
Case Count 157 78 235
I% 66.8% 33.2% 100.0%
Landon Count 90 15 105
% 85.7%I 14.3% 100.0%
Yakeley-Gilchrist Count 127 40 167
% 76.0% 24.0% 100.0%

 

 

 

 

 

 

 

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

lCampbell Count 62 23 85
% 72.9%I 27.1% 100.0%
Wilson Count 180I 39 ml
% 82.2% 17.8% 100.0%
Mayo Count 39 16 55
% 70.9% 29.1% 100.0%
Wonders Count 1 53 55 208
I% 73.6% 26.4% 100.0%
Total Count 2687 1087 3774
% 71.2% 28.8% 100.0%
Table 4.66 (cont’d)

The chi-square value for this tabulation is less than .0005 indicating that we can

reject the null hypothesis that students are equally likely to study with peers of a different

racial background regardless of which residence hall they live in (Table 4.67).

Table 4.67 Chi-Square Tests

 

Value Ide Asymp. Sig. (2-sided) I

 

Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N Of Valid Cases

 

 

98.9773'20

99.974
18.963
3774

 

20
1

.000
.000
.000

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 15.84.

Results from the one-way analysis Of variance indicate that the racial composition

of the residence hall is an important predictor for whether students will study with

students Of a different racial background (see Table 4.68). When we test the null

hypothesis, the Observed significance level for the comparison of means between halls is

less than .005, suggesting that there is a significant relationship between these two

variables.

152

 

Table 4.68 Study Bivariate AN OVA 7
Sum Of Squares df IMean Square F Sig.I

 

 

 

 

 

 

 

Between Groups 20.297 20 1.015 5.054 .000
Within Groups 753.622 3753 .201
Total 773.919 3773

 

I socialize or party with students of a different racial background than myself.

For the seventh item on the Interactional Diversity scale, the majority of
respondents describe socializing or partying with students Of a different racial
background occasionally or Often (see Table 4.69). Of the participants who answered this
question (n= 3,907) 44.2 percent Of students describe engaging in this activity
occasionally compared with 10.8 percent who report never socializing across race lines,
28.3 percent who report doing so Often, and 16.7 percent Of participants who socialize or

party with peers Of a different racial background very Often.

Table 4.69 Socialize/Party

 

 

 

 

 

 

 

I Frequency Percent Valid Percent
ever 422 9.2 10.8
Occasionally 1727 37.7 44.2
Often I 107 24.2 28.3
Very Often 651 14.2 16.7
Total 3907 85.4 100.0I
[dissing 669 14.6
Total 4576 100.0]

 

The chi-square value for this measure is significant at less than .0005 (Table
4- 70). Therefore, we can reject the null hypothesis that all students are equally likely to

SOCialize and party across race lines.

153

 

Table 4.70 Socialize/Party: Chi-Square Test Frequencies

 

 

 

 

 

 

 

 

 

 

 

 

I Observed N Expected N Residual

ever 422 976.8 -554.8
Occasionally 1727 976.8 750.2
Often 1107 976.8 130.2
Very Often 651 976.8 -325.8
Total 3907
Socialize/Party: Chi-Square Test Statistics

Socialize/PartyI

Chi-Square 1017.354(a)
df 3
Asymp. Sig. .000
a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
fi-equency is 976.8.

Socialize/Par_ty: Race/Ethnicigy

For this variable, race again appears to play a role; however, the pattern is slightly
different than the previous items that have been reviewed (see Table 4.71). On this item,
Black students were more likely than any other racial/ethnic group to report never
socializing or partying with students of a different racial background (17.3%). White
students exhibit similarly low levels of socializing with non-White students with the
majority never or only occasionally engaging in these activities. Latino and Mixed Race
students describe socializing across race lines more frequently than other racial/ethnic

groups with more than 70 percent engaging in this activity often or very often. The chi-

 

Square for this variable is significant at .000; therefore, we reject the null hypothesis that
stl-ldents are equally likely to socialize and party with students from a different racial

background than themselves (see Table 4.72).

154

 

Table 4.71 Socialize/Party * Race/Ethnicity Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Socialize/Party Race/Ethnicity
Native Mixed
White/Caucasian Black Latino American Asian Race Total
ever Count 314 46 4 01 27 5 396
% 10.6% 17.3% 3.5% .0% 9.2% 6.5% 10.7%
Occasionally Count 1399 109 29 5 87 18 1647
I % 47.3% 41.0% 25.4% 50.0% 29.8% 23.4% 44.3%
loncn Count 823 64 45 2 89 26 1049
I% 27.8% 24.1% 39.5% 20.0% 30.5% 33.8% 28.2%
Very Often |Count 421 47 36 3 89 28 624
% 14.2% 17.7% 31.6% 30.0% 30.5% 36.4% 16.8%
Total Count 2957 266 1 14 10 292 77 3716
% 100.0% 100.0% 100.0% 100.0%100.0% 100.0% 100.0%

 

 

 

Table 4.72 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided) 1
Pearson Chi-Square 1.429E2 15 .000
Likelihood Ratio 135.355 15 .000
Linear-by-Linear Association 80.678 1 .000
N Of Valid Cases 3716

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.07.1

Socialize/Pam: Gender

Crosstabulation results for this variable suggest that men are somewhat more

likely than women to socialize and party with students of a different racial background

(see Table 4.73). Twelve percent Of women describe never socializing across race lines

compared with 8.1 percent of men. In contrast, 18.4 percent of men describe socializing

across race lines very Often compared to 15.8 percent Of women. For both groups,

respondents were most likely tO report socializing or partying with students of a different

racial background occasionally. The chi-square value for this tabulation is less than .005

(see Table 4.74). Therefore, we can reject the null hypothesis that men and women are

equally likely tO engage in this activity.

155

Table 4.73 Socialize/Party * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Socialize/Party Gender
Male Female Total
Never Count 1 07 302 409
I% 8.1% 12.0% 10.7%
Occasionally Count 584 1117 1701
% 44.3% 44.4% 44.4%
IOften ICount 384 699 1083
% 29.1% 27.8% 28.3%
Very Often Count 243 396 639
% 18.4% 15.8% 16.7%
Total Count 1318 2514 3832
% 100.0% 100.0% 100.0%
Table 4.74 Chi-Square Tests
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 16.57081 3 .001
Likelihood Ratio 17.077 3 .001
Linear-by-Linear Association 12.179 1 .000
N Of Valid Cases 3832

 

 

 

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 140.67.—

Socialize/Pam: Class Standing

Crosstabulations suggest that class standing does play a role in the likelihood that
students will socialize or party with peers of a different racial background (Table 4.75).
However, compared with other variables in the Interactional Diversity scale, the
likelihood of engaging in this activity actually diminishes over time until the senior year.
Juniors were the least likely to socialize or party across race lines, followed by
sophomores, and then freshmen: 56.7 percent Of juniors, 55.6 percent of sophomores, and
54.9 percent Of freshmen report never or only occasionally engaging in this activity.

Seniors were the most likely to socialize or party with students of a different racial

156

 

 

hhulliallha

fl‘k.

background. Approximately half Of all seniors (49.4%) describe engaging in this activity

often or very Often. The chi-square value for this tabulation is less than .0005 indicating

we can reject the null hypothesis that students are equally likely to socialize and party

across race lines regardless Of their academic standing (Table 4.76).

Table 4.75 Socialize/Party * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

Socialize/Party Class Standing
Freshman Sophomore Junior Senior Total
Never Count 214 1 1 l 67 1 6 408
% 12.1% 9.1% 12.8% 5.1% 10.7%
Occasionally Count 755 564 229I 142 1690
% 42.8% 46.5% 43.9% 45.5% 44.3%
IOften Count 535 3201 134 87 1076
% 30.3% 26.4% 25.7% 27.9% 28.2%
Very Often Count 259‘ 219 92 67 637
% 14.7% 18.0% 17.6% 21.5% 16.7%
Total Count 1763 1214 522 312 3811
% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

 

 

Table 4.76 Chi-Square Tests

 

 

 

 

Value Idf] Asymp. Sig. (2-sided) ]
Pearson Chi-Square 35.2968' 9 .000
Likelihood Ratio 37.145 9 .000
Linear-by-Linear Association 6.787 1 .009
N OfValid Cases 3811

 

 

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 33.40.

Socialize/Pam: Current Hall Percent White

Crosstabulations for this variable reveal some differences between residence halls

(see Table 4.77). Respondents living in Yakeley-Gilchrist Hall (63.7%), followed by

Wilson (62.4%), Emmons (62.3%) and Holden (62.2%) halls were the most likely to

describe never or only occasionally socializing across race lines. More than three-quarters

157

Of residents in Yakeley-Gilchrist are female. In addition, Yakely-Gilchrist and Wilson
halls are two Of the least diverse residence balls on campus. Aside from this, however,
there is nothing immediately distinctive about these four residence halls that could
explain these findings. Residents in Rather and Holmes were the most likely tO report
socializing and partying with students of a different racial background: 54.4 percent Of
residents in Rather Hall and 53.7 percent Of residents in Holmes Hall describe engaging
in this activity often or very Often. AS discussed in previous sections, Rather Hall is the
most racially diverse residence hall on campus and Holmes Hall contains the largest
living-learning program. Compared with other halls on campus with a high percentage Of
White residents, participants in Wonders Hall are considerably more likely to socialize
across race lines. This may be due, in part, to the fact that Of the five least racially diverse
halls on campus, Wonders Hall has the highest percentage Of freshmen and the lowest
percentage Of juniors. The chi-square value for this tabulation is less than .005 (see Table
4.78). As a result, we can reject the null hypothesis that students are equally likely tO
socialize or party with peers of a different racial background regardless Of which hall they
reside in.

Table 4.77 Current Hall Percent White without Van Hoosen * Socialize/Party
Bivariate Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ICurrent Hall Percent White without Socialize/Party Bivariate

Van Hoosen Never or Occasionally Often or Very Often Total

Rather Count 47 56 103
°/o 45.6% 54.4% 100.0%

Hubbard Count 155 126 281
°/o 55.2% 44.8% 100.0%

Butterfield Count 45 36 81
% 55.6% 44.4% 100.0%

McDonel Count 135 136 27 l
°/o 49.8% 50.2% 100.0%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bryan Count 67 66 133
% 50.4% 49.6% 100.0%
Akers Count 1 59 138 297
% 53.5% 46.5% 100.0%
Shaw Count 123 96 219
% 56.2% 43.8% 100.0%
Holden Count 1 55 94 249
% 62.2% 37.8% 100.0%
lBailey Count 52 47 99
% 52.5% 47.5% 100.0%
Holmes Count 192 223 415
% 46.3% 53.7% 100.0%
Williams Count 40 26 66
% 60.6% 39.4% 100.0%
Emmons Count 66 40I 106
% 62.3% 37.7% 100.0%
Mason-Abbot Count 1 52 1 13 265
% 57.4% 42.6% 100.0%
Armstrong Count 59H 56 1 15
% 51.3% 48.7% 100.0%
ICase Count 135 100 235
% 57.4% 42.6% 100.0%
Landon Count 62 43 105
% 59.0% 41.0% 100.0%
Yakeley-Gilchrist Count 107 61 168
% 63 .7% 36.3%. 100.0%
ICampbell Count 52 33 85
"/6 61.2% 38.8% 100.0%
Wilson Count 13 8 83 221
% 62.4% 37.6% 100.0%
Mayo Count 34 21 55
% 61 .8% 38.2% 100.0%
Wonders Count 107 101 208
% 5 l .4% 48.6% 100.0%
Total Count 2082 1695 3777
% 55.1% 44.9% 100.0%

Table 4.77 (cont’d)

159

Table 4.78 Chi-Square Tests

 

 

Value I dfl Asymp. Sig. (2-sided)
Pearson Chi-Square 45.533(a)l 20 .001
Likelihood Ratio 45.679 20 .001
Linear-by-Linear Association 7.717 1 .005
N Of Valid Cases 3777

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 24.68.

Results from the one-way analysis Of variance indicate that the racial composition
of the residence hall is an important predictor for whether students will socialize or party
with students Of a different racial background (see Table 4.79). When we test the null
hypothesis, the Observed significance level for the comparison Of means between halls is
less than .005, suggesting that there is a significant relationship between these two

variables.

Table 4.79 Socialize/Party Bivariate ANOVA

 

 

Sum Of Squares df IMean Square F Sig.
Between Groups 11.264 20 .563 2.292 .001
Within Groups 923.073 3756 .246
Total 934.337 3776

 

 

 

 

 

 

 

I have intellectual discussions with students of a dlfferent racial background than

myself outside of class.

For the eighth item Of the Interactional Diversity scale, two-thirds Of students
report having intellectual discussions with peers Of a different racial background
occasionally or never (Table 4.80). Of the participants who responded to this question
(n=3,900), 19.7 percent describe never engaging in this activity and an additional 44.7
percent describe having intellectual discussions across race lines occasionally. However,

a sizeable minority Of respondents report having these discussions Often or very often

160

(23.6% and 12.4% respectively). The chi-square test for this variable is significant at a

value of less than .0005 which suggests that there are meaningful differences between

students for this measure (Table 4.81).

Table 4.80 Intellectual Discussions

 

 

 

 

 

 

FrequencyIPercent Valid Percent
ever 769 16.8 19.7
Occasionally 1730 37.8 44.4
Ofien 919 20.1 23.6
Very Often 482 10.5 12.4
Total 3900 85.2 100.0j
Missing 676 14.8 I
Total 4576I 100.0I

 

Table 4.81 Intellectual Discussions: Chi-Square Test Frequencies

 

 

Observed N I Expected N I Residual I

ever 769 975.0 -206.0

Occasionally 1730 975.0 755.0

Often 919 975.0 -56.0

Very Often 482 975.0 -493.0
Total 3900

 

 

Intellectual Discussions: Chi-Square Test Statistics

 

Intellectual Discussions I

Chi-Square 880.663(a)
(if 3
Asymp. Sig. .000

 

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
fi"i=quency is 975.0.

Intellectual Discussions: Race/Ethnici

 

Results from the crosstabulations suggest that the racial/ethnic background of

resDondents does play a role on this variable (see Table 4.82). Two-thirds of White

Stllclents described never or only occasionally engaging in these discussions. Compared to

161

other non-White groups, Black students were the least likely to have intellectual

discussions across race lines: nearly 60 percent report never or occasionally participating

in this activity. In contrast, Mixed Race students were the most likely to engage in these

discussions with nearly a third reporting that they have intellectual discussions with peers

of a different racial background very Often. Except for Mixed Race students, all students

were most likely to describe engaging in these discussions occasionally. The chi-square

for this variable is significant at .000; therefore, we reject the null hypothesis that

students are equally likely to have intellectual discussions with students from a different

racial background than themselves outside of class (see Table 4.83).

Table 4.82 Intellectual Discussion with Students of Different Race * Race/Ethnicity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C rosstabulation
alfillectual Discussion Race/Ethnicity
It Students of - -
Different Rm White/Caucasian Black I Latino I 14:12:12; Asian Iiiiiiiitd Total
eVer Count 643 501 101 OI 26 8 737
I% 21.8% 18.9% 8.5% .0% 9.0% 10.4% 19.9%.
Occasionally Count 1357 105 46 5 112 21 1646
% 46.0% 39.6% 39.3% 50.0% 38.6% 27.3% 44.4%
Often Count 655 68 31 2 85 23 864
% 22.2% 25.7% 26.5% 20.0% 29.3% 29.9% 23.3%
Very Often Count 295 42 30 3 67 25 462
1% 10.0% 15.8% 25.6%I 30.0% 23.1% 32.5% 12.5%
Total Count 2950 265 117 10 290 77 3709
% 100.0% 100.0% 100.0% 100.0%I 100.0%I100.0% 100.0%I
Table 4.83 Chi-Square Tests
Value df Asymp. Sig. (2-sided)I
Pearson Chi-Square 1.424132 15 .000
Likelihood Ratio 135.311 15 .000
Lillear-by-Linear Association 1 15.810 1 .000
I! of Valid Cases 3709

 

 

 

 

a. 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.25.

162

Intellectual Discussions: Gender

Crosstabulations for this variable indicate that men are more likely than women to

have intellectual discussions with peers of a different racial background outside Of the

classroom (Table 4.84). Women were the most likely to report never engaging in these

discussions (21.7%). In comparison, men were more likely than women to describe

having intellectual discussions across race lines occasionally, often, and very Often. For

both groups, however, respondents were most likely to report engaging in this activity

occasionally. The chi-square value for this tabulation is less than .005; therefore, we can

1‘6j ect the null hypothesis that men and women are equally likely to have intellectual

discussions with peers Of a different racial background outside Of class (Table 4.85).

Table 4.84 Intellectual Discussions * Gender Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Intellectual Discussions Gender
Male Female Total
ever Count 212 546 758
% 16.1% 21.7% 19.8%
Occasionally Count 609 1093 1702
% 46.4% 43.5% 44.5%
Ofien Count 3191 573 892
% 24.3% 22.8% 23.3%
Very Often Count 173 299 472
_ % 13.2% 11.9% 123%1
Total Count 1313 2511 3824
L % 100.0% 100.0% 100.0%
Ta ble 4.85 Chi-Square Tests
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 17.137(a) 3 .001
Likelihood Ratio 17.577 3 .001
LiIlear-by-Linear Association 9.413 1 .002
MValid Cases 3824

 

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 162.06.

163

Intellectual Discussions: Class Standing

As with most Of the variables that comprise the Interactional Diversity scale, the
likelihood that students will hold intellectual discussions with peers of a different racial
background appears to increase with class standing (Table 4.86). Freshmen were the
least likely to have these discussions: 23.6 percent Of freshmen describe never engaging
in intellectual discussions with peers of a different racial background, compared with
l 8 -4 percent Of sophomores, 15.9 percent of juniors, and 11.3 percent of seniors. Seniors
were the most likely to have these discussions, with 19.6 percent reporting they engage in
this activity very often, compared with 15.7 percent of juniors, 12.0 percent Of
Sophomores, and 10.5 percent of freshmen. The chi-square value for this tabulation is less
than .0005, indicating we can reject the null hypothesis that students are equally likely to
hOId intellectual discussions with peers Of a different racial background regardless of

tl‘leir academic standing (Table 4.87).

Table 4.86 Intellectual Discussions * Class Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I Intellectual Discussionsl Class Standing
Freshman Sophomore Junior Senior Total
Never Count 415 222 83 35 755
% 23.6% 18.4% 15.9% 11.3% 199%1
Occasionally Count 7701 564 236 ”91 16891
% 43.7% 46.7% 45.3% 38.3% 44.4%
Often Count 392 278 120- 96 886
% 22.2% 23.0% 23.0% 30.9% 23.3%
Very Often Count 185 145 82 61 473
\ % 10.5% 12.0% 15.7% 19.6% 12.4%
Total Count 1762 1209 521 311 3803
L\ % 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

164

Table 4.87 Chi-Square Tests

 

 

 

Value Idf I Asymp. Sig. (2-sided) I
Pearson Chi-Square 64.932(a) 9 .000
Likelihood Ratio 64.149 9 .000
Linear-by-Linear Association 53.717 1 .000
N of Valid Cases 3803

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 38.68.

Intellectual Discussions: Current Hall Percent White

 

The crosstabulations for this variable suggest that residents in Mayo and Emmons
hall are the least likely to hold intellectual discussions with peers of a different racial
background (Table 4.88). For both halls, 76.4 percent of residents describe never or only
occasionally participating in these types Of discussions. Although more than two-thirds Of
the residents in Mayo Hall are upperclassmen, nearly 90 percent Of its residents are
White—the second highest proportion on campus. Emmons Hall is more racially diverse
than many Of the halls on campus; however, more than 90 percent of its residents are
fi‘eshmen—also the second highest proportion on campus. Residents in Rather Hall,
followed by those in Bryan Hall are the most likely to describe having conversations with
students Of a different racial background. Rather Hall is the most racially diverse hall on
file campus of Michigan State University. In spite of the fact that nearly 90 percent of its

r‘esidents are freshmen, 48.0 percent of respondents describe participating in intellectual
di Seussions with peers of a different racial background outside of class. Bryan Hall is one
of the more diverse halls on campus, particularly with regard to Asian American and
International residents. More than 47 percent of participants living in Bryan Hall report

engaging in these types of discussions.

165

Table 4.88 Current Hall Percent White without Van Hoosen * Intellectual

Discussions Bivariate Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[Current Hall Percent White without Intellectual Discussions Bivariate
Van Hoosen Never or Occasionally Often or Very Often I Total
Rather Count 53 491 102
% 52.0% 48.0% 100.0%
Hubbard Count 176 104 280l
% 62.9% 37.1% 100.0%
Butterfield Count 50! 29I 79
% 63.3% 36.7% 100.0%
McDonel |Count 155 1 16 271
% 57.2% 42.8% 100.0%
Bryan Count 701 63 133
% 52.6% 47.4% 100.0%
Akers Count 189 106 295
% 64.1% 35.9% 100.0%
Shaw Count 1 56 62 2 18
% 71.6% 28.4% 100.0%
Holden |Count 169 83 252
% 67.1% 32.9% 100.0%
Bailey Count 62 37 99
% 62.6% 37.4%j 100.0%
Holmes Count 224 190' 414
% 54.1% 45.9% 100.0%
Williams Count 4 l 25 66
% 62.1% 37.9% 100.0%
Emmons Count 81 25 106
% 76.4% 23.6% 100.0%
Mason-Abbot |Count 174 90 264
% 65.9% 34.1%I 100.0%
Armstrong Count 86 29I 1 15
% 74.8% 25.2% 100.0%
lCase Count 136 98 234
% 58.1% 41.9% 100.0%
Landon ICount 73 32 105
% 69.5% 30.5% 100.0%
Yakeley-Gilchrist Count 113 54 167
% 67.7% 32.3% 100.0%
Campbell Count 62 23 85
% 72.9% 27. 1% 100.0%

 

 

 

 

 

166

 

Wilson

 

 

 

 

 

 

 

 

 

 

Count 164 56 2201

% 74.5% 25.5% 100.0%
IMayo |Count 42 13 55

% 76.4%. 23.6% 100.0%
Wonders Count 143 66 209F

% 68.4% 3 1 .6% 100.0%
Total Count 2419 1350 3769

% 64.2% 35.8% 100.0%
Table 4.88 (cont’d)

The chi-square value for this variable is significant at less than .0005 (Table 4.89).
Therefore, we can reject the null hypothesis that students are equally likely to hold

intellectual discussions with peers of a different racial background outside of class.

Table 4.89 Chi-Square Tests

 

 

 

 

 

Value I (If Asymp. Sig. (2-sided) I
Pearson Chi-Square 82.01 1(a) 20 .000
Likelihood Ratio 82.664 20 .000
Linear-by-Linear Association 21.778 1 .000
N Of Valid Cases 3769

 

a 0 cells (.0%) have expected count less than 5. The minimmn expected count is 19.70.

Results from the one-way analysis of variance indicate that the racial composition
of the residence hall is an important predictor for whether students will hold intellectual
discussions with peers Of a different racial background outside of class (see Table 4.90).
When we test the null hypothesis, the Observed significance level for the comparison of
means between halls is less than .005, suggesting that there is a significant relationship

between these two variables.

Table 4.90 Intellectual Discussions Bivariate ANOVA

 

 

 

Sum Of Squares df IMean Square F Sig.I
Between Groups 18.853 20 .943 4.168 .000
Within Groups 847.597 3748 .226
Total 866.450l3768

 

 

 

 

 

167

I attend events with students of a different racial background than myself.

For the ninth item Of the Interactional Diversity scale, more than half Of the

respondents describe attending events with peers of a different racial background

occasionally or never (Table 4.91). Of the participants who responded to this question

(n=3,901), 14.5 percent describe never engaging in this activity and an additional 43.9

percent describe attending events with peers of a different racial background

occasionally. However, a sizeable minority Of respondents report attending events with

students of a different racial background often or very Often (27.0% and 14.5%

respectively). The chi-square test for this variable is Significant at a value of less than

.0005 which suggests that there are meaningful differences between students for this

measure (Table 4.92).

Table 4.91 Attend Events Different Racial Backgrounds

 

 

 

 

 

I Frequency Percent Valid Percent

ever 567 12.4 14.5
Occasionally 1714 37.5 43.9
Often 1053 23.0} 27.0
Very Often 567 12.4 14.5
Total 3901 85.2 100.0
Missing 675 14.8 I
Total 4576 100.01

 

 

 

 

Table 4.92 Attend Events Different Racial Backgrounds: Chi-Square Test

 

 

Frequencies
r Observed N Expected N Residual

ever 567 975.2 -408.2
Occasionally 1714 975.2 738.8
Often 1053 975.2 77.8
Very Often 567 975.2 -408.2
Total 3901

 

 

 

 

168

 

_Test Statistics .

 

 

 

 

L Attend Events Different Racial Backgrounds I
Chi-Square 907.596(a)
f 3
symp. Sig. .000
Table 4.92 (cont’d)

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 975.3.

Attend Events with Students of Different Racial Background: Race/Ethnicigy
Crosstabulation results suggest that race plays a role in the likelihood that students
will engage in this activity (see Table 4.93). More than 60 percent of White students
describe attending events with students Of a different racial background occasionally or
never making them the least likely of all groups to participate in this activity. Black
students were less likely than other non-White students to attend events with students of a
different racial background than themselves: more than half report never or only
occasionally engaging in this activity. More than two-thirds of Native American and
Mixed Race students report attending events with students of a different racial
background than themselves Often or very often. The chi-square for this variable is
significant at .000; therefore, we reject the null hypothesis that participants are equally

likely to attend events with students from a different racial background than themselves

(see Table 4.94).

169

Table 4.93 Attend Events with Students of Different Racial Backgrounds *
Race/Ethnicity Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Attend Events with Race/Ethnicity 7
Students of Different Native Mixed
Racial Backgrounds White/Caucasian Black Latino AmericanI Asian Race Total
Never ICount 457 38 1 1 01 27 9 542
% 15.5% 14.3% 9.4% .0% 9.3% 11.7% 14.6%
Occasionally Count 1367 104 37 3 106 17 1634
% 46.4% 39.1% 31.6% 30.0% 36.6% 22. 1% 44.1%
IOften Count 764 77 39 4 74 28 986
% 25.9% 28.9% 33.3% 40.0% 25.5% 36.4% 26.6%
Very Often Count 361 47 30 3 83 23 547
% 12.2% 17.7% 25.6% 30.0% 28.6% 29.9% 14.7%
Total Count 2949 266 1 17 10 290 77 3709
% 100.0% 100.0% 100.0% 100.0% 100.0%I 100.0% 100.0%I
Table 4.94 Chi-Square Tests
Value df Asymp. Sig. (2-sided) 1
Pearson Chi-Square 1.145E2 15 .000
Likelihood Ratio 106.827 15 .000
Linear-by-Linear Association 80.315 1 .000
N of Valid Cases 3709

 

 

 

 

 

a 4 cells (16.7%) have expected count less than 5. The minimum expected count is 1.46.

Attend Events with Students of Different Racial Background: Gender

Crosstabulation results for this variable reveal that men and women are very

similar in their likelihood to engage in this activity (Table 4.95). Approximately 14

percent of men and women describe never attending events with students Of a different

racial background than themselves. Men and women were most likely to report engaging

in this activity occasionally with more than 40 percent of participants in each group

selecting this answer. However, more than a quarter of students in both groups describe

attending events with students of a different racial background Often. The chi-square

170

value for this tabulation is greater than .05 indicating that we cannot reject the null

hypothesis that men and women are equally likely to engage in this activity (Table 4.96).

Table 4.95 Attend Events with Students of Different Racial Backgrounds * Gender

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crosstabulation
Attend Events with Students of Gender
Different Racial Backgrounds Male Female Total
ever ICount 184 369 553
% 14.0% 14.7% 14.5%
Occasionally Count 591 1100 1691
% 44.9% 43.9% 44.2%
IOften Count 363 660T 1023
% 27.6% 26.3% 26.8%
Very Often Count 179 378 557
% 13.6% 15.1% 14.6%
Total Count 1317 2507 3824
% 100.0% 100.0% 100.0%
Table 4.96 Chi-Square Tests
Value Idf Asymp. Sig. (2-sided)
Pearson Chi-Square 2.331(a) 3 .507
Likelihood Ratio 2.345 3 .504
Linear-by-Linear Association .102 1 .749
N Of Valid Cases 3824

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 190.46.

Attend Events with Students of Different Racial Background: Class Standing

Crosstabulations on this variable suggest that upperclassmen are more likely to
attend events with students of a different racial background than freshmen (Table 4.97).
Freshmen are the least likely to engage in this activity: 16.3 percent of freshmen report
never attending events with students Of a different racial background compared with 14.7
percent of sophomores, 11.8 percent of juniors, and 8.4 percent Of seniors. In contrast,

seniors are the most likely to attend events with students Of a different racial background:

171

19.1 percent Of seniors describe engaging in this activity very Often, compared with 16.6

percent Of juniors, 14.2 percent of sophomores, and 13.5 percent Of freshmen. For all

groups, however, participants were most likely to report engaging in this activity

occasionally. The chi-square results for this tabulation are less than .0005; therefore, we

can reject the null hypothesis that students are equally likely to attend events with peers

of a different racial background, regardless of their academic standing (Table 4.98).

Table 4.97 Attend Events with Students of Different Racial Backgrounds * Class

Standing Crosstabulation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Attend Events with Students Class Standing
of Different Racial
Backgrounds Freshman Sophomore Junior Senior Total
Never Count 288 1 77 62 26 553
% 16.3% 14.7% 11.8% 8.4% 14.5%
Occasionally Count 785 552 212 128 1677
% 44.5% 45.7% 40.5% 41.4% 44.1%
IOften Count 452 307 163 96 1018
% 25.6% 25.4% 31.1% 31.1% 26.8%
Very Often Count 239 171 87 59 556
% 13.5% 14.2% 16.6% 19.1% 14.6%
Total Count 1764 1207 524 309 3804
% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 4.98 Chi-Square Tests 7
Value df Asymp. Sig. (2-sided) J
Pearson Chi-Square 32.059(a) 9 .000
Likelihood Ratio 32.907 9 .000
Linear-by-Linear Association 25.192 1 .000
of Valid Cases 3804

 

 

 

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 44.92.

172

 

Attend Events with Students of Different Racial Backgrounds: Current Hall Percent
m

Crosstabulation results for this variable suggest there are some differences
between respondents that may be related to their current residence hall (see Table 4.99).
Residents living in Emmons Hall, followed by those living in Wilson Hall, were the least

likely to attend events with students of a different racial background: for both halls, more

than two-thirds of respondents describe never or only occasionally engaging in this

activity. More than 90 percent Of the residents in Emmons Hall are freshmen. Over 40

percent of residents in Wilson Hall are upperclassmen; however, nearly 90 percent of its

residents are White. Residents in Rather and Holmes halls were the most likely to report

attending events with students of a different racial background: 48.5 percent of

respondents living in Rather Hall and 48.4 percent of respondents living in Holmes Hall

describe engaging in this activity Often or very Often. As discussed previously, Rather

Hall has the greatest percentage of students of color and the largest living-learning

program on campus is located within Holmes Hall.

Table 4.99 Current Hall Percent White without Van Hoosen * Attend Events With
Bivariate Crosstabulation

 

 

 

 

 

 

 

lCurrent Hall Percent White without Attend Events With bivariate 7
Van Hoosen Never or Occasionally Often or Very Often I Total

Rather Count 53 50 103
% 51.5% 48.5% 100.0%
Hubbard Count 167 1 12 279
% 59.9% 40.1% 100.0%
Butterfield Count 48 33 81
% 59.3% 40.7% 100.0%
McDonel Count 147 125 272
% 54.0% 46.0% 100.0%
Bryan Count 71 62 133
% 53.4% 46.6% 100.0%

 

 

 

 

 

 

173

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Akers Count 163 133 296
% 55. 1% 44.9% 100.0%
Shaw Count 1 39 8 1 220
% 63.2% 36.8% 100.0%
Holden Count 159 91 250
% 63.6% 36.4% 100.0%
Bailey Count 56 42 98
% 57.1% 42.9% 100.0%
WHolmes Count 214 201 415
% 51 .6% 48.4% 100.0%
Williams Count 37 29 66
% 56.1% 43.9% 100.0%
Emmons Count 72 34 106
% 67.9% 32.1% 100.0%
Mason-Abbot Count 160 105 265
% 60.4% 39.6% 100.0%
Armstrong Count 70 45 1 15
% 60.9% 39. 1% 100.0%
iCase Count 137 96 233
% 58.8% 41.2% 100.0%
Landon Count 61 44 105
% 58. 1% 41 .9% 100.0%
Yakeley-Gilchrist Count 99 67 166
% 59.6% 40.4% 100.0%
ICampbell Count 55 30 85
% 64.7% 35.3% 100.0%
Wilson Count 148 73 221
% 67.0% 33.0% 100.0%
Mayo Count 36 19 55
% 65.5% 34.5% 100.0%
Wonders Count 120 87 207
% 58.0% 42.0% 100.0%
Total Count 2212 1559 3771
% 58.7% 41 .3% 100.0%

Table 4.99 (cont’d)

Results from the chi-square are less than .05 indicating that we can reject the null

hypothesis that students are equally likely to attend events with peers of a different racial

background, regardless of their current residence hall (Table 4.100).

Table 4.100 Chi-Square Tests

 

 

 

 

 

Value df I Asymp. Sig. (2-sided)
Pearson Chi-Square 34.178(a) 20 .025
Likelihood Ratio 34.359 20 .024
Linear-by-Linear Association 5.620 1 .018
N Of Valid Cases 3771

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 22.74.

Results from the one-way analysis Of variance indicate that the racial composition
of the residence hall is an important predictor for whether students will attend events with
peers of a different racial background (see Table 4.101). When we test the null
hypothesis, the Observed significance level for the comparison of means between halls is
less than .005, suggesting that there is a significant relationship between these two

variables.

Table 4.101 Attend Events With Bivariate ANOVA
Sum Of Squares df Mean Square F Sig.

 

 

Between Groups 3.681 2 1.840 7.613 .001
Within Groups 909.257 3761 .242
Total 912.938 3763

 

 

 

 

 

 

 

 

I attend events sponsored by other racial/ethnic groups that differ from my own
racial/ethnic background.

For the tenth and final variable Of the Interactional Diversity scale, the vast
majority of respondents describe attending events sponsored by peers Of a different racial
background occasionally or never (Table 4.102). Of the participants who responded to
this question (n=3,903), 40.5 percent describe never engaging in this activity and an
additional 42.1 percent describe attending events sponsored by other racial groups on

campus occasionally. A small number of respondents report attending such events often

175

or very often (11.4% and 5.9% respectively). The chi-square test for this variable is
significant at a value of less than .0005 which suggests that there are meaningful

differences between students for this measure (Table 4.103).

Table 4.102 Attend Events Sponsored by Other Racial Groups

 

 

 

 

I Frequency Percent Valid Percent
ever 1584 34.6 40.6
Occasionally 1642 3 5 .9 42.1
Often 446 9.7 1 1.4
Very Often 231 5.07 5.9
Total 3903 85.3 100.0I
Missing 673 14.7
Total 4576 100.0 I

 

 

 

 

 

Table 4.103 Attend Events Sponsored Other Racial Groups: Chi-Test Frequencies

 

 

 

 

 

 

 

 

 

I Observed N Expected N Residual
ever 1584 975.8 608.2
Occasionally 1642 975.8 666.2
Often 446 975.8 -529.8
ery Often 231 975.8 -744.8
Total 3903
Test Statistics 7
I Attend Events Sponsored Other Racial Groups I
Chi-Square 1690. l 30(a)
(If 3
Asymp. Sig. .000

 

 

a 0 cells (.0%) have expected frequencies less than 5. The minimum expected cell
frequency is 975.8.

Attend Events Sponsored by Other Racial Groups: Race/Ethnicigy

Crosstabulation results for this variable indicate significant racial differences (see
Table 4.104). Nearly half of all White students describe never attending such events and

an additional 40 percent report only attending these events occasionally. Native American

176

students were the only other racial/ethnic group whose most frequent response on this

variable was “Never”. For all other groups, the majority of students report attending these

events occasionally. Among non-White students, Latinos and African Americans were

the least likely to attend these events, with approximately 70 percent of students in both

groups never or only occasionally attending events sponsored by racial/ethnic groups

different from their own. The chi-square for this variable is significant at .000; therefore,

we reject the null hypothesis that participants are equally likely to attend events

sponsored by students from a different racial/ethnic background than themselves (see

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 4.105).
Table 4.104 Attend Events Sponsored by Other Racial Groups * Race/Ethnicity
Crosstabulation
Attend Events Race/Ethnicity
Sponsor ed by Other Native Mixed
Racial Groups White/Caucasian Black Latino American Asian Race Total
Never Count 1342 58 4 60I 17 1503
% 45.5% 21 .7% 18.8% 40.0% 20.5% 22.1% 40.5%I
Occasionally Count 1219 127 2 129 33 1571
I % 41 .3% 47.6% 52.1% 20.0% 44.2% 42.9% 42.3%
IOften Count 267 59 3 57 14 418I
% 9.1% 22. 1% 15.4% 30.0% 19.5% 18.2% 11.3%
Very Often Count 122 23 l 46 13 221
% 4.1% 8.6% 13.7% 10.0% 15.8% 16.9% 6.0%
Total Count 2950 267 10 292 77 3713
% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0"/fl
Table 4.105 Chi-Square Tests
Value df Asymp. Sig. (2-sided) I
Pearson Chi-Square 2.57 3E2 15 .000
Likelihood Ratio 240.683 15 .000
Linear-by-Linear Association 184.068 1 .000
N of Valid Cases 3713

 

 

 

 

 

a 5 cells (20.8%) have expected count less than 5. The minimum expected count is .60.

177

Attend Events Sponsored by Other Racial Groups: Gender

Crosstabulation results suggest that majority of men and women do not attend
events on campus sponsored by racial groups different from their own (Table 4.106).
Men were somewhat less likely to attend these events: 42.5 percent Of men describe
never attending events sponsored by other racial groups on campus compared with 39.7
of women. Similarly, 43.2 percent of women report attending these events occasionally
compared with 39.9 percent of men. The chi-square value for this tabulation is greater
than .05; therefore, we cannot reject the null hypothesis that men and women are equally

likely to attend events sponsored by racial groups different from their own (Table 4.107).

Table 4.106 Attend Events Sponsored by Other Racial Groups * Gender
Crosstabulation

 

 

 

 

 

 

 

 

Attend Events Sponsored by Other Gender
Racial Groups Male Female Total
Never ICount 559 996 1555
% 42.5% 39.7% 40.6%
Occasionally Count 525 1086 1 61 1
% 39.9% 43.2% 42.1%
|Often Count 153 282 435
% 11.6% 11.2% 11.4%
Very Often Count 79 147 226
% 6.0% 5.9% 5.9%
Total Count 1316 251 1 3827
% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

Table 4.107 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 4.142(a) 3 .247
Likelihood Ratio 4.149 3 .246
Linear-by-Linear Association .534 1 .465
1‘1 of Valid Cases 3827

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 77.72.

178

 

Attend Events Sponsored by Other Racial Groups: Class Standing

Crosstabulations reveal the likelihood that students will attend events sponsored
by racial groups different from their own increases with academic standing (Table 4.108).
Freshmen are the least likely to attend these events: 46.1 percent Of freshmen report never
attending such events, compared with 38.8 percent of sophomores, 34.5 percent of
juniors, and 28.1 percent of seniors. Seniors were the most likely to attend these events:
28.7 percent Of seniors describe attending events sponsored by racial groups different
from their own Often or very Often compared with 23.5 percent Of juniors, 16.7 percent Of

sophomores, and 13.7 percent Of freshmen.

Table 4.108 Attend Events Sponsored by Other Racial Groups * Class Standing
Crosstabulation

 

 

 

 

 

 

 

 

 

 

Attend Events Sponsored Class Standing
by Other Racial Groups Freshman I Sophomore I Junior Senior Total
Never Count 8101 4701 181 88 1549
% 46.1% 38.8% 34.5% 28.1% 40.7%
Occasionally Count 706 538 221 135 1600
% 40.2% 44.5% 42.1% 43.1% 42.0%
IOften Count 166 129 76 59 430|
% 9.4% 10.7% 14.5% 18.8% 11.3%
Very Often Count 76 73 47 3 1 227
% 4.3% 6.0% 9.0% 9.9% 6.0%
Total Count 1758 1210 525 313 3806
% 100.0% 100.0% 100.0% 100.0% 100.0%

 

 

 

 

 

 

 

 

The chi-square value for these variables is less than .0005; therefore, we can

reject the null hypothesis that students are equally likely to attend such events regardless

of their class standing (Table 4.109).

Table 4.109 Chi-Square Tests

 

 

 

 

 

 

Value df Asymp. Sig. (2-sided) I
Pearson Chi-Square 84.179(a) 9 .000
Likelihood Ratio 81.134 9 .000
Linear-by-Linear Association 77.657 1 .000
N of Valid Cases 3806

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 18.67.

Attend Events Sponsored by Other Racial Groups: Current Hall Percent White

Crosstabulations on this variable suggest that current residence hall may play a

role in students’ likelihood to attend events sponsored by racial groups different from

their own (Table 4.110). Residents living in Armstrong Hall, followed by residents in

Wilson and Mayo halls, are the least likely to report engaging in this activity. More than

90 percent of respondents from each hall describe never or only occasionally attending

events sponsored by other racial groups on campus. Armstrong Hall has the highest

percentage of freshmen residents on campus. As discussed previously, nearly 60 percent

of the residents in Wilson Hall are freshmen. Only 30 percent of residents in Mayo Hall

are freshmen. However, Wilson and Mayo Hall have a high percentage of White

residents compared with other halls on campus. Residents living in Rather and McDonel

halls were the most likely to report attending events sponsored by racial groups different

from their own. For both halls, more than 25 percent Of respondents describe attendng

these events Often or very Often. Although Rather Hall has a large number Of freshmen

residents (88%), it also has the highest percentage of students of color on campus. More

than 80 percent Of the residents in McDonel Hall are upperclassmen, it is one of the more

racially diverse halls on campus, and it has the highest percentage of international and

transfer residents on campus.

180

Table 4.110 Current Hall Percent White without Van Hoosen * Attend Events
Sponsored Bivariate Crosstabulation

 

lCurrent Hall Percent White

Attend Events Sponsored Bivariate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

without Van Hoosen Never or Occasionally Often or Very Often Total

Rather |Count 76 27 103
% 73.8% 26.2% 100.0%

Hubbard Count 236 45 28 1
% 84.0% 16.0% 100.0%

Butterfield ICount 69 12 81
% 85.2% 14.8% 100.0%

McDonel Count 203 69 272
% 74.6% 25.4% 100.0%

Bryan Count 107 26 133
% 80.5% 19.5% 100.0%

Akers Count 238 58 296
% 80.4% 19.6% 100.0%

Shaw Count 1 86 33 2 19
% 84.9% 15. 1% 100.0%
Holden Count 206 44 2501
% 82.4% 17.6% 100.0%

Bailey Count 87 l l 98
% 88.8% 1 1.2% 100.0%

Holmes Count 33 1 82 41 3
% 80.1% 19.9% 100.0%

Williams Count 52 1 3 65
% 80.0% 20.0% 100.0%

Emmons Count 9 l 15 106
% 85.8% 14.2% 100.0%

Mason-Abbot Count 212 53 265
% 80.0% 20.0% 100.0%

Armstrong Count 105 10' 1 15
% 91.3% 8.7% 100.0%

ICase Count 1 89 46 235
% 80.4% 19.6% 100.0%

Landon 1Count 92 13 105
% 87.6% 12.4% 100.0%

Yakeley-Gilchrist Count 139 28 167
% 83.2% 16.8% 100.0%

Campbell Count 75 10 85
% 88.2% 1 1.8% 100.0%

 

181

 

 

Count

 

 

 

 

 

 

 

 

 

 

Wilson 199 21 22d
% 90.5% 9.5% 100.0%
Mayo [Count 50I 5 55
% 90.9% 9.1% 100.0%
Wonders Count 1 75 33 208
% 84.1% 15.9% 100.0%
Total Count 3 1 18 654 3772
% 82.7% 17.3% 100.0%
Table 4.110 (cont’d)

The chi-square value for this measure is less than .0005 (Table 4.111). As a result,
we can reject the null hypothesis that students are equally likely to attend events
sponsored by racial groups different from their own, regardless Of their current residence

hall.

Table 4.111 Chi-Square Tests

 

 

 

 

Value I (if I Asymp. Sig. (2-sided)
Pearson Chi-Square 50.409(a) 20 .000
Likelihood Ratio 52.140 20 .000
Linear-by-Linear Association 13.590 1 .000
N Of Valid Cases 3772

 

 

a 0 cells (.0%) have expected count less than 5. The minimum expected count is 9.54.
Results from the one-way analysis of variance indicate that the racial composition
of the residence hall may not be an important predictor for whether students will attend
events sponsored by peers of a different racial background (see Table 4.112). When we
test the null hypothesis, the Observed significance level for the comparison of means
between halls is greater than .05 suggesting that there is not a significant relationship

between these two variables.

182

Table 4.112 Attend Events Sponsored Bivariate ANOVA
Sum Of Squares df Mean Square F Sig.

 

 

 

 

 

 

 

 

 

Between Groups .820 2 .410 2.863 .057
Within Groups 538.924 3762 .143
Total 539.744 3764

 

Interactional Diversity Scale Analysis

After reviewing the individual items within the Interactional Diversity scale by
participants’ racial/ethnic background, gender, class standing, and current residence hall,
we will now consider the scale as a whole.

Interactional Diversity Scale: Race/Ethnicigy

A one-way analysis Of variance test was conducted on the interactional scale
using race as the factor (see Table 4.113). Scale means indicate that Mixed Race students
(27.3) were the most likely to participate in meaningful cross-racial interactions, followed
by Chicano/Latino students (26.7), Asian students (26.5) and Native American students
(26.5). White/Caucasian students (21.4), were the least likely to engage in interactional
diversity, followed by Black students (23.5). Results from the analysis suggest that the
racial/ethnic background of participants is an important determinant for interaction.

The observed significance level for the comparison Of means between
racial/ethnic groups is less than .0005, indicating that we can reject the null hypothesis
that the likelihood Of interactional diversity will be equal across racial/ethnic groups
(Table 4.114). This suggests that there is a significant relationship between these two

variables.

183

Table 4.113 Interactional Diversity Scale Descriptives: Race/Ethnicity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

95% Confidence
Interval for Mean
Race/ Std. Std. Lower Upper
Ethnicity N Mean Deviation Error Bound Bound Minimum Maximum
White!
Caucasian 2978 21.3753 6.69139 .12262 21.1348 21.6157 10.00 40.00
Black 268 23.4660 7.01415 .42846 22.6624 24.3096 10.00 40.00
Chicano/Latino 118 26.6893 7.09014 .65270 25.3966 27.9819 10.00 40.00
Native
American 10 26.5000 6.07819 1.92209 22.1519 30.8481 20.00 36.00
Asian 293 26.5465 7.31551 .42738 25.7053 27.3876 11.00 40.00
Mixed Race 78 27.3077 7.35255 .83251 25.6499 28.9654 10.00 40.00
Total 3745 22.2341 7.02973 .11487 22.0089 22.4594 10.00 40.00
Table 4.114 Interactional Diversity Scale AN OVA
Sum Of Squares df Mean Square F Sig.
Between Groups 12583.956 5 2516.791 54.573 .000
Within Groups 1724338 3739 46.118
Total 185017.7 3744

 

 

 

 

 

 

 

 

Interactional Diversify Scale: Gender

After reviewing the role of participants’ gender on the individual items within the

Interaction Scale, we now turn to the analysis of the entire scale. One-way analysis of

variance for the interactional scale using gender as the factor suggests that men and

women have similar levels Of cross-racial engagement: the mean score for men (22.8)

was slightly higher than women (22.0) (see Table 4.115). However, the Observed

significance level for the comparison of means between men and women is less than

.0005 (see Table 4.116). Therefore, we reject the null hypothesis that the likelihood of

interactional diversity will be equal for both men and women.

Table 4.115 Interactional Diversity

Scale Descriptives: Gender

 

 

 

 

95% Confidence Interval
for Mean
Std. Std. Lower Upper
Gender N Mean Deviation Error Bound Bound Minimum Maximum
Male 1331 22.7804 6.95579 . 19066 22.4604 23.1544 10.00 40.00
Female 2533 21.9309 7.03710 .13982 21.6567 22.2050 10.00 40.00
Total 3864 22.2235 7.01992 .1 1293 22.0021 22.4449 10.00 40.00

 

 

 

 

 

 

 

 

 

184

 

 

 

Table 4.116 Interactional Diversity Scale AN OVA
Sum Of Squares df Mean Square F Sig.
Between Groups 629.697 1 629.697 12.817 .000
Within Groups 1897359 3862 49.129
Total 1903656 3863

 

 

 

 

 

 

 

 

Interactional Diversig Scale: Class Standing

The one-way analysis Of variance using class standing as the factor suggests that

interactional diversity increases with class standing (see Table 4.117). The mean score for

freshmen was the lowest (21.6), followed by sophomores (22.2) and juniors (23.0).

Seniors had the highest mean scores on this scale with 24.4.

Table 4.117 Interactional Diversity Scale Descriptives: Class Standing

 

 

 

95% Confidence

Interval for Mean
Class Std. Std. Lower Upper
Standing N Mean Deviation Error Bound Bound Minimmn Maximum
Freshman 1777 21.6257 6.7782 . 16081 21.3103 21.9411 10.00 40.00
sophomore 1225 22.2089 7.05131 .20147 21.8137 22.6042 10.00 40.00
Junior 525 22.9738 7.45506 .32537 22.3346 23.6129 10.00 40.00
Senior 317 24.3680 7.03887 .39534 23.5902 25.1458 10.00 40.00
T0t8| 3844 22.2218 7.02469 .1 1330 21.9997 22.4439 10.00 40.00

 

 

 

 

 

 

 

 

 

 

The Observed significance level for the comparison of means between freshmen,

sophomores, juniors and seniors is less than .0005 leading us to reject the null hypothesis

that the likelihood of interactional diversity will be equal across groups (Table 4.118).

Once again we find that class standing is an important determinant.

 

 

Table 4.118 Interactional DiveristV Scale AN OVA
Sum Of Squares df Mean Square F Sig.
Between Groups 2388.705 3 796.235 76.329 .000
Within Groups 187249.] 3840 48.763
Total 1896378 3843

 

 

 

 

 

 

 

 

185

 

Interactional Diversity Scale: First-Year Status

The one-way analysis of variance using first-year status as the factor confirms that
interactional diversity increases with time spent on campus (see Table 4.119). While first-
year students exhibited a mean score of21.7 on the interaction scale, returning students
revealed a mean score Of 22.8.

Table 4.119 Interactional Diversity Scale Descriptives: First-Year Status

 

95% Confidence
Interval for Mean

Std. Std- Lower Upper
N Mean Deviation Error Bound Bound Minimum Maximum

 

 

 

Yes 1948 21.6598 6.80955 .15429 21.3572 21.9624 10.00 40.00
NO 1897 22.7701 7.21323 .1656] 22.4453 23.0949 10.00 40.00
Total 3845 22.2076 7.03265 .11342 21.9852 22.4300 10.00 40.00

 

 

 

 

 

 

 

 

 

 

The Observed significance level for the comparison Of means between first-year
and returning students is less than .0005 leading us to reject the null hypothesis that the
likelihood of interactional diversity will be equal across groups (Table 4.120). Once again
we find that class standing is an important determinant.

Table 4.120 Interactional Diversity Scale ANOVA

 

Sum of Squares df Mean Square F Sig.

 

Between Groups 1184.738 1 1184.738 24.098 .000
Within Groups 188932.580 3843 49.163
Total 190117.318 3844

 

 

 

 

 

 

 

 

Interactional Diversin Scale: Current Hall Percent White

The one-way analysis Of variance for the interactional scale and current hall does
not present a clear pattern of interaction (see Table 4.121). In general, mean scores for
interaction decrease as the percentage of White residents in the hall increases. For
example, the mean scores in those halls with the most students of color hover near 23

points while the mean scores for balls with the greatest number on non-Hispanic White

186

students are between 20 and 21. The highest mean score appears in Rather Hall (24.4),
which has the lowest percentage of White residents in any Of the halls (49.4%). Holmes
Hall, which is 79.0 percent White, has the second highest mean on the interaction scale
(24.3). This hall is unique on campus in that it houses the Lyman-Briggs Living-Learning
Program: 86 percent of its residents are in the College Of Natural Science, many of whom
are also enrolled in the Honors College. McDonel Hall, which is 68.2 percent White, has
the third highest mean on the interaction scale (23.9). This hall is unique for two reasons.
First, McDonel Hall houses two special Living-Leaming Programs, La Casa, and the
lntemational Culture and Language Residence. As a result, it contains the largest number
Of international students of any of the residence halls included in this study. Second,
twenty percent of all transfer students at Michigan State University reside in McDonel
Hall, giving it the largest number Of transfer residents of any residence hall.

The lowest mean on the interaction scale appears in Emmons Hall (20.0) followed
by Wilson Hall (20.3). Emmons Hall is 84.0 percent White and has the second highest
percentage of freshmen residents of any hall on campus (92%). Wilson Hall is one Of the
least racially diverse halls on campus and nearly 60 percent of its residents are freshmen.

The Observed significance level for the comparison Of means between residence
halls is less than .0005 leading us to reject the null hypothesis that the likelihood Of
interactional diversity will be equal across halls (see Table 4.122). This suggests that
patterns of interaction are related to the racial demographics of students’ current

residence hall and that this relationship Should be explored further.

187

       

otal

Table 4.122 Interactional Diversity Scale ANOVA

Table 4.121 Interactional

Std.
Deviation
7.69
7.
6.
7

8.49

6.

7
6.65401
7

5.

7.2

7.1 .
7.03231 .1

Scale

.74461

Current Hall Percent White

95% Confidence Interval for

Std.
Error

22.

1.0452
.593

19.1

Mean

Lower Bound

20.6151

21.6851

Bound

23.

24.
24.6901
23.
22.4

21.

24.6531
21.1
22.5531
21

23.

22.0

21.0471

22.

 

 

 

 

 

 

 

18mm of Squares: df Mean Square F Sig.
Between Groups 6696.111 20 334.806 6.983 .000
Within Groups 181523.508 3786 47.946
Total 188219.619 3806

 

 

 

Interactional Diversity Scale: Current Hall Assignment

Minimum

11

The one-way analysis of variance for the interactional scale based on hall

assignments suggests that students who requested to live in their current hall are less

likely to engage in interactional diversity than students who are required to live in their

188

 

current hall, due to participation in a living-leaming program, and students who were

randomly assigned to their current hall (see Table 4.123). Students who requested their

current hall had a mean score Of 21.7 on the Interactional Diversity scale compared with a

mean score of 22.6 for those students randomly assigned to their hall. Students

participating in a living-learning program and, therefore, required to live in their current

hall had the highest mean on the Interactional Diversity scale (24.1). The observed

significance level for the comparison of means between students required to live in their

current hall and those who are not is less than .0005 (see Table 4.124). AS a result, we

reject the null hypothesis that the likelihood of interactional diversity will be equal for all

residents.

Table 4.123 Interactional Diversity Scale Descriptives: Current Hall Assignment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

95% Confidence
Interval for Mean
N Mean Deviation Error Bound Bound Minimum Maxrmum
Hall Required 559 24.0915 7.09471 .30007 23.5021 24.68091 10.001 40.00
Rammed 2505 21.7292 6.95942.13905 21.4565 22.0019 10.00 40.00
Current Hall
Did Not
Request 926 22.6236 7.08332 .23277 22.1667 23.0804 10.00 40.00
Current Hall
Total 3990 22.2677 7.05382 .11167 22.0488 22.4867 10.00 40.00
Table 4.124 Interactional Diversity Scale ANOVA
Sum of Squares df Mean Square F Sig.
Between Groups 2703.145 2 1351.573 27.525 .000
Within Groups 195774.958 3987 49.103
Total 198478.103 3989

 

 

 

 

 

 

 

189

Interactional Diversify Scale: First-Year Roommate

One-way analysis of variance was conducted to test the null hypothesis that

participants are equally like engage in interactional diversity regardless Of whether they

had requested their first-year roommate or the racial background of that roommate (see

Table 4.125). Results suggest that students who requested their first-year roommate, and

who share the same racial background with that roommate, are the least likely to engage

in meaningful and sustained interactions across race lines; these students had a mean

score Of 20.2 on the Interactional Diversity scale. With a mean score Of 21.4, students

who did not request their first-year roommate, and who share the same racial background

with that roommate, were somewhat more likely to engage in interactional diversity.

Students whose first-year roommate is Of a different racial background report higher

levels of interactional diversity, particularly if they requested that roommate. Students

who were randomly assigned their first-year roommate, and whose roommate is of a

different racial background, had a mean score of 24.6 on the Interactional Diversity scale.

Students who requested their first-year roommate, and whose roommate is of a different

racial background had the highest mean (25.7) on the Interactional Diversity scale.

Table 4.125 Interactional Diversity Scale Descriptives: First-Year Roommate

 

Std.

Std.

95% Confidence
Interval for Mean

 

 

 

 

 

 

 

 

 

 

 

. . Lower Upper . . .

N Mean Devratlon Error Bound Bound Mmlmum Max1mum
Requesmd's‘y’ 858 20.1504 6.20919 .21198 19.7344 20.5665 10.00 40.00
roommate/same race
Reques‘edlfty’ 213 25.6695 7.67607 .52596 24.6327 26.7063 10.00 40.00
roommate/drfferent race
B'“““’" 1732 21.3797 6.74358 .16204 21.0618 21.6975 10.00 40.00
roommate/same race
B'““'“”'. 1046 24.5771 7.10554 .21970 24.1460 25.0082 10.00 40.00
roommate/dlfferent race
Total 3849 22.2120] 7.03155 .11334 21.9898 22.4342 10.00 40.00

 

190

 

The Observed significance level for this test is less than .0005 indicating that we
Should reject the null hypothesis that the likelihood Of engaging in interactional diversity

and the first-year roommate are independent (see Table 4.126).

Table 4.126 Interactional Diversity Scale ANOVA

Sum of Squares df Mean Square F Sig.I
Between Groups 13243.764 3 4414.588 95.892 .000
Within Groups 177011.762 3845 46.037
Total 190255.526 3848

 

 

 

 

 

 

 

 

 

Interactional Diversify Scale: Current Roommate (Upperclassmen)

One-way analysis of variance was conducted to test the null hypothesis that
upperclassmen are equally like engage in interactional diversity regardless of whether
they had requested their current roommate or the racial background Of that roommate (see
Table 4.127). Results indicate that the pattern among first-year roommates remains
consistent for upperclassmen. Students who requested their current roommate, and who
share the same racial background with that roommate, are the least likely to participate in
interactional diversity, with a mean score Of21.2 on the Interactional Diversity scale.
Upperclassmen who were randomly assigned their current roommate, and who share the
same racial background with that roommate were somewhat more likely to engage in
interactional diversity, with a mean score of 22.8. Upperclassmen whose current
roommate is of a different racial background were more likely to experience interactional
diversity. Students who were randomly assigned their current roommate had a mean

score Of 25.2 on the Interactional Diversity scale While students who requested their

191

current roommate, and whose roommate is of a different racial background, had the

highest mean score (27.0).

Table 4.127 Interactional Diversity Scale Descriptives: Current Roommate
(U pperclassmen)

 

95% Confidence
Interval for Mean

Std; Std- Lower Upper . . _
N Mean Devratlon Error Bound Bound Mmrmum Max1mum

 

 

 

Did not request current
roommate/Same race

 

649 22.7630 7.10396 .27885 22.2155 23.3106 10.00 40.00

Requested current
roommate/Same race

Did not request current
roommate/Different race

862 21.2200 6.72171 .22894 20.7707 21.6694 10.00 40.00

167 25.2029 7.27689 .56310 24.0912 26.3147 11.00 40.00

Requested current
roommate/Different race

Total 1897 22.7701 7.21323 .16561 22.4453 23.0949 10.00] 40.00

219 27.0370 7.17088 .48456 26.0820 27.9921 10.00 40.00

 

 

 

 

 

 

 

 

 

The Observed significance level for this test is less than .0005 indicating that we
should reject the null hypothesis that the likelihood Of upperclassmen engaging in

interactional diversity and the current roommate are independent (see Table 4.128).

Table 4.128 Interactional Diversity Scale AN OVA

Sum Of Squares df Mean Square F Sig.
Between Groups 7046.865 3 2348.955 48.542 .000
Within Groups 91603.394 1893 48.391
Total 98650.259 1896

 

 

 

 

 

 

 

 

 

 

Interactional Diversity Scale: Overall Racial Environment

One-way analysis of variance was conducted to test the null hypothesis that
participants are equally like engage in interactional diversity regardless of the overall
racial environment of their current residence hall (see Table 4.129). The observed
significance level for this test is less than .0005 indicating that we should reject the null

hypothesis. It appears that students who score highest on the Overall Racial Environment

192

scale score higher on the Interaction scale as well. Although this relationship is not linear,
it is significant.

Table 4.129 Interaction Scale ANOVA: Overall Racial Environment
Sum of Squares df Mean Square F Sig.

 

 

 

 

 

 

 

 

 

Between Groups 9107.781 49 185.873 3.843 .000
Within Groups 193637.599 4004 48.361
Total 202745.381 4053

 

Interactional Diversity Scale: Classroom Experience

One-way analysis Of variance was conducted to test the null hypothesis that
participants are equally like engage in interactional diversity regardless of their
experiences inside the classroom (see Table 4.130). The Observed significance level for
this test is less than .0005 indicating that we should reject the null hypothesis. It appears
that students who score highest on the Classroom Experience scale score higher on the
Interaction scale as well, and that this relationship is linear (see Table 4.131). While we
cannot determine a causal relationship between these variables with the data that is

available, results suggest that there is meaningful relationship between them.

Table 4.130 Interactional Diversity Scale ANOVA: Classroom Experience
Sum of Squares df Mean Square F Sig.

 

 

Between Groups 54074.387 22 2457.927 66.592 .000
Within Groups 148821.469 4032 36.910
Total 202895.856 4054

 

 

 

 

 

 

 

193

Figure 4.1 Classroom Experience Scale * Interactional Diversity Scale

3500 "'- Dot/Lines show Means

3000 -*

25. 00 1 . 1‘

into raction scale

20001

1500-1

1

TI’___‘ T# — [#417 T "if" ' "T
5 10 15 20

Classroom Experience Scale

Summagy

The item analysis of the Interactional Diversity scale reveals that students’
racial/ethnic background plays a significant role in predicting their participation in each
activity. White students were significantly less likely to describe engaging in the activity
than students of color on every item within the scale except for socializing or partying
with students of a different racial background. For this item, Black students were the least
likely to describe participating. Compared to all other students Of color, Black students
were the least likely to report engaging in any of the activities. Mixed Race tended to
have higher rates Of participation on each of the items within the Interactional Diversity

scale, followed by Latino and Native American students.

194

Gender is not a consistently significant predictor for items within the Interactional
Diversity scale. Men were significantly more likely than women to dine or share meals
with students of a different racial background, to visit in the rooms of students of a
different racial background, to visit with students of a different racial background in their
own room, socializing or partying with students of a different racial background, having
intellectual discussions outside of class with students of a different racial background.
Gender was not significantly related to having meaningful discussions about racial/ethnic
issues outside of class, sharing personal feelings and problems, studying or preparing for
class, and attending events with students of a different racial background. Gender was
also not a significant predictor of students who attend events sponsored by racial/ethnic
groups different from their own background.

Academic standing was significantly related to most of the items within the
Interactional Diversity scale. When academic standing was a significant predictor,
upperclassmen were typically more likely to participate in the activity than freshmen.
Academic standing was not significantly related to students’ likelihood to visit in the
rooms of residents of a different racial background or their likelihood to share personal
feelings and problems across race lines.

The current residence hall was significantly related to each of the items within the
Interactional Diversity scale. Residents living in Rather and Holmes Halls were more
likely than residents of other halls to participate in these activities. Halls with lower
levels of racial diversity and higher percentages of freshmen typically exhibited the

lowest mean scores for each item of the scale.

195

When we considered the effects of independent variables on the Interactional
Diversity scale, a number of significant relationships were discovered. Not surprisingly,
race was a significant predictor of students’ scores on the Interactional Diversity scale.
White students had the lowest mean score of all racial/ethnic groups. Among students of
color, Black students had the lowest mean score. Mixed Race students had the highest
mean score of all racial/ethnic groups.

Gender was also significantly related to the Interactional Diversity scale: the
mean score for men was significantly higher than the score for women.

Academic standing was significantly and positively related to the Interactional
Diversity scale, indicating that the likelihood of engaging in interactional diversity
increases as students progress through their academic careers. Similar results were found
when we compared first-year students with returning students. As a result, first-year
students will not be considered in the multivariate analysis. For the remainder of this
study, academic standing will be the only independent variable used to address students’
status.

The current residence hall was also significantly related to the Interactional
Diversity scale. Rather Hall had the highest mean score of all the residence halls,
followed by Holmes Hall. As discussed previously, Rather Hall has the highest
percentage of residents of color and Holmes Hall is where the largest living-learning
program on campus is located. McDonel Hall had the third highest mean score on the
Interactional Diversity scale. McDonel Hall has the highest percentage of international
students of any of the residence halls included in this study and houses two living-

learning programs: La Casa and the lntemational Culture and Language Residence.

196

Emmons Hall had the lowest mean score on the Interactional Diversity scale. Although
Emmons Hall is not one of the least racially diverse halls on campus, it has the second
highest percentage of freshmen on campus. Wilson Hall had the second lowest mean
score. Wilson Hall is one of the least racially diverse balls on campus and more than half
of its residents are freshmen. It is likely that the mean score for both of these halls is
reflective of both racial diversity as well as the academic standing of its residents.

Hall assignment and Interactional Diversity were significantly related. Students
who were required to live in their current hall because of their participation in a living-
learning program were significantly more likely to engage in interactional diversity than
others. Students who were randomly assigned their current hall exhibited higher levels of
engagement across race lines than those who requested to live in their current hall.
Because the interactional diversity of students participating in living-learning programs
can also be seen in the Current Hall variable, the hall assignment will not be included in
future analyses.

First-year roommates are significantly related to students’ levels of interactional
diversity. Students who requested their first-year roommate, and whose roommate shares
their same racial background, were the least likely to engage in interactional diversity.
Students who were randomly assigned their first-year roommate were more likely to
engage in interactional diversity, particularly if that roommate is of a different racial
background. Students who requested their first-year roommate, and whose roommate is
of a different racial background, were the most likely to participate in interactional
diversity. Similar results were found with upperclassmen with regard to their current

roommate. As a result, only the first-year roommate will be considered in future analyses.

197

The Overall Racial Environment was significantly and positively related to the
Interactional Diversity scale. Students are more likely to engage in interactional diversity
when they live in a hall where they observe little racial conflict, where they feel
emotionally and physically safe, and where they are comfortable with the racial
composition of the hall.

Finally, the Classroom Experience scale was significantly and positively related
to the Interactional Diversity scale. Students who report that their values and beliefs have
been challenged in class, and that have been given the opportunity to interact with peers
of a different racial background in class, are also more likely to engage in interactional
diversity outside of the class.

In the next chapter, we will consider the relationship between these independent
variables and the Interactional Diversity scale using multivariate regressions. We will
also consider the possibility that some of these independent variables are interacting with

one another.

198

CHAPTER 5: MULTIVARIATE REGRESSIONS

The bivariate analyses from the previous chapter revealed that each of the items
within the Interactional Diversity scale varies significantly among respondents.
Furthermore, mean scores on the scale were found to be significantly related to a number
of independent variables. We will now consider the role of these independent variables
when we control for additional factors. In this chapter, the correlations of meaningful
variables will be considered, a series of multiple regressions will be conducted, and the
possibility of interactions between variables will also be explored. Following these

analyses will be a discussion of the findings.
Correlations

A correlation matrix was calculated to ensure that the independent variables to be
included in the multivariate regressions are not highly correlated with each other. None
of the correlations were found to be out of the ordinary, indicating that there are not
conceptual or statistical overlaps between these variables. Therefore, the table will not be
included here. Now we’ll look at the correlation matrix that only contains the bivariate
correlations between the dependent variable—the Interactional Diversity scale—and all
of the independent variables (Table 5.1). The discussion following relates to this
correlation matrix. Many of the independent variables are significantly correlated with
the Interactional Diversity scale. The Pearson correlations for Black, Latino, Asian, and
Mixed Race students are positive and significant at less than .01. This suggests that
compared with students of a different racial background, these students are significantly
more likely to engage in interactional diversity. The correlation for gender is significant

at less than .01 and negative, suggesting that men are more likely than women to engage

199

in interactional diversity. The correlation with academic standing is positive and
significant at less than .01 which suggests that upperclassmen are more likely than
freshmen to engage in interactional diversity. Hours Socializing is significantly and
positively correlated with interactional diversity, indicating that students who spend more
time socializing with their peers are more likely to socialize across race lines.

Demographics for neighborhood, elementary, junior high, and high schools are
positively and significantly correlated at less than .01. This suggests that students who
grew up in more racially diverse neighborhoods, and students who attended more racially
diverse schools prior to attending Michigan State University, report higher levels of
interactional diversity. While parents’ education is not significantly correlated with
interactional diversity, family income is negatively and significantly related at less than
.01. This suggests that students fi'om more aflluent backgrounds are less likely to engage
in interactional diversity.

The racial composition of the current residence hall is negatively correlated with
interactional diversity and significant at less than .01. As a reminder, Current Hall is an
interval variable. Residence halls on the campus of Michigan State University have been
ranked according to the percentage of residents within each hall that are White. This
variable ranges from 1 (Rather Hall; 49.4% non-Hispanic White) to 21 (Wonders Hall;
89.9% non-Hispanic White). Van Hoosen Hall has been eliminated from this measure
because it does not accurately represent a residence hall environment: its rooms are
independent units which include their own kitchens and baths. This correlation suggests

that as the percentage of White residents increase, and the percentage of students of color

200

decrease, within the residence hall, students are less likely to engage in interactional
diversity.

The Overall Racial Environment scale is found to be significantly correlated with
the Interactional Diversity scale at less than .05 and this relationship is positive. The
Overall Racial Environment scale refers to conflicts within the residence hall related to
racial/ethnic differences as well as the level of comfort and safety the participant feels
living in the residence hall. Once again, Van Hoosen Hall has been omitted from this
scale. The items within the Overall Racial Environment scale were coded in a direction
to reflect a positive, conflict-free environment. Therefore, students who witness little
racial conflict within their hall, who feel emotionally and physically safe within their hall,
and who would recommend their current hall to fiiends of their same racial background,
are more likely to engage in interactional diversity.

Regarding the first-year roommate, students who did not request their first-year
roommate, and whose roommate is of their same racial background, are significantly less
likely to engage in interactional diversity than other students. In contrast, students whose
first-year roommate is of a different racial background are more likely to engage in
interactional diversity.

The Pearson correlation for the Classroom Experience scale, although it is
significant, is unusually high at .511. Looking at the individual items within this scale
and those within the Interactional Diversity scale, there does not appear to be any
conceptual overlap. For each of the items within the Classroom scale, the phrase “in my
classes” was used to encourage students to think only about those occasions that occur

within the classroom. For items on the Interactional scale which might also occur in a

201

classroom setting, such as intellectual discussions and discussions about racial/ethnic

issues, the phrase “outside of class” was included to discourage students from

considering occasions within the classroom. Despite these precautions, it is possible that

some participants conflated these items. As a result, the validity of these measures may

be have been somewhat compromised.

Table 5.1 Interactional Diversity Scale Correlations

 

Variable

Interactional Diversity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Scale
Interactional Diversity Scale 1.000
Black (0=White) .045“
Latino (0=White) .108**
Native American (0=White) .030
Asian (0=White) .169”
Mixed Race (0=White) .100M
Gender (0=Male) -.058**
Academic Standing .110“
Hours Socializing .153“
Neighborhood Demographics (0=Segregated) .122"
Elementary Demographics (0=Segregated) .101 **
JH/HS Demographics (0=Segregated) .108“
Father’s Education .003
Mother’s Education .019
FamiLy Income -.054**
Current Hall Percent White without Van Hoosen -.112**
Overall Racial Environment Scale without Van Hoosen .142**
1St year roommate: Blind/Same Race (0=Requested/Same) -.108**
lSt year roommate: Blind/Different Race .193"
(0=Requested/SamQ
1St year roommate: Requested/Different Race .1 13**
(0=Requested/Same)
Classroom Environment .511“

 

Regressions: Current Hall Percent White

Multivariate regressions were conducted to determine the significance of the

independent variables once we control for other factors. In Model 1, we consider the

202

 

relationship between interactional diversity and the residence hall (r2=.191) (see Table

 

 

 

 

 

 

 

 

5.2).
Table 5.2 Coefficientsa

Unstandardized Standardized

Coefficients Coefficients

Model 1 B Std. Error Beta 1 Sig.
(Constant) 6.629 1 .060 6.254 .000
Black“ * 1.872 .456 .068 4.103 .000
Latino" 3.218 .655 .078 4.9101 .000
Native American 1.751 2.130 .013 .822 .411
Asian" 3.752 .442 .143 8.486 .000
Mixed Race" 4.495 .749 .094 5.999 .000
Gender (0=Male) -.129 .232 -.009 -.556 .578
Academic Standing“ .868 .116 .117 7.497 .000
Hours Socializing" 1.024 .096 .169 10.683 .000
Neighborhood Demographics
(0=Segregate (1),, * 1.126 .291 .075 3.870 .000
”Emma” 133”?”th .404 .298 .027 1.356 .175
(O—Segregated)
JH/HS Demographics
(O=Segregated)"' .554 .273 .039 2.032 .042
Father’s Education -.004 .082 .000 -.051 .959
Mother’s Education“ .192 .085 .041 2.253 .024
Family Income“ -.147 .070 -.037 -2.093 .036
Current Hall Percent White without -.088 .020 -.073 _ 4. 4 62 .000
Van Hoosen"
Overall Racial Environment Scale
without Van Hoosen" .316 .030 .167 10.538 .000
1St Year Roommate: Blind/Same
Race" 1.423 .284 .101 5.015 .000
1St Year Roommate:
B1ind/DifferentRace** 3.482 .328 .221 10.619 .000
1St Year Roommate:
Request ed/Di fferent Race" 4.527 .530 .148 8.544 .000

 

 

 

 

 

 

 

a. Dependent Variable: interactional diversity scale

In addition to basic demographics, measures related directly to the residence hall,

such as Current Hall Percent White and the Overall Racial Environment scale, were also

203

 

 

 

included. The role of first-year roommate is also considered in this model. Dummy
variables were created to compare participants who did not request a first-year roommate
of their same racial background with those that did. For these variables, participants
residing in Van Hoosen Hall were included.
fia_c_g

The multivariate regression reveals that even when controlling for such factors as
gender, academic standing, and current residence hall, race continues to be a significant
predictor of interactional diversity. Given that race proved to be significant for each of
the individual measures within the Interactional Diversity scale, this is not surprising.
Compared with White students, Black, Latino, Asian, and Mixed Race students are
significantly more likely to engage in meaningfirl and ongoing activities across race lines.
Among these groups, the effects of race are greatest for Mixed Race and Asian students.
Native American responses were not significantly different from White students in this

model.
@212

During the bivariate analysis, gender was not a consistently significant predictor
for the Interactional Diversity scale. Once other variables are controlled for, gender is no
longer significant.
Academic Standing

Academic standing was found to be significant for all of the bivariate measures
and in the multivariate analysis it continues to be a significant positive predictor of
interactional diversity. Upperclassmen that live on campus are more likely than fi'eshmen

to engage in interactional diversity. In other words, as time spent on campus increases, so

204

does the likelihood that students will interact regularly, and in meaningful ways, with
students of a different racial background than themselves.
Pre- C allege Demographics

Regardless of race, gender, or academic standing, students raised in a racially
diverse neighborhood exhibit a higher rate of interactional diversity than those raised in
racially segregated areas. While the racial demographics of the elementary school is not a
significant predictor, the racial demographics of students’ junior high and high school are
significant. Students who describe their junior high and high school as racially mixed are
more likely to engage in interactional diversity than those who describe these schools as

segregated.
Parents ’ _Backgro and

In this multivariate analysis, Father’s Education is not a significant predictor of
students’ engagement in interactional diversity. However, Mother’s Education is
meaningful. The likelihood that students will participate in interactional diversity
increases as mother’s education increases. Relating to family income, this variable has a
significant negative effect on interactional diversity such that students from more affluent
homes are less likely to engage in interactional diversity, regardless of race, gender,

academic standing, or current residence hall.
Residence Hall

In this model, the residence hall demographics appear to be a significant predictor
of interactional diversity. This measure is negatively significant at less than .0005
suggesting that as the percentage of White residents in a residence hall increases—and

thus, the percentage of students of color decreases—the level of interactional diversity

205

decreases. This relationship proves to be significant even after controlling for the race,
gender, and academic standing of the participants.
Overall Racial Environment

The Overall Racial Environment within the residence hall is also a meaningful
predictor of student behavior. A significant and positive relationship was found between
these two variables after controlling for other factors, indicating that residents who
perceive less racial conflict to be occurring within their halls, and who feel emotionally
and physically safe in their halls, are more likely to engage in interactional diversity.

F irst- Year Roommate

First-year roommate continues to be a significant predictor of interactional
diversity in the multivariate analysis. Compared to all other students, those students who
requested their first-year roommate, and whose roommate is of their same racial
background, were significantly less likely to engage in interactional diversity. Students
whose first-year roommate is of a different racial background, regardless of whether they
had specifically requested that roommate, are more likely than students who share the
same racial background with their first-year roommate but did not request that roommate.
Students who requested their first-year roommate, and whose roommate is of a different
racial background, were the most likely to participate in interactional diversity even after

controlling for race, gender, academic standing, and current residence hall.
Interactions: Current Hall Percent White

Although a number of variables proved to be significantly related to interactional
diversity in the multivariate regression, it is important to test for the possibility of

interactions between them.

206

_R_q_c_e

The first interaction tested was between the race variable and Current Hall Percent
White to determine whether interactional diversity levels in more racially diverse halls
result from the fact that non-White students are more likely to engage in interactional
diversity. A series of regressions equations were run for each of the racial categories that
were significant in the original regression model to test the possibility of interaction
between the racial category and Current Hall Percent White. Results from these equations
proved that the interaction between the racial category and Current Hall were significant
and positive for Black, Asian, and Mixed Race students. This suggests that compared to
other racial groups, the racial composition of the residence hall affects the interactional
diversity of these groups differently. To clarify the extent of this relationship, I ran a
separate round of subgroup regressions based on race. For each regression model, I
included all of the same independent variables and removed the dummy variables for
race. I used the self-reported race variable as the selection variable and for each equation,
I chose one racial group to focus on. When the selected cases for the regression model
were those students who self-identified as White or Caucasian, we see that the Current
Hall Percent White is significant at less than .0005 and negative (see Table 5.3). This
indicates that for White students, as the percentage of White residents in their current hall

increases, the likelihood of interactional diversity decreases.

207

Table 5.3 Coefficients”

 

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta 1 Sig.
(Constant) 6.767 1 .224 5.529 .000
Gender (0=Male) -.279 .258 -.020 -1.080 .280
Academic Standing" .844 .130 .119 6.483 .000
Hours SocializingM .982 .106 .171 9.278 .000
Neighborhood Demographics
(0=Segregated)" 1.117 .325 .077 3.434 .001
Elementary Demographics
(0=Segregated)* .667 .336 .047 1.983 .047
JH/HS Demographics
(0=Segregate (1),, .827 .307 .062 2.699 .007
Father’s Education -.058 .094 -.013 -.620 .536
Mother’s Education“ .197 .096 .042 2.048 .041
Family Income -.093 .079 -.024 -1.181 .238
Current Hall Percent White
without Van Hoosen" -.134 .022 -.115 -6.157 .000
Overall Racial Environment
Scale without Van Hoosen" .333 .035 .176 9.572 .000
1St Year Roommate:
Blind/Same Race“ 1.277 .299 .096 4.272 .000
1St Year Roommate:
Blind/Different Race" 3.039 .376 .181 8.085 .000
1"“t Year Roommate:
Requested/Different Race" 4.560l .714 .121 6.387 .000

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = White/Caucasian

When the selected cases for the regression model were those students who self-

identified as Black, we see that the Current Hall Percent White is greater than .05 (see

Table 5.4). This indicates that for Black students, the percentage of White residents in

their current hall does not significantly affect their levels of interactional diversity.

208

Table 5.4 Coefficients”

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B I Std. Error Beta t Sig.
(Constant) 9.356 3.574 2.618 .009
Gender (0=Male) -1.660 1.028 -.105 -1.615 .108
Academic Standing .356 .441 .051 .806 .421
Hours Socializing“ 1.059 .450L .150- 2.355 .019
Neighborhood Demographics
(0=Segregated) .422 1.101 .028 .383 .702
Elementary Demographics
(0=Segregat ed) .599 1.064 .040 .562 .574
JH/HS Demographics
(0=Segregated) -.923 1.020 -.065 -.905 .367
Father’s Education .280' .284 .075 .987 .325
Mother’s Education .004 .327 .001 .013 .9901
Family Income .005 .297 .001 .017 .987
Current Hall Percent Wlnte Without .112 .084 .086 1.334 .183

an Hoosen
Overall Racial Environment Scale
without Van Hoosen" .340 .103 .209 3.297 .001

st , -
1 Year R°°mmate' Bum/same 1.274 1.386 .067 .919- .359
Race

st . . .
1 Year Roommate. Blind/Different 2.280 1.100 .163 2.072 .039
1” Year Roommate:
Requested/Different Race 2.963 . 1.604 .131 1.848 .066

 

 

 

 

 

 

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Black

When the selected cases for the regression model were those students who self-

identified as Latino, we see that the Current Hall Percent White is greater than .05 (see

Table 5.5). This indicates that for Latino students, the percentage of White residents in

their current hall does not significantly affect their levels of interactional diversity.

Similar results were found for those students who self-identified as Asian (see Table 5.6)

and those who identified as Mixed Race (see Table 5.7).

209

Table 5.5 Coefficients”

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Std. .
Model B Error Beta t 818-
(Constant) 19.354 5.956 3.249 .002
Gender (0=Male) .205 1.437 .014 .142 .887
Academic Standing -.598 .801 -.080I -.747 .457
Hours Socializing .479 .642 .075 .746 .457
Neighborhood Demographics r
(0=Segregated) -.362 1.759 -.025 -.206 .837
Elementary Demographics _ _ _
(0=Segregated) 3.719 1.998 .262 1.861 .066
JH/HS Demographics (0=Segregated) 2.289 1.841 .159 1.244 .217
Father’s Education .202 .500I .056 .404 .687
Mother’s Education .481 .556 .118 .865 .389I
Family Income -.802 .415 -.219I -1.934 .056
Current Hall Percent White without Van 403 .155 _.132 _1 .3 12 .193
Hoosen
Overall Racial Envrronment Scale .114 .179 .065 .636 .526
Without Van Hoosen
1st Year Roommate: Blind/Same Race 3.784 2.644 .218 1.431 .156
1” Year Roommate: Blind/Different
R H 8.312 2.419 .584 3.437 .001
1" Year Roommate:
Reque 1 'Di fi‘erent Race" 8.349 2.947 .455 2.833 .006

 

 

 

 

 

 

 

a. Dependent Variable: interactional diversity scale

b. Selecting only cases for which race final self-reported = Latino

210

Table 5.6 Coefficients“

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Std. .
Model B Error Beta 1 318-
(Constant) -.154 3.715 -.042 .967
Gender (0=Male) .958 .818 .065 1.171 .243
Academic Standing” 1.002 .427 .134 2.346 .020
Hours Socializing” 1.568 .385 .234 4.078 .000
N‘Eghmho‘“ Demg‘aph’cs 1.307 1.153 .089 1.134 .258
(0—Segregated)
Elementary Demographics
(0=Segregated) -.749 1.152 -.050 -.6501 .516
JH/I-IS Demographics (0=Segregated) -.785 1.070 -.053 -.734 .464
Father’s Education .130 .322 .032 .404 .687
Mother’s Education .331 .320 .079 1.035 .302
Family Income -.163 .246 -.041 -.661 .5091
Current Hall Percent Whrte wrthout Van .082 .081 .0 60 1.012 .312
Hoosen
Overall Racial Environment Scale
without Van Hoosen“ .484 .110 .251 4.393 .000
1St Year Roommate: Blind/Same Race" 6.184 2.020 .229 3.061 .002
18’ Year Roommate: Blind/Different
R H 5.812 1.508 .380 3.853 .000
1”t Year Roommate:
Requested/Different Race" 6.674 1.716 .358 3.889 .000

 

 

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Asian

211

 

Table 5.7 Coefficients“

 

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta 1 Sig.
(Constant) 12.524 7.793 1.607 .1 13
Gender (0=Male) 2.666 1.848 .159 1.443 .154
Academic Standing 1.944 .993 .225 1.958 .055
Hours Socializing .713 .641 .127 1.114 .2701
Neighborhood Demographics
(0=Segregated) 3.109 2.314 .210 1.344 .184
Elementary Demographics
(0=Segregate d) .092 2. 108 .006 .044 .965
”1st Demg’aphms -2.799 1.895 -.192 -1477 .145
(O—Segregated)
Father’s Education -.463 .527 -.115 -.879 .383
Mother’s Education 1.049 .648 .218 1.618 .111
Family Income -.901 .556 -.234 g -1.620 .1101
Cl’m’m Ha" Percent WM" .082 .150 .064 .546 .587
Without Van Hoosen
Overall Racial Environment 1
Scale without Van Hoosen .210 .196 .120 1.072 .288
st , -
1 Year R°°mmate° Bhnd/SW’ -1.248 2.877 -.O76 -.434 .666
Race
1St Year Roommate:
Blind/Different Race 3.160 2.549 .216 1.239 .220
1St Year Roommate:
Requested/Different Race“ 8.027 3.908 .275 2.054 .044

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Mixed

Race

Taken together, these results suggest that the racial composition of the residence

halls does not significantly affect any of the non-White students. Therefore, the

significance of the variable Current Hall Percent White results primarily from its strong

effect on White students.

212

Academic Standing

The next interaction tested for is between academic standing and current
residence hall. The percentage of upperclassmen ranges substantially between residence
halls. To ensure that the levels of interactional diversity within residence halls is not a
function of the academic standing of its residence, a regression model was run to test this
interaction. The interaction variable for academic standing and current hall was not
significant; therefore, we can assume that the interactional diversity within the residence
halls is not the result of the academic standing of its residents.

Hours Socializing

The possibility of interaction between hours socializing and current residence hall
was also considered. It is possible that the residents in some halls spend more time
socializing than in other halls and that this is contributing to higher rates of interactional
diversity in particular halls. This variable is significant at less than .005 indicating that
the relationship between the racial demographics in the residence hall and its interactional
diversity vary by the hours its residents spend socializing. Individual subgroup
regressions were run to determine how the number of hours students spend socializing
influences their interactional diversity. When the cases included in this analysis were for
those students who reported socializing for 0 hours of socializing per week, the racial
composition of the residence hall was not significant. The relationship between 1 to 5
hours of socializing per week and Current Hall was also not significant. Hours of
socializing is significantly and negatively related to Current Hall Percent White for those
students who report socializing 6 to 10 hours per week. This suggests that for students

who socialize 6 to 10 hours per week, the likelihood of their engagement in interactional

213

diversity decreases as the percentage of White residents living in their current hall
increases. Similar results were found for those students who spend 11 to 15 hours per
week socializing. When the cases included for analysis were students who report
socializing 16 to 20 hours per week, the racial composition of the residence hall no longer
significantly affects interactional diversity. For students who describe spending more
than 20 hours per week socializing, Current Hall is found to again be significantly and
negatively related to interactional diversity.

Pre- College Demographics

The possibility of interaction between neighborhood demographics and current
residence hall was also considered. This variable is not significant, indicating that the
relationship between current residence hall and interactional diversity is not related to the
neighborhood demographics of the residents within that hall. The demographics of the
elementary school were not found to be significant in the original model; therefore, it is
not necessary to conduct an interaction test with this variable. An interaction between the
racial demographics of the junior high and high school and current residence hall was
tested. This variable is greater than .05 indicating there is no significant interaction
between the racial demographics of the respondents’ junior high and high school and the

racial composition of the residence hall related to interactional diversity.
Parents ’fiBaLkgLound

Father’s education was not found to be significant in the original regression
variable; therefore, this variable was not considered in the tests for interactions. Because
Mother’s education was significant in the original model, I tested for the possibility of

interaction between mother’s education and the racial composition of the residence hall.

214

This variable was not found to be significant indicating that the effect of racial
composition in the residence halls on interactional diversity does not vary for students
based on their mothers’ education level. The possibility of interaction between Current
Hall and Family Income was also tested. This variable was significant at less than .05 and
negative suggesting that the effect of Current Hall on interactional diversity varies across
income brackets. Subgroup regressions were run to determine the relationship between
family income and Current Hall. The relationship between family income and Current
Hall is not significant for students who describe their family income as less than $60,000.
When the regression model is run to include only those cases where students report their
family income between $60,000 and $79,999 the relationship between these variables is
significant at less than .05 and negative. This suggests that for students in this income
bracket, the likelihood that they will engage in interactional diversity decreases as the
percentage of White residents in their current hall increases. When we include only those
students who report their family income as $80,000 to $99,999 the relationship between
family income and the racial composition of the residence hall is no longer significant.
For students who describe their family income as greater than $100,000 the relationship

again becomes significant at less than .0005 and negative.

Other Interactions: Race

While testing for the possibility of interaction between race and the racial

composition of the residence halls, a number of other meaningful findings were

discovered that should be addressed.

215

Academic Standing

Although academic standing is found to be a significant predictor of interactional
diversity in the original regression model, the effects of academic standing are not
consistent for all racial groups. Subgroup regressions based on race reveal that academic
standing is only significantly related to interactional diversity for White (p<.0005) and

Asian (p<.05) students. It is not a significantly related to any of the other racial groups.

Hours Socializing

As mentioned, the number of hours that students spend socializing per week was
found to be a positively significant predictor of interactional diversity in the original
regression model. Furthermore, this variable was found to interact with the racial
composition of the residence hall. Subgroup regressions based on race suggest that the
effect of hours spent socializing also varies across racial groups. These equations reveal
that hours spent socializing is only significantly related to interactional diversity among

White (p<.0005), Black (p<.05), and Asian (p<.0005) students.

Pre- College Demographics

Neighborhood demographics were found to be a significant predictor of
interactional diversity in the original model, suggesting that students who report growing
up in integrated neighborhoods were more likely to engage in interactional diversity.
Upon closer examination, however, it is only White students who are significantly
affected by neighborhood demographics. When we include only those students who self-
identify as White in the equation, neighborhood demographics are significant at less than
.005. The effects of neighborhood demographics on interactional diversity are not

significant for any other racial groups. Elementary school demographics and junior

216

high/high school demographics were also found to be significant at less than .05 for
White students but were not significantly related to any other racial groups. These
findings suggest that students of color are equally likely to engage in interactional
diversity regardless of whether they were raised in a segregated community and attended
segregated schools. For White students however, their pre-college environment is a
meaningful determinant of interactional diversity in college. The implications of these
findings will be discussed further in the next chapter.

Parents ’ Background

Mother’s education was found to be significantly and positively related to
interactional diversity in the original regression model. Subgroup regressions based on
race indicate that this variable is only significantly related to interactional diversity
among White students (p<.05). This suggests that for White students, the likelihood that
they will engage in interactional diversity increases as the educational level of their
mother increases. The relationship between these variables was not found among other
racial groups. Although family income was found to be significantly and negatively
related to interactional diversity in the original model and interaction tests found that the
effects of family income vary across residence halls, there does not appear to be a
relationship between family income and race. This indicates that levels of interactional
diversity found within a particular racial group do not differ by family income. African
American students, for example, are no more or less likely to engage in interactional

diversity if they are from a poor or affluent family.

217

Overall Racial Environment

In the original regression model, the Overall Racial Environment scale was found
to be a significant positive predictor of interactional diversity. Subgroup regressions
based on race reveal that the effects of the racial environment within the residence halls
differ across racial groups. The effects of the Overall Racial Environment on interactional
diversity were significant at less than .005 for White, Black, and Asian students. For
these students, the likelihood that they would engage in interactional diversity increased
with a more positive racial environment in their current hall. Levels of interactional
diversity were not significantly affected by this variable for Latino and Mixed Race

students.

F irst- Year Roommate

When compared with students who requested their first-year roommate, and
whose roommate is of their same racial background, the effects of first-year roommate on
interactional diversity were found to be significant. Subgroup regressions based on race
reveal that these effects vary across racial groups. White students were significantly more
likely to engage in interactional diversity if they did not request their first-year roommate
and share that roommate’s racial background (p<.0005). The effects of first-year
roommate were only significant for Black students who were randomly assigned a first-
year roommate of a different racial background (p<.05). These students were more likely
than any other Black students to engage in interactional diversity. Levels of interactional
diversity among Latino students were significantly higher if their first-year roommate is
of a different racial background, whether they had requested that roommate or been

randomly assigned. Among Latino students who were randomly assigned a first-year

218

roommate from a different racial background, the effects on interactional diversity were
significant at less than .005. Among Latino students who requested their first-year
roommate, and whose roommate is of a different racial background, the effects were
significant at less than .05. The effects of first-year roommate on interactional diversity
for Asian students were similar to that of White students. Compared with students who
requested their first-year roommate, and who share the same racial background as that
roommate, all other Asian students were significantly more likely to engage in
interactional diversity. For those students who were randomly assigned a roommate of
their same racial background, the effect on interactional diversity was significant at less
than .005. For those students whose first-year roommate is of a different racial
background, whether they had requested that roommate or been randomly assigned, the
effect on interactional diversity was significant at less than .0005. Finally, among Mixed
Race students, only those students who requested their first-year roommate, and whose
roommate is of a different racial background, were significantly more likely to engage in

interactional diversity (p<.05).
Regrgssion: Classroom Experience

In the next section, we will consider the relationship between interactional
diversity and experiences within the classroom (r’=.322) (see Table 5.8). In addition to
the basic demographics, the Classroom Experience scale was included in this equation.
As a reminder, the Classroom Experience scale considers the extent to which students
interact with peers across race lines in the classroom, students’ values and beliefs are
challenged within the classroom, and students’ understanding of others is broadened in

the classroom. In this equation, we find that when we control for experiences within the

219

classroom, Latino, Asian, and Mixed Race students are significantly different from White
students in their levels of interactional diversity (p<.0005). Men are significantly less
likely than women to engage in interactional diversity once we consider the Classroom
Experience scale (p<.0005). Academic standing appears to be significantly and positively
related to interactional diversity in this model (p<.0005). Here the significance of
neighborhood demographics disappears but the racial composition of junior high and
high school remains a significant and positive predictor of interactional diversity
(p<.005). Mother’s education continues to be a significantly and positively related to
interactional diversity (p<.005) while family income continues to significantly and
negatively related to this variable (p<.05). Finally, the Classroom Experience scale is

significantly and positively related to interactional diversity (p<.0005).

220

Table 5.8 Coefficients”

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model 2 B Std. Error Beta 1 Sig.
(Constant) 3.687 .635 5.803 .000
Black .651 .402 .024 1 .620 .105
Latino" 2.419 .586 .059 4.129 .000
Native American 3.057 1.926 .022 1.587 .1 13
Asian" 2.9401 .380 .1 12 7.740 .000
Mixed Race" 3.899 .676 .081 5.768 .000
hGender (0=Male)* * -.786 .208 -.053 -3 .781 .000
Academic Standing" * .573 .104 .078 5.530 .000
Hours Socializing“ * 1 .002 .086 .166 1 1 .671 .000
Nei borhood Demo a bios
(0=Sleigregated) gr 1’ .475 .262 .032 1.813 .070
Elfmenm’y Demographics .255 .268 .017 .9501 .342
(O—Segregated)
JH/HS Demo a bios
(0=Segregatecgl;*g .792 .245 .056 3.231 .001
Father’s Education -.008 .074 -.002 -.105 .917
Mother’s Education* * .245 .077 .053 3.194 .001
Family Income“ -.154 .063 -.039 -2.431 .015
[Classroom Experience Scale“ 1.231 .038 .474 32.591 .000J

 

 

 

a. Dependent Variable: interactional diversity scale

Interactions: Classroom Experience Scale

Although the classroom experience proved to be significantly related to
interactional diversity in the multivariate regression, it is important to test for the
possibility of interactions between the classroom experience and other variables as we did

in the previous model.
Race

A regression equation was run to test for the possibility of interaction between,

race and the Classroom Experience scale. When all of the dummy variables based on race

221

were included in this model, only the interaction between Mixed Race students and
Classroom Experience was significant at less than .05. This suggests that the effects of
what happens in the classroom on interactional diversity are different for Mixed Race
students than all other racial groups. When the regression equation was run to test the
interaction between race dummy variables and Classroom Experience individually, so
that only one interaction was tested in each equation, the interaction between Classroom
Experience and race was significantly and negatively related for Latino students (p<.05)
and significantly and positively related for Mixed Race students (p<.05) compared to all
other racial groups. These findings suggest that the Classroom Experience effects
interactional diversity differently across racial groups. In order to gain a clearer
understanding of how these variables are related, subgroup regressions based on race
were analyzed.

When the regression equation is run to include only those students who self-
identify as White or Caucasian, the effects of the Classroom Experience on interactional
diversity is significant at less than .0005 and positive (see Table 5.9). This indicates that
as the level of interaction and engagement that White students experience in the
classroom across race lines increases, their likelihood of engaging in interactional
diversity outside of the classroom increases significantly.

When the regression equation is run to include only those students who self-
identify as Black, the relationship between the Classroom Experience scale and
interactional diversity is significant at less than .0005 and positive (see Table 5.10).
Similar to White students then, the experiences that Black students have within the

classroom are predictive of their interactional diversity outside the classroom.

222

 

Table 5.9 Coefficients“

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta 1 Sig.
(Constant) 4.580 .718 6.375 .000
Gender (0=Male)" -1.179 .233 -.084 -5.065 .000
Academic Standing“ .527 .116 .075 4.522 .000
Hours Socializing“ .919 .095 .161 9.670I .000
Nei hborhood Demo a hics
(0=Segregated) gr 1’ .196 .295 .014 .664 .507
Elfmema’y Demc’graphics .332 .303 .023 1.093 .274
(O-Segregated)
JI-I/HS Demo a hics
(0=Segregategx 1.031 .276 .077 3.730- .000
Father’s Education -.050 .085 -.011 -.585 .559
Mother’s Education” .242 .087 .052 2.788 .005
Family Income“ * -. 193 .072 -.049 -2.695 .007
Classroom Experience Scale" 1.227 .043 .473 28.546 .000

 

 

 

 

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = White/Caucasian

223

 

 

Table 5.10 Coefficients“

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B lStd. Error Beta t Sig-
(Constant) 2.179 2.563 .850 .396
Gender (0=Male) -1.310 .877 -.083 -1.494 .137
Academic Standing .438 .382 .063 1.146 .253
Hours Socializing" 1.240 .392 .174 3.166 .002
giggggggpemograpmcs 1.218 .958 .081 1.271 .205
g:’§:;fgt25m°gmphlcs .921 .929 .063 .991 .323
“LE/HS Demographics -.431 .887 -.030 -.486 .627
(O—Segregated)
Father’s Education .237 .248 .063 .958 .339
Mother’s Education .195 .288 .047 .677 .499
Family Income .058 .258 .014 .226 .822
Classroom Experience Scale" 1.320] .146 .504 9.044 .000

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Black

Subgroup regressions for Latino students reveal a significant and positive

relationship with the Classroom Experience as well (p<.005) (see Table 5.11). Similar

results were also found for Asian (p<.0005) and Mixed Race (p<0005) students (see

Tables 5.12 and 5.13). As was the case for the subgroup regressions related to Current

Hall, the degree of multicollinearity was too high to test for interactions among Native

American students.

224

 

Table 5.11 Coefficients“

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta 1 Sig.
(Constant) 10.263 5.125 2.002 .048
Gender (0=Male) 1.616 1.446 .1 12 1.118 .267
Academic Standing .309 .747 .041 .414 .680I
Hours Socializing .456 .631 .071 .723 .472
gigggggfemomh’“ .020 1.671 .001 .012 .991
$§§§§$§5m0gmpmcs -2.186 1.937 -.154 -1.128 .262
JH/HS Demo a hics
(0=Segregatei; 1’ 2.483 1.800- .173 1.380- .171
Father’s Education -.287 .493 -.080 -.582 .562
Mother’s Education .886 .536 .220 1.652 .102
Family Income -.343 .419 -.094 -.819 .415
Classroom Experience Scale" .922 .284 .337 3.250 .002

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Latino

225

Table 5.12 Coefficients”

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients
Model B | Std. Error Beta t Sig.
(Constant) 4.0201 2.248 1.788 .075
Gender (0=Male) .752 .750i .051 1.003 .317
Academic Standing .546 .388 .073 1.405 .161
gours SocializingM 1.723 .348 .258 4.957 .000
ei borhood Demo a hics

(ggggegmd) gr 1’ .377 1.044 .026 .361 .719
Elemen Demo a hics
(0=Segrfgued) g p -.117 1.041 -.008 -.113 .910
JH/HS Demo hics
(0=Segregategap -.033 .985 -.002 -.033 .974
Father’s Education .092 .293 .023 .315 .753
Mother’s Education .018 .295 .004 .062 .951
Family Income .197 .223 .050 .884 .378
[Classroom Experience Scale" 1.237 .132 .489 9.356 .000]

 

 

 

 

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Asian

Table 5.13 Coefficients”

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized
Coefficients Coefficients

Model B Std. Error Beta t Sig.
(Constant) 1.845 5.096 .362 .718
Gender (0=Male) 1.064 1.502 .064 .708 .481
Academic Standing 1.044 .840 .121 1.243 .218
Hours Socializing" .861 .543 .153 1.585 .118
Niighbmhmd DemOg’aphics 3.604 1.845 .243 1.954 .055
(O—Segregated)
g:g’:;’:g£5m°graph‘cs -.828 1.780- -.057 -.465 .644
JHIHS Demographics -1.262 1.657 -.086 -.762 .449
(O—Segregated)
Father’s Education .329 .421 .082 .783 .437
Mother’s Education .456 .544 .095 .837 .406
Family Income -.740 .472 -.192 -1.567 .122
Classroom Experience Scale” 1.561 .279 .554 5.588 .000]

 

a. Dependent Variable: interactional diversity scale
b. Selecting only cases for which race final self-reported = Mixed Race

Taken together, these findings suggest that the effects of Classroom Experience

on interactional diversity are significant and positive for every racial group. However, the

degree of influence that these experiences have on students varies across race.

Academic Standing

The possibility of interaction between academic standing and the Classroom

Experience scale was considered to ensure that this relationship was not influencing the

significant findings. The interaction variable was not significantly related to interactional

diversity.

Hours Socializing

The possibility of interaction between the number of hours that students describe
socializing per week and the Classroom Experience was explored. It is possible that
students who spend more time socializing are more outgoing in the classroom setting and
may benefit more from the interactional opportunities provided for them in the classroom
setting. The interaction variable was not significantly related to interactional diversity.

Pre-College Demographics

In the original regression model, neighborhood demographics and the
demographics of students’ elementary school were not significantly related to
interactional diversity. As a result, the possibility of interaction between these variables
and the Classroom Experience will not be explored. The demographics of students’ junior
high and high schools were significantly related to interactional diversity in the original
model. The possibility of interaction between this variable and the Classroom Experience

was considered; however, this relationship was not significant.

Parentflackgo und

Although father’s education was not significantly related to interactional diversity
in the original regression model, mother’s education was found to be significantly and
positively related to interactional diversity. When we include the interaction between
Mother’s education and Classroom Experience in the regression equation this variable is
not significantly related to interactional diversity. The interaction between family income
and experiences within the classroom was also explored. This variable was not

significantly related to interactional diversity.

228

Other Interactions: Race

While testing for the interactions between Classroom Experience and race, a number of

other variables appear to be interacting with race in this model that should be discussed.
Gender

Although gender is significantly and negatively related to interactional diversity
in the original regression model, subgroup regressions reveal that gender is only a
significant predictor among White students. When we control for experiences within the
classroom, White women are significantly more likely to engage in interactional diversity
than White men (p<.0005). The relationship between gender and race is not significant
for any other racial groups.
Academic Standing

Academic standing is a significant predictor of interactional diversity in the
original regression model; however, similar to the case of students’ gender, academic
standing is only significant for White students (p<.0005). Subgroup regressions based on
race indicate that academic standing does not significantly influence interactional

diversity levels for students of control, when we control for Classroom Experience.

Hours Socializing

 

In the original regression model, Hours Socializing was significantly related to
interactional diversity. Subgroups regressions based on race reveal that the number of
hours students spend socializing is significant among White (p<.0005), Black (p<.005),
and Asian (p<.0005) students. This indicates that for these students, the likelihood of

engaging in interactional diversity increases with the number of hours that they spend

229

socializing, when we control for Classroom Experience. This variable was not

significantly related to interactional diversity for Latino and Mixed Race students.
Pre-College Environment

In the regression model, junior high and high school demographics were
significantly and positively related to the Interactional Diversity when controlling for the
Classroom Experience. Sub-group regressions reveal the racial composition of students’
junior high and high schools were only significantly related to Interactional Diversity

among White students (p<.0005).

Parents ’ Rae/ground

 

Mother’s education was significantly and positively related to interactional
diversity in the original regression model; however, subgroup regressions suggest that
this variable is only meaningful for White students. When we run the regression to
include only those students who self-identify as White or Caucasian, the relationship
between Mother’s education and interactional diversity is significant at less than .005 and
positive. This relationship is not significant among students of color. Similarly, although
Family Income was a meaningful variable in the original regression, subgroup
regressions reveal that it is only a significant predictor for White students. Among White
students, the likelihood of interactional diversity decreases as family income increases
(p<.005).

These findings suggest that most of the significant variables found in the original
regression model are based on the strength of the effects on White students. Other than

the Classroom Experience scale, none of the other variables are significant predictors

230

across all racial groups. This is a meaningful finding, however, because it suggests that

what happens inside the classroom effects the interactional diversity levels of all students.

m

Multiple regression models suggest that levels of interactional diversity are
significantly related to a ntunber of independent variables including race, academic
standing, the environment in which students were raised prior to attending college,
parental background, the racial composition and dynamics within the residence halls, and
experiences within the classroom. Upon closer inspection, however, it became evident
that interactional diversity rates are affected differently across racial groups. The primary
finding from subgroup regressions was that the level of interactional diversity among
White students is especially susceptible to external variables.

When we consider the relationship between interactional diversity and campus
housing, every independent variable that was significant in the original regression
model—except for family income—was also significant in the subgroup regression
model was for White students. In particular, the racial composition of the current
residence hall was only significantly related to White students’ levels of interactional
diversity. For these students, as the percentage of White residents living in their current
hall increased, the likelihood of interactional diversity decreased significantly. This
relationship was not found among any other racial group.

Analysis of the relationship between interactional diversity and the classroom
experience revealed similar results. While the interactional diversity rates of White
students were significantly related to a number of independent variables, this was not the

case for students of color. However, all students, regardless of race, were significantly

231

affected by their experiences within the classroom. When students’ ideas and values are
challenged within the classroom, and when they are given the opportunity to engage with
peers of a different racial background in the classroom, their levels of interactional
diversity outside of the classroom increases as well. Although cross-sectional data cannot
determine if there is a causal relationship between these two variables, we can be sure
that they are related to one another.

In the following chapter, we will consider the implications of these findings and
how they relate to the existing literature. We will also review the limitations of this study

and consider the future directions that interactional diversity research should pursue.

232

CHAPTER 6: CONCLUSION

Discussion

The first portion of this survey asked students a series of questions related to their
experiences on and perceptions of the campus at Michigan State University. Less than
half of the respondents perceive racial tension on campus. The racial/ethnic background
of students was significantly related to this measure. African American students were the
most likely to perceive this tension on campus while White students were the least likely.
Men were significantly less likely to perceive racial tension. Academic standing was also
significantly related with upperclassmen significantly more likely than freshmen to
perceive racial tension. Current hall was a significant predictor of perceived racial tension
on campus. Residents living in halls with the largest percentage of students of color as
well as Black residents—Rather, Hubbard, and Butterfield—were the most likely to
perceive racial tension. In contrast, residents living in the halls with less racial
diversity—were significantly less likely to perceive racial tension on campus.

More than half of the respondents described the campus racial climate as good.
Significant differences were found between racial/ethnic groups. Native American
students, followed by White students, were the most likely to describe the racial climate
favorably. Black students were the most likely to describe the racial climate negatively.
Men were significantly more likely to view the campus racial climate favorably.
Academic standing was also significantly related to perceived campus climate. Freshmen
were significantly more likely to view the racial climate favorably. Current residence hall

was a significant predictor of perceived campus climate. Students in the most racially

233

diverse halls are more likely to describe the current climate as poor or fair when
compared with students from other, less diverse halls.

The majority of participants describe broadening the diversity (race, sexual
orientation, background, religious, culture, etc.) of their immediate circle of fi'iends since
attending Michigan State University. Native American students, followed by Asian
students, were the most likely to strongly agree with this statement while Black and
White students were the most likely to disagree or strongly disagree with this statement.
Gender was not significantly related to this variable. Academic standing was significantly
and positively related to this variable. Freshmen were significantly less likely to agree
with this statement than upperclassmen. Seniors were the most likely to strongly agree.
Current hall was significantly related to broadening the diversity of friends. Residents of
Williams Hall—which has the highest percentage of upperclassmen on campus—and
Holmes Hall—with the largest living-learning program—were the most likely to agree
with this statement.

Most respondents report that they have become more aware of their personal
prejudices since attending Michigan State University. Black and Native American
students were more likely than other racial/ethnic groups to agree or strongly agree with
this statement. White and Asian students were the most likely to disagree or strongly
disagree with this statement. Mixed Race students were the most likely to strongly
disagree and the least likely to strongly agree with this statement. Women were
significantly more likely to agree with this statement than men. Freshmen were
significantly less likely to agree with this statement than upperclassmen. Seniors were the

most likely to strongly agree. Current hall was significantly related to this variable;

234

however, a clear pattern did not emerge. In general, those residence halls with the
greatest diversity, and the greatest percentage of Black residents, tend to show higher
rates of agreement than those halls with fewer students of color.

The primary dependent variable in this dissertation is the Interactional Diversity
scale. Out of the ten items in this scale, eight were adapted from Preparing Students for a
Diverse Democracy Project (Saenz, Ngai, and Hurtado 2007). These items include the
fi'equency with which respondents dine, have meaningful discussions outside of class
about racial/ethnic issues, share personal feelings and problems, study or prepare for
class, socialize or party, have intellectual discussions outside of class, and attend events
with students of a different racial background; as well as the frequency with which they
attend events sponsored by racial/ethnic groups different from their own background. The
two additional items that were added to this scale by the researcher were the frequency
with which respondents visit the rooms of residents of a different racial background and
visit in their own rooms with students of a different racial background.

Item analysis of the Interactional Diversity scale revealed that students vary
significantly in their likelihood to participate in these activities. For each item on the
scale, there were significant racial/ethnic differences in the responses. For example,
nearly two-thirds of White students and at least half of Native American and Black
students describe never or only occasionally dining or sharing meals with peers of a
different racial background. In contrast, more than two-thirds of Asian students, and more
than half of Latino and Mixed Race students report engaging in this activity ofien or very

often.

235

The most dramatic racial/ethnic differences in the item analysis were found on the
items related to visiting in the dorm rooms. More than two-thirds of White students and
more than 60 percent of Black students describe never or only occasionally visiting the
rooms of residents of a different racial background. Mixed Race students were
significantly more likely to engage in this activity than any other racial/ethnic group with
40 percent reporting that they engage in this activity very often, followed by Latino,
Native American, and Asian students. Among students who visit with students of a
different racial background in their own rooms, more than 70 percent of White students
and nearly 60 percent of Black students describe never or only occasionally engaging in
this activity. In contrast, 40 percent of Mixed Race students report engaging in this
activity very often.

Students rarely engage in meaningful discussions about racial/ethnic issues with
peers of a different racial background outside of class: more than 75 percent of
respondents describe never or only occasionally engaging in this activity. White students
were the least likely and Native American students were the most likely to participate in
these discussions.

Nearly 70 percent of students report never or only occasionally sharing their
personal feelings and problems with students of a different racial background. White and
Black students were the least likely to engage in this activity while Native American
students, followed by Mixed Race students, were the most likely to do so.

Substantial racial differences were found among respondents who study or
prepare for class with students of a different racial background. Three-quarters of White

students and more than two-thirds of Black students report never or only occasionally

236

engaging in this activity. All other students were most likely to describe studying with
peers of a different racial background often.

Significant racial/ethnic differences were found among students who socialize
and/or party with peers of different racial backgrounds, but the pattern is somewhat
different than for other items on the scale. Students were more likely to report engaging
in this activity than any other item on the scale. Black students were the least likely to
describe socializing or partying across race lines, followed by White students. Latino and
Mixed Race students were the most likely to do so.

The majority of students report never or occasionally having intellectual
discussions outside of class with peers of a different racial background. Two-thirds of
White students and more than 60 percent of Black students describe never or only
occasionally participating in these discussions. Mixed Race students were the most likely
to do so with more than a third reporting that they have these discussions very often.

One-third of White students and less than half of Black students report attending
events on or around campus with peers of a different racial background often or very
often. In contrast, more than two-thirds of Native American and Mixed Race students
report engaging in this activity.

Students rarely attend events on campus sponsored by racial/ethnic groups
different from their own background. Of all the items on the scale, students were the least
likely to engage in this activity. White students were significantly less likely to attend
these events than other racial groups: more than 45 percent report never attending these
events. Among students of color, Native Americans were the most likely to report never

attending events sponsored by other racial groups; however, they were also the group

237

most likely to report attending these events often. Mixed Race and Asian students were
the most likely to report attending these events very often.

Gender was not significantly related to the items within the Interactional Diversity
scale on a consistent basis. Men were significantly more likely than women to dine or
share meals with students of a different racial background, to visit in the rooms of
students of a different racial background, to visit with students of a different racial
background in their own room, socializing or partying with students of a different racial
background, having intellectual discussions outside of class with students of a different
racial background. Gender was not significantly related to having meaningful
discussions about racial/ethnic issues outside of class, sharing personal feelings and
problems, studying or preparing for class, and attending events with students of a
different racial background. Gender was also not a significant predictor of students who
attend events sponsored by racial/ethnic groups different from their own background.

Academic standing was also not a consistently significant predictor of items
within the Interactional Diversity scale. Academic standing was significantly and
positively related to dining and sharing meals with students of a different racial
background, visiting with students of a different racial background in their own rooms,
having meaningful discussions about racial/ethnic issues outside of class, studying or
preparing for class, socializing and partying, having intellectual discussions outside of
class, and attending events with students of a different racial background. Academic
standing was also significantly related to attending events sponsored by different
racial/ethnic groups. In general, students were more likely to participate in these activities

as their academic standing increased. Academic standing was not significantly related to

238

students who visit in the rooms of residents of a different racial background and sharing
personal feelings and problems with students of a different racial background.

The current residence hall played a significant role for each of the items within
the Interactional Diversity scale and a few basic trends emerged. First, students living in
Holmes Hall were consistently more likely to participate in each activity than most other
residents. Holmes Hall is unique in that it houses the largest living-learning program on
the campus of Michigan State University, the Lyman-Briggs Program: 86 percent of its
residents are in the College of Natural Science enrolled in programs such as physics,
chemistry, and biology. The most likely explanation for this finding is that students who
participate in these living-learning programs spend a great deal of time together, both
formally and informally, when compared to other students on campus. There are two
additional explanations. First, students in the Lyman-Briggs program are among the
smartest in the university. Many are enrolled in the Honors College and a large
percentage of the students go on to earn professional and graduate degrees. According to
the contact hypothesis, equal status between groups is a key component of intergroup
friendships and reducing prejudice (Allport 1954; Chavous 2005; Pettigrew 1998).
Another possible explanation is that the majority of non-White students that participate in
the Lyman-Briggs program are of Asian descent. It may be that White students feel more
comfortable interacting with these students than with other non-White groups.

The second housing trend that emerged was that residents in halls with greater
racial diversity were typically more likely to engage in interactional diversity than
residents of halls with less racial diversity. For example, students living in Rather Hall—

the most racially diverse hall on campus—were more likely to participate in the activity

239

than residents of other halls. Residents of Rather and Holmes Halls were more likely than
most other students to participate in each of the items within the Interactional Diversity
scale except dining and sharing meals, having meaningful discussions about racial/ethnic
issues outside of class, and sharing personal feelings and problems with students of a
different racial background. The students who were least likely to participate in these
activities varied between halls but, in general, these residents lived in the least racially
diverse halls.

Similarly, a strong relationship was found between the current residence hall and
the likelihood that students would visit with residents of a different racial background in
the each others’ rooms. For both items, students living in Rather Hall and Holmes Hall
were significantly more likely to engage in these activities than the residents of any other
hall. Students living in Mayo and Wilson Hall were the least likely to visit in the rooms
of residents of a different racial background. Mayo Hall has the second highest
proportion and Wilson Hall has the third highest proportion of non-Hispanic White
residents on campus. Students living in Wilson Hall were also the least likely to visit with
residents of a different racial background in their own rooms.

Comparable results were found among residents who study with students of a
different racial background. Students living in Rather and Holmes Hall were the most
likely to participate in this activity while students residing in Landon and Wilson Halls
were the least likely to do so.

On occasion, the proportion of freshmen compared with upperclassmen appeared
to play a role as well. For instance, residents of Rather and Holmes Hall were

significantly more likely to attend events on and around campus with students of a

240

different racial background. Residents of Emmons Hall, followed by those of Wilson Hall
were the least likely to do so. More than 90 percent of the residents in Emmons Hall are
freshmen. Over 40 percent of residents in Wilson Hall are upperclassmen; however,
nearly 90 percent of its residents are White. For this item, it appears that both race and
academic standing play a meaningful role in determining which residence halls exhibit
the highest rates of participation.

When we consider the Interactional Diversity scale, the mean score for this scale
was 22.3 out of 40 possible points. This indicates that the average student on the campus
of Michigan State University engages in interactional diversity occasionally. Upon closer
inspection, the data suggests that for all racial groups, the levels of interactional diversity
are most easily influenced among White students. The bivariate relationship between the
Interactional Diversity scale and race revealed that White students were the least likely of
all racial groups to engage in interactional diversity. This finding is consistent with the
literature (Antonio 2001; Chang 1996; Chang, Astin, and Kim 2004; Saenz, Ngai, and
Hurtado 2007).

It is important to remember that levels of interactional diversity among White
students may not be entirely reflective of White racial attitudes. Part of interactional
diversity is opportunity, and if White students are not given the opportunity to come into
frequent contact with students of color, the likelihood that they will engage in
interactional diversity diminishes. At a diverse institution such as Michigan State
University, the issue of interactional diversity is still largely optional for White students
who comprise three-fourths of the student body; whereas for students of color, interacting

with peers of a different racial background is a part of their day-to-day experiences.

241

National data on interactional diversity among college students has found
inconsistent results when comparing students of color. Data from the 1994 Cooperative
Institutional Research Program, Chang, Astin and Kim (2004) revealed that African
American students were less likely to engage in interactional diversity than other non-
White students (Antonio 2001; Chang, Astin, and Kim 2004). In contrast, data from the
Preparing Students for a Diverse Democracy Project revealed that Afiican Americans
were more likely to engage in interactional diversity than Latino and Asian students
(Saenz, Ngai, and Hurtado 2007). Using data from the National Longitudinal Study of
Adolescent Health, Joyner and Kao (2000) found that when the opportunity structure of
secondary schools was controlled for, Hispanic and Native American students were the
most likely to have interracial fiiendships, followed by White students. They found
African American and Asian students were the least likely to have interracial fiiendships.

In the current study, Black students were less likely than other students of color to
participate in meaningful and ongoing activities across race lines. Research in the
California State University system suggests that students of color are less likely to
interact across race lines as their proportions increase on campus (Cowan 2005). For
example, on campuses where Black students were the largest minority group, they were
the least likely to engage in interactional diversity; on campuses where Latinos were the
largest non-White group, they exhibited the lowest levels; and so on. This might explain
why on the campus of Michigan State University, Black students were the least likely
students of color to interact across race lines.

Among all racial groups in the study, Mixed Race students were the most likely to

engage in interactional diversity. This information is a new contribution to the literature

242

because multiracial participants are typically not included in the analysis. One possible
explanation for this finding is that Mixed Race persons, in general, have higher rates of
interactional diversity than other racial groups. By their very nature, multiracial
individuals have been forced to navigate within and across racial groups throughout their
lives. Challenges and questions are likely to surface from within their own families, fi'om
their peers and teachers throughout their education, and from the public at large (Kerwin,
Ponterotto, Jackson, and Harris 1993; Stephan 1992; Tatum 1997). In addition,
multiracial children tend to be more culturally sensitive than other racial/ethnic groups
which may also contribute to their increased willingness to develop meaningful and
sustained relationships with peers of a different racial background than their own
(Kerwin, Ponterotto, Jackson, and Harris 1993). Another possible explanation is that
Mixed Race students comprise a small proportion of the student body on the campus of
Michigan State University. Although university records do not currently have
information on the number of students that identify as multiracial, it can be assumed that
their numbers are lower than African American (9.6%) and Asian American (6.0%)
students on campus. As discussed above, their levels of interactional diversity may also
be related to their proportionately low numbers on campus.

The relationship between interactional diversity and gender was also considered.
Although the bivariate correlations indicated a significant relationship between these
variables, the relationship disappeared in the multivariate model when we controlled for
the current residence hall and racial enviromnent of the hall.

When we controlled for the classroom experience, the relationship between

gender and interactional diversity persisted among White students. White men were

243

significantly more likely than White women to engage in interactional diversity when the
classroom experience is controlled for. This relationship was not found for any of the
other racial/ethnic groups. Other research has found inconsistent results for the effects of
gender on interactional diversity (Antonio 2001; Cowan 2005; Johnson and Marini 1998;
Koehler 2001; McClelland and Linnander 2006; Milem, Umbach, and Liang 2004;
Saenz, Ngai, and Hurtado 2007).

Bivariate correlations and multivariate regressions suggested that academic
standing was positively and significantly related to interactional diversity. These findings
have also been found in other research. Chang, Astin, and Kim (2004) found that levels
of interactional diversity tended to increase with academic standing for all racial groups.
In the current study, the effects of academic standing were only significantly related to
the interactional diversity levels of White and Asian students. These effects were not
found for Black, Latino, or Mixed Race students.

The amount of time that students spend socializing was found to be significantly
and positively related to interactional diversity in bivariate and multivariate regressions.
Saenz, Ngai and Hurtado (2007) found these variables to be significantly related among
Latino, Asian, and White students but not among African Americans. In the current
study, hours spent socializing was significant for White, Black, and Asian students but
not for Latino and Mixed Race students, when we controlled for current residence hall. It
was found to be significant for all groups except Latinos when we controlled for the
classroom experience.

In the multivariate analysis, students’ neighborhood demographics and the

demographics of students’ junior high and high schools were found to be significantly

244

related to interactional diversity when we controlled for current residence hall. Subgroup
regressions in the current study revealed that the neighborhood demographics were only
significant predictors for the interactional diversity levels of White students. Similarly,
the racial demographics of students’ junior high and high schools was significantly
related to interactional diversity when we controlled for the classroom experience. Once
again, this relationship proved to be significant only among White students. Previous
research has shown that segregation which occurs in educational settings is likely to be
replicated throughout the life course (Braddock 1985; Gurin 1999; Milem, Umbach, and
Liang 2004). Similarly, Bonilla-Silva (2003) argues that segregation frequently leads to
the development of a White habitus for many Whites. As a result, Whites’ attitudes,
perceptions, and feelings towards minorities greatly reduce their likelihood of developing
interracial friendships over time. In their analysis of the Diverse Democracy Project,
saenz, Ngai, and Hurtado (2007) found that the pre-college environment was a
significant predictor of interactional diversity for Black students but not for Asian,
Latino, or White students. The findings from this study suggest the effects of segregation
differ substantially between White students and students of color, at least in terms of their
desire and willingness to develop friendships across race lines.

Regarding family background, multivariate regressions indicated that mother’s
education was significantly and positively related to interactional diversity, while family
income was significantly and negatively related to interactional diversity, when we
controlled for current residence hall. Subgroup regressions indicate that the effects of
mother’s education are only significant for White students. In contrast, there was no

interaction between race and family income. This suggests that once we control for the

245

 

current residence hall, the likelihood of a student engaging in interactional diversity
diminishes as their family income increases, regardless of the student’s racial/ethnic
background.

Mother’s education was found to significantly increase the likelihood for
interactional diversity, and family income was found to significantly reduce the
likelihood of interactional diversity, when we controlled for experiences within the
classroom. In both cases, subgroup regressions revealed that these variables were only
significant for White students. Results from this study related to parents’ education are
consistent with the literature while those related to family income are not. McClelland
and Linnander (2006) found that White students, whose parents were more educated,
were more likely to hold positive feelings towards African Americans. In their
longitudinal research of students at a mid-Atlantic public research university, Milem,
Umbach, and Liang (2004) found that family income was negatively related to
interactional diversity among White students.

Bivariate and multivariate regressions suggest students’ experiences within the
residence hall system directly affect the likelihood that they will engage in meaningful
and ongoing contact with residents of a different racial background. The bivariate
relationship between the racial composition of the current residence hall and interactional
diversity indicated the tendency for students in more racially diverse halls to report
higher levels of interactional diversity. For examples, students living in Rather Hall—the
most racially diverse hall on campus—had the highest mean score on the Interactional

Diversity scale. Residents of Emmons Hall—one of the least racially diverse halls on

246

 

campus with a high percentage of first-year students—exhibited the lowest mean score on
this scale.

There are two primary exceptions to this trend to consider. The first appeared in
halls where Living-Learning programs were located. Students participating in these
programs spend far more time together, both formally and informally, than other students
on campus. Consistent with the organizational literature on friendship, the effects of
tracking on student friendships, and research on interracial fiiendships, students living in
these halls report higher levels of interactional diversity despite the lower levels of racial
diversity within the halls (Antonio 2004; Feldman 1981; Kubitschek and Hallinan 1998;
Moody 2001 ).

The second discrepancy occurred in Hubbard Hall. If there was a direct
relationship between the racial composition of the residence hall and the interactional
diversity of its residents, we would expect that Hubbard Hall, the second most racially
diverse hall on the campus of Michigan State University, would have higher rates of
interaction than most other halls. However, this is not the case. The most likely
explanation for this finding relates to the critical mass of Black students in Hubbard Hall
(Cowan 2005). Sigelman et al. (1996) found that casual contact and close friendships
among Whites and Blacks increased as the percentage of Black residents in a residential
neighborhood increased. However, as the ratio of Black residents approached 50 percent,
White residents became increasingly uncomfortable with their environment. Similarly,
Tyson (2004) found that students of color were significantly less likely to have interracial
friendships as the percentage of non-White residents in their residence hall increased.

Although Black residents only comprise 31 percent of Hubbard Hall’s total population,

247

48 percent of Hubbard Hall residents are students of color. Furthermore, Hubbard Hall is
quite large compared to most of the other residence halls on campus with 1,055 residents.
At the time of this study, approximately one-quarter of the Black population residing on
campus lived in Hubbard Hall. In this situation, the need for Black students to interact
across race lines is reduced while the potential for perceived threat on the part of White
students is increased.

The current residence hall continued to significantly and negatively affect the
interactional diversity scale in the multivariate analysis. Subgroup regressions revealed
that the racial composition of the residence hall is only significantly related to the
interactional diversity levels of White students. The residence halls with the greatest
percentage of White residents replicate the White habitus that Bonilla-Silva warns against
and, as a result, White students living in these halls are significantly less likely to engage
in interactional diversity (2001; 2003).

Multivariate regressions suggest that students’ perceptions of the Overall Racial
Environment of their current residence halls significantly affect their levels of
interactional diversity. This scale includes the frequency with which students perceive
and/or participate in conflicts within the hall related to racial/ethnic differences, the
extent to which students feel emotionally and physically safe in their residence hall, the
likelihood that they would recommend their current hall to fiiends of their same racial
background, the likelihood that they would live in the same hall the following year, how
easy it was for them to make fiiends with residents in the hall, and their comfort level
with the racial composition of the hall. Subgroup regressions revealed that this scale was

only significant for White, Black, and Asian students. For these students, the more

248

 

 

 

positively they viewed the current residence hall, the more likely they were to engage in
interactional diversity.

Research consistently finds that Whites are less likely to perceive incidents of
discrimination than non-Whites. The majority of Whites today believe that African
Americans are doing as well or better than they actually are in terms of jobs, income,
education, and access to health care (Cashin 2004). Furthermore, while the majority of
Whites concede that incidents of discrimination may still occur, most feel that Blacks no
longer face serious institutional discrimination (Bobo 2001; Feagin 2000; Feagin and
O'Brien 2003; Rubin 1994; Schuman and Steeh 1996).

Similar attitudes have been found on college on predominantly White university
campuses where White students are less likely to perceive racial tension than students of
color (Ancis, Sedlacek, and Mohr 2000; Chavous 2005; Hughes, Anderson, Cannon,
Perez, and Moore 1998; Hurtado 1992; Nora and Cabrera 1996). Koehler (2001) found
that Black students were more tolerant of the racial composition of their residence halls
than White students. This difference is consistent with the broader housing literature
(Bobo 2001; Cashin 2004; Schuman and Steeh 1996). Koehler (2001) also found a
positive relationship between interactional diversity and students’ desire to live in a
racially diverse residence hall, regardless of the student’s race. Previous research has
found that White and Black students are more likely to have interracial friendships if they
perceive such friendships as normative on campus (Chavous 2005). Students that
perceive racial tension on campus are less likely to engage in interactional diversity
(Saenz, Ngai, and Hurtado 2007). Research also suggests that perceived racial conflict in

the residence hall leads to a decrease in interaction across race lines as well as a decrease

249

in students’ comfort (Johnson-Durgans 1994). The current findings suggest that White,
Black, and Asian students are especially susceptible to the effects of the residence hall
environment.

Bivariate and multivariate regressions indicated that the first-year roommate plays
a significant role in the likelihood that students will engage in meaningful and ongoing
contact with peers fi'om different racial backgrounds. Previous studies have shown first-
year roommates positively affect interracial fiiendships although the strength of this
relationship has been inconsistent (McClelland and Linnander 2006; Tyson 2004; Van
Laar, Levin, Sinclair, and Sidanius 2005).

In the current study, subgroup regressions reveal that the effects of the first-year
roommate vary across racial groups. Among White and Asian students, those that were
randomly assigned a first-year roommate, regardless of race, and those that requested
their first-year roommate, and whose roommate is of a different racial background than
themselves, were significantly more likely to engage in interactional diversity than those
who requested a first-year roommate that shares their same racial background. Among
Black students, only those students who were randomly assigned a roommate of a
different racial background were significantly more likely to engage in interactional
diversity. Among Latino students, having a first-year roommate of a different racial
background—whether that roommate had been requested or randomly assigned—
significantly increased the likelihood that they would engage in interactional diversity.
Finally, among Mixed Race students, only those students who had requested their first-
year roommate, and whose roommate is of a different racial background—were

significantly more likely to engage in interactional diversity.

250

The last independent variable to consider is the effects of the Classroom
Experience scale on students’ levels of interactional diversity. The Classroom Experience
scale measures the extent to which students’ values and beliefs have been challenged in
classes, the extent to which their understanding of students different from themselves has
been broadened, the extent to which their views have been changed through
conversations with peers of a different racial/ethnic background in their classes, the
frequency with which they have conversations with peers of a different racial/ethnic
background in their classes, and the frequency with which they sit next to students of a
different racial/ethnic background in their classes. The bivariate and multivariate analyses
indicated that the experiences which students have in the classroom significantly predict
the likelihood that they will engage in interactional diversity outside of the classroom.
Subgroup regressions reveal the Classroom Experience significantly affects the
interactional diversity levels of all racial groups.

In the following section, I will address the implications of these findings.
Implications

The findings from this study point to three key implications. First, the levels of
interactional diversity differ significantly between racial groups. While this finding in
itself is not unique, the specific details related to this issue make a meaningful
contribution to the literature. White students were the least likely to report engaging in
the activities outlined in the Interactional Diversity scale. It is important to reiterate that
levels of interactional diversity among White students may not be entirely reflective of
White racial attitudes. Part of interactional diversity is opportunity, and if White students

are not given the opportunity to come into frequent contact with students of color, the

251

likelihood that they will engage in these activities diminishes. At an institution like
Michigan State University, where Whites comprise three-fourths of the student body,
interactional diversity is largely optional for White students; whereas for students of
color, interacting with peers of a different racial background is a part of their day-to-day
experience.

With that in mind, the levels of interactional diversity among White students were
more easily influenced than any other students. Every independent variable considered in
this study significantly affected the likelihood of interactional diversity among White
students. What this suggests is that if White students are not exposed to racial diversity
prior to attending college, they are significantly less likely to take advantage of that
diversity once they reach college. Additionally, if White students are not consistently
exposed to racial diversity on the college campus, either through the racial background of
their roommate, the racial composition of their residence hall, or in the classroom setting,
they are far less likely to seek out interracial relationships on their own.

Among students of color, Black students were the least likely to engage in
interactional diversity. Research on the interactional diversity on non-White students
varies in its findings. It may be that the proportion of Black students on the campus of
Michigan State University is large enough to reach a critical mass, reducing their
likelihood to regularly interact with students of a different racial background.

In the current study, Mixed Race students were the most likely to engage in
interactional diversity. This finding is a new contribution to the literature which has
largely excluded this population of students. There a number of possible explanations for

this finding related to the multiple identities of mixed-race individuals. Additionally,

252

 

 

 

because Mixed Race students comprise a small percentage of the student body at L
Michigan State University, their higher levels of interactional diversity may be a function
of their numerical status.

The second key finding from this study is that levels of interactional diversity are
significantly affected by first-year roommates for all students. Students who request their
first-year roommate, and whose roommate is of their same racial background, are
significantly less likely than other students to engage in interactional diversity. This is
especially true for White and Asian students. The vast majority of colleges and
universities allow their students to request their first-year roommate. Particularly on a
large campus, such as Michigan State University, having a familiar face in the dorm room
can substantially reduce students’ feelings of alienation and isolation. It is unlikely that
this housing policy will change. However, at public institutions, where the majority of
students come from within the state, it is important to consider students’ exposure to
racial diversity prior to attending college. The state of Michigan, for example, is the most
segregated in the country. The likelihood that students will have attended a racially
diverse high school is, therefore, significantly lower than for students from any other
state. It follows, then, that students will also be less likely to request a first-year
roommate who is not of their same racial background. If campus housing policies
continue to allow students to request their first-year roommate (and there is good reason
to continue allowing students to do so), administrators should be mindful that this policy
could substantially reduce students’ likelihood to engage in interactional diversity. As a
result, increased efforts should be made to regularly bring students from different

backgrounds together in other ways.

253

The third and, in this researcher’s opinion, most encouraging finding from this
study is that the classroom is a critical component of the interactional diversity puzzle.
Previous research has shown the importance of exposing students to a diversified
curricula and providing them with opportunities to interact with peers who are different
from themselves. The current study confirms that the effects of these courses are
cumulative. The Classroom Experience scale—which takes into account both curriculum
and pedagogy—proved to be a significant predictor of interactional diversity for all
students. The analyses indicate that the more often students are exposed to these types of
courses the more likely they will be to engage in interactional diversity outside of the
classroom, regardless of their racial or ethnic background. While it is certainly important
to encourage students to enroll in courses which focus on issues of diversity and
inequality, faculty across disciplines should also be encouraged to incorporate an active

pedagogy into their courses.
Limitations and Direction for Futgre Studies

There are several limitations that can arise when conducting survey research,
particularly its focus is racial attitudes. First, the data was only collected one time.
Research on racial attitudes has shown that longitudinal data is more reliable than cross-
sectional data (Pettigrew 1998; Pettigrew and Tropp 2000). Understanding the extent of
the relationship between the independent variables and interactional diversity is limited.
For example, while there is clearly a significant relationship between the Overall Racial
Environment scale and the Classroom Experience scale and the Interactional Diversity
scale, cross-sectional data prohibits us from knowing the causal relationship between

these variables.

254

 

 

The second limitation to the data is the issue of social desirability bias. Research
on racial attitudes and behaviors are typically vulnerable to social desirability bias,
especially from White and highly educated respondents (Krysan 2000; Sears 1997).
Moreover, research on interracial fiiendships is fraught with reporting bias (Bonilla-Silva
2003; Smith 2002). While many steps were taken to include a variety of student and
campus organizations when promoting this survey, it is possible that some participants
continued to perceive the survey in a particular light. As a result, those students who
participated in the study may have inflated the extent to which they interact with peers
across race lines. Questions that appear in the Interactional Diversity scale are expected
to yield a minimum response bias. While students may exaggerate the frequency with
which they socialize or party with students of a different racial background, questions
that relate to having intellectual discussions and sharing personal feelings are usually
answered honestly (Bonilla-Silva 2003). By incorporating all of these single-item
questions into a scale, the overall effects of social desirability should be minimized.

The third limitation is the potential for self-selection bias. One of the steps taken
to minimize a response error was in sampling the entire population. By employing a
nonprobability sample design, the likelihood of nonresponse bias is reduced (Groves
2006). The racial background, gender, class standing, and current hall of the 33 percent of
students who participated in this study are very similar to the total on-campus population.
Non-Hispanic White students were somewhat more likely to participate in the study than
students of color, but this difference was not significant. In terms of basic demographics,
the effects of nonresponse are minimal. However, what we do not know is how these

students differ in terms of values, beliefs, and prejudice from those students that did not

255

participate in the study. This could be especially important for students of color. For
example, how do the attitudes and experiences of Black students living in Hubbard Hall
compare between those who participated in the study and those who did not? Are the
levels of interactional diversity for those students even lower than what the data already
suggests?

Finally, there is the issue of missing data. There are at least 400 cases missing
from each question of the survey and this number increases to more than 700 by the end
of the survey. The overall proportion of those who completed the last question, however,
is very similar to the proportion that completed the first question. These proportions are
also similar to the overall on-campus population, with non-Hispanic Whites and
lntemational students somewhat overrepresented and African Americans and
Chicano/Latinos slightly underrepresented. The increased drop-out rate does not appear
to be related to any particular issue; rather, it seems to be an indication of survey fatigue.
Women were somewhat more likely to complete the survey than men. Completion rates
based on gender are likely to be related to several issues: survey fatigue, higher rates of
survey participation among women more generally, as well as the salience of the topic.

From a methodological perspective, there are a number of important adjustments
to make in future studies. First, there are two demographic issues that I intend to change.
In the current study, the domestic status of respondents was not considered. On the one
hand, less than 4 percent of the participants were identified through university records as
lntemational students. Furthermore, Tyson (2004) did not find significant differences in
the rate of interracial fiiendships among first-year college students. On the other hand,

there is strong evidence that the experiences and perspectives of immigrants differ

256

 

substantially fiom U.S. born citizens (Ishii-Kuntz 2000; Stepick 1997; Waters 1999). It is
a limitation of this study that responses of African students cannot be teased out from
those of Afiican Americans, for example, or Chinese students from Asian Americans.
Furthermore, it is recommended that the open-ended racial/ethnic question be replaced
with a categoric variable. Although the open-ended question is more inclusive, the
extensive coding required for this item did not prove methodologically advantageous. In
future projects, I intend to include all of the standard Office of Management and Budget
categories, as well as other marginalized populations—for example, Jews and Arab
Americans——and allow participants to identify with all applicable categories.

The second methodological concern that presented itself in this study was
validity. There were several “perceptions” that I attempted to measure using one-item
questions on issues related to campus climate, the racial environment on campus, and
others. Although the bivariate analysis revealed significant differences for each of these
variables, it cannot be assumed that participants interpreted these questions similarly.
While the frndings discussed here are consistent with much of the literature and certainly
contribute to our understanding of race relations on the campus of Michigan State
University, they cannot be used to draw any meaningful conclusions. For this reason, no
further analyses were conducted on these items. In future projects, it will be necessary to
design a series of questions that can be collapsed into issue-specific scales.

Upon reviewing the item-analysis within the Interactional Diversity scale, two
topics strike me as worthy of closer inspection. First, among all students, the level of
interactional diversity was greatest for socializing and/or partying with students of a

different racial background. Students were more likely to engage in this activity than any

257

 

other on the scale. What is it about this particular activity that makes students more
willing to participate? Related to this, White students were less likely than all other
students to engage in every activity within the scale except this item. Among students
who socialize and/or party across race lines, Black students were the least likely to
describe doing so. Why?

Second, of the items within the Interactional Diversity scale, students were least
likely to attend events on campus sponsored by racial/ethnic groups different fiom their
own and to have meaningful discussion about racial/ethnic issues with students of a
different racial background. In the current study, there is not a “control” scale. In other
words, participants were not asked the extent to which they engage in these activities
overall. Therefore, it is impossible to know if students are less likely to engage in a
particular activity with peers of a different racial background than they would with peers
of their same racial background. I intend to include this additional scale in future studies.

At the outset of this project, I hoped to provide insight to these findings from the
students themselves. Despite taking steps to encourage student involvement, participation
in the focus groups was extremely limited. In order to maximize our understanding of the
data, it will be important to continue reaching out to participants. This will most likely
require increasing the incentive opportunities available to students (James and Bolstein
1990)

The last significant lesson of this study to consider is the importance of social
capital in effectively conducting this research. Although a 33 percent response rate is not
especially high compared to surveys of the general public, this rate is consistent with

online surveys conducted at large, research-focused, public institutions such as Michigan

258

State University (Porter and Umbach 2006). These results could not have been achieved
without the substantial support of student organizations, the Residence Halls Association,
and the diligent recruitment efforts of the MRULE student leaders. At an institution like
Michigan State University that is so large, gaining access to the students first requires the
support of their peers. Additionally, my ability to conduct this research would have been
greatly hindered had I not had the support of several departments at Michigan State
University including Student Affairs; Residence Life; Housing; and the Office for
Inclusion and Intercultural Initiatives (formerly the Office of Affirmative Action,
Compliance, and Monitoring). Their support enabled me to coordinate the data collection
process as well as facilitate a more in-depth analysis of the findings. In future research
projects, it will be critical to achieve similar connections.

Contributions of the Research

The focus of my dissertation contributes to a variety of fields. First and foremost,
it highlights significant issues that campuses nationwide will be forced to consider as
their student populations inevitably become more diversified. Results from this
dissertation provide insight into a topic that, thus far, few researchers have considered. In
fact, this is the first study to explore the relationship between campus housing and student
race relations to this extent. As a case study, it also provides direction for future studies to
build upon. As the number of studies increases and consider a variety of campus
environments, these results will help enable the implementation of sound housing and
diversity policies on college campuses. By examining the extent to which students

interact across racial lines, the quality of that interaction and the role that residential

259

segregation plays in shaping both of these issues, this dissertation also helps to shed light

 

on the future of race relations more generally.

260

 

 

 

APPENDIX A: IPPSR Notification Letter E

 

August 18, 2009
Dear «first»:

I am writing to ask you to participate in The 2006 MSU Community Survey that is being
conducted with Michigan State University undergraduate students currently living in the
residence halls. This study is part of an on-going effort to better understand student
relationships and experiences at MSU as they relate to living in the residence balls. The
study is being conducted by Jessica Mills, PhD Candidate, and the Multi-Racial Unity
Living Experience (MRULE). It has the support of several departments at MSU including
Student Affairs, Residence Life, Housing, Affirmative Action, Compliance and
Monitoring; as well as the Residence Halls Association.

Every undergraduate student at MSU who is currently living on—campus has been invited
to be part of this important project. We will begin data collection on «Date». I will be
sending an email to your MSU account on that day which will contain the link to access
the web survey and your ID number. The survey will be available on-line for about 10
days for your convenience.

Please accept the enclosed coupons for free and discounted items from East Lansing
merchants as a token of our appreciation for completing the survey. In addition to the
enclosed items, if you complete the survey before noon on «Date», we will enter
your name into a drawing to win an iPod nano 2GB.

We will be asking questions about classroom, work, and residence hall experiences,
particularly as they relate to interacting with people of different backgrounds. Although
this is primarily a dissertation, the results of this study may also be useful to current and
prospective MSU students regarding campus life and will be available to the University
administration in their efforts to better understand students’ perceptions and experiences.

The software being used to collect the data has several security features ensuring that
your responses are completely confidential. The database containing participant names
will be destroyed upon completion of data collection and no identifying information will
ever be part of the data. We will protect your privacy to the maximum extent allowable
by law.

 

Participation in this research is completely voluntary, but greatly appreciated and vital
to the overall success of this project. You may choose to not participate, to skip any
question that you do not want to answer, and to end your participation at any time.

If you have any questions about this study, please contact the investigator, Jessica Mills,
at xxx-xxxx or by email at millsies@msu.edu. If you have any questions or concerns
regarding your rights as a study participant, or are dissatisfied at any time with any
aspect of this study, you may contact — anonymously, if you wish - Peter Vasilenko, Ph.D.,

261

 

Director of the Human Subject Protection Programs at Michigan State University, by
phone: (517) 355-2180, fax: (517) 432-4503, email: irb@msu.edu, or regular mail:
202 Olds Hall, East Lansing, MI 48824.

Thank you in advance for your valuable cooperation.

Sincerely,

Karen Clark
Project Manager

262

APPENDIX B: IPPSR First E-Mail Notification

 

Dear «Name»:

Last week I sent you an email inviting you to participate in The 2006 MSU Community Survey. I
would like to again invite you to participate in this important research.

This research is part of an on-going effort to better understand student relationships and
experiences as they relate to living in the residence halls. It is important that our research includes
opinions fi'om a wide range of MSU students.

Even though participation in this research is completely voluntary, I am hoping you will take 10
minutes in the next couple of days to participate.

To access the on-line questionnaire please click on the following link: «link»

Your logon ID is: «logon ID»

We will enter your name in a drawing for an iPod nano 20B if you complete the survey by noon
on «Date».

If you have any questions about this research or would like a paper copy of the survey to
complete, please contact me at 517.353.1764 or by email at: clarkk@msu.edu.

Thank you!

Karen Clark

Project Manager

Office for Survey Research
Michigan State University

263

APPENDIX C: IPPSR Second E-Mail Notification

 

Dear «Name»:

Before I end data collection on the 2006 MSU Community Survey on «Date» at 12:00 pm, I
wanted to give you a final opportunity to participate in this important research and to be placed in
the drawing for the nano iPod ZGB.

This study is part of an ongoing effort to better understand student relationships and experiences
at MSU as they relate to living in the residence halls.

To participate, please click on the following link: «Link».

Your logon ID is: «Logon ID»
Thank you!
Karen Clark

Project Manager
Office for Survey Research

264

APPENDIX D: The 2006 MSU Community Survey

 

The 2006 MSU Community Survey

This survey is being conducted by Jessica Mills, PhD Candidate, and the Multi-Racial
Unity Living Experience (MRULE). This survey has the support of several
departments at Michigan State University including Student Affairs, Residence Life,
Housing, the Office of Affirmative Action, Compliance, and Monitoring as well as
MRULE (Multl-Racial Unity Living Experience) and the Residence Halls Association
(RHA).

We are interested in assessing your experiences and perceptions in living, classroom
and work environments particularly as they relate to interacting with people of
different backgrounds.

The survey will take approximately 10 minutes of your time to complete. After you
complete each section, click the “next" button and proceed to the following section.
All of your responses will be kept strictly confidential and your privacy will be
protected to the maximum extent allowable by law.

If you have any questions about the research, please contact Jessica Mills at
m1l_lsj_es@_msg._e_d_u. If you have any technical problems with the survey or would
prefer a paper copy to complete, please contact Karen Clark at glarkamsggdu.

 

I indicate my voluntary consent to participate in this research.

Yes
No

265

Section A: Current Classroom Experience

1. For this section, please think about your daily experiences that occur outside of
your residence hall. For each of the following, please indicate how often you have
experienced each in your classes at MSU -- never, occasionally, often, or very
often.

Very

Never Occasionally Often Often

The majority of students in my classes are of
my same racial background.

I have experiences in my classes that challenge
my values/beliefs.

I have experiences in my classes which
broaden my understanding of students
different from myself.

My views have been changed because of
conversations with students of different
racial/ethnic backgrounds than myself in my
classes.

I have conversations with students of a
different racial background than myself in my
classes.

I sit next to students of a different racial
background than myself in my classes.

Section B: Current Work Experience
2. Have you been employed at any time during this academic year?

Yes
NO

266

3. The next set of statements refers to experiences that you might have in a work
environment. For each, please indicate how often you have experienced each at
your job -- never, occasionally, often, or very often.

. Very
Never Occasnonally Often Often
I work with people at my job of my same racial
background.

I have experiences at work that challenge my
values/beliefs.

I have experiences at work which broaden my
understanding of people different from myself.

My views have changed because of
conversations with people of different
racial/ethnic backgrounds than myself at work.

I have had conflicts with an employee or
customer at work that I believe are based on
the difference in our race/ethnicity.

There are conflicts with people at work related
to racial/ethnic differences.

Employees talk directly to one another about
racial/ethnic issues of conflict.

I have conversations with people of a different
racial background than myself at work.

I socialize outside of work with co-workers of
different racial backgrounds than myself.

267

Section C: Current Residential Experience

4. Now, please think about the experiences you have had in your current
residence hall. ,

For each of the following examples, please indicate how often each occurs in your
current residence hall -- never, occasionally, often, or very often.

Very

Never Occasionally Often Often

There are conflicts between groups on my
floor and/or in my residence hall related to
racial/ethnic differences.

There are conflicts between individuals on my
floor and/or in my residence hall related to
racial/ethnic differences.

Residents talk directly to one another about
racial/ethnic issues of conflict in my residence
hall.

I dine or share meals with students of a
different racial background than myself.

I visit with residents of a different racial
background than myself in my room.

I visit with residents of a different racial
background than myself in their room.

I have had conflicts with someone on my floor
and/or in my residence hall that I believe are
based on the difference in our race/ethnicity.

268

S. For the next set of questions, please indicate how accurately each statement
reflects your experience in your current residence hall.

Please indicate how strongly you agree or disagree with each statement.

Strongly

. Strongly
Disagree Disagree Agree Ag

ree

I feel emotionally safe in my residence
hall.

I feel physically safe in my residence hall.

I would recommend this residence hall to
my friends of the same racial background
as myself.

If I were to live on campus next year, I
would want to live in the same residence
hall.

It has been easy to make friends in my
current residence hall.

I feel that my choice of residence hall is
strongly influenced by my peers of my
same racial background.

I am comfortable with the racial
composition of my residence hall.

6. Within your residence hall, how many students do you think are of your same
racial background?

None or very few

Many, but less than half
About half

More than half

Most or all

7. Are you required to live in your current residence hall due to your participation In
a Living-Learning program such as Lyman Briggs, RISE, ROSES, or James Madison?

Yes
No

269

 

 

8. Did you request to live in your current residence hall?

Yes
No

9. Which of the following factors influenced your choice of residence hall?

I already knew students who lived here.
I lived here last year.
Located close to my classes or other campus events.

For the amenities/services offered (e.g., cafeteria, computer lab, exercise
facilities).

A desire to live with other students of my same racial background.
A desire to live with students of racial backgrounds different than my own.

A preference to live in a diverse environment.

270

Yes No

Section D: Overall Experiences at MSU

10. Next, please think about your experiences both at MSU and within your
current residence hall. Please indicate how often each of the following occurs --

never, occasionally, often, or very often.

I have meaningful and honest discussions
about race and ethnic relations with students
from racial backgrounds different than my own
outside of class.

I share personal feelings and problems with
students of a different racial background than
myself.

I study or prepare for class with students of a
different racial background than myself.

I socialize or party with students of a different
racial background than myself.

I have intellectual discussions with students of
a different racial background than myself
outside of class.

I attend events with students of a different
racial background than myself.

I attend events sponsored by other
racial/ethnic groups that differ from my own
racial/ethnic background.

271

Very

Never Occasionally Often Often

11. Still thinking about your experiences both at MSU and within your residence
hall, please indicate how strongly you agree or disagree with each statement.

I feel connected to a formal or informal
group of other students.

(Examples of formal or informal groups
include fraternities, sororities, intramural
sports teams, or student organizations).

I have broadened the diversity (e.g., race,
sexual orientation, background, religious,
culture, etc.,) of my immediate circle of
friends.

I have become more aware of my personal
prejudices and stereotypes towards
others.

There is racial tension on campus.

Strongly
Disagree

Strongly

Disagree Agree Agree

12. In your opinion, would you say that the current racial climate on campus is

poor, fair, good, or excellent?

Exceflent
lGood
Fan
Poor

272

Section E: Issues of Prejudice and Discrimination
13. Below is a list of examples in which you may have experienced discrimination.

For each of the following situations, please indicate whether you have ever
experienced discrimination, have been prevented from doing something, or have
been hassled or made to feel inferior because of your race.

Yes No
Getting a job?
At work?
From your family?
Getting medical care?
Shopping at a store?
From the police?
From teachers or professors?
From an employer or your boss?
From a peer?
From a stranger?

From a neighbor?

14. Have you ever been . . .

Yes No
Accused or suspected of doing something wrong because of your race?
Called racist names by someone?

Made fun of or harmed because of your race?

273

Section F: Demographics
15. Are you . . .

Male
Female

16. What is your age as of your last birthday?
years

17. What is your race/ethnicity?

18. What is your current academic standing?

Freshman
Sophomore
Junior
Senior

19. What is your current GPA?

I I
EFNNPP’
Hmomom

“NNPPP
DINO-.500

g.

.5
20. In what residence hall do you currently live?
21. How many semesters have you lived in your current residence hall?

22. On average, how many hours per week do you spend socializing and/or
partying?

0 hours

1-5 hours

6-10 hours

11-15 hours

16-20 hours

More than 20 hours

274

23. In your first year at MSU, did you request a specific roommate?

Yes
No

24. In your first year at MSU, was your roommate of a different racial background
than yourself?

Yes
No

25. Is this your first year at MSU?

Yes
No

26. Did you request your current roommate?

Yes
No

27. Is your current roommate of a different racial background than yourself?

Yes
No

28. What was the racial composition of . . .

Evenly
A" More 32233.2. 9.2:. A"
or nearly students of or nearly
. White students _
all color students than all
students than White and students of White
of color students students of color students
color
The neighborhood
where you grew
up?

Your elementary
school?

Your junior high
and high school?

275

29. Prior to coming to MSU, what was the racial composition of your closest friends?

All or nearly all students of color

More students of color than White students

Evenly divided between White students and students of color
More White students than students of color

All or nearly all White students

30. Since coming to MSU, what is the racial composition of your closest friends?

All or nearly all students of color

More students of color than White students

Evenly divided between White students and students of color
More White students than students of color

All or nearly all White students

31. What is the highest education level your father completed?

Did not complete high school

High school graduate or equivalent

Some college

Associate's degree

College graduate

Post-graduate/Professional degree (e.g., JD, MD, MA, MS, PhD)
Not sure

32. What is the highest educational level your mother completed?

Did not complete high school

High school graduate or equivalent

Some college

Associate’s degree

College graduate

Post-graduate/Professional degree (e.g., JD, MD, MA, MS, PhD)
Not sure

33. What is your best estimate of your family's annual income?

Less than $20,000
$20,000 to $ $39,999
$40,000 to $59,999
$60,000 - $79,999
$80,000 - $99,999
$100,000 - $119,999
Over $120,000

276

We are planning to conduct several focus groups to further address the issues that
are raised in this survey. If you are interested in participating in a focus group next
semester, please click on the link below.

 

YES
NO

277

Thank you for your interest in participating in a focus group regarding how race and
ethnicity affect your decisions and experiences at MSU.

1. Please provide the following information so that we may be in touch with you at
a later date.

Email address:
First Name:

2. If it is possible to arrange, would you prefer to participate in a focus group with
students that share
your same racial background?

Yes
No

3. Which one of the following racial backgrounds do you identify with most closely?

Non-Hispanic White

African American

Arab American

Asian American

Latino/Chicano

Native American

Mixed Raced/Multiracial
International/Non-U.S. student

4. Focus groups will be conducted on-campus during the Spring Semester (late
January/early February). In order to help us schedule the focus group meetings,
please choose four (4) of the following meeting times based on your preference
and availability.

1st Choice
2nd Choice
3rd Choice
4th Choice

Thank you very much for completing the 2006 MSU Community Survey!

278

 

APPENDIX E: MSU Community Survey Recruitment Flyer

 

  
 

campus?

MAKE YOUR VOICE HEARD
MSU Ccmnuniw Sway

279

APPENDIX F: MSU Community Focus Group Questions

 

1. What is your impression of the racial climate on campus? Explain.
2. Do you think there is any racial tension on campus? Explain.

3. Do you think housing assignment plays a role in students’ relationships with peers
of different racial/ethnic backgrounds? Explain.

280

REFERENCES

1994. "News and Views: Notes on the New Segregation in Higher Education." The
Journal of Blacks in Higher Education: 13.

1995. "Minority Dorms Serve Real Needs of Students; Pressure to Segregate." Pp. A24
in New York Times.

1996. "Not for everyone." Pp. 4 in Bay State Banner, vol. 31.

1998. "News and Views: Harvard's Self-Examination on Race Relations." The Journal of
Blacks in Higher Educationz69.

2001. "News and Views: Racial Segregation in Award Ceremonies at Louisiana State
University." The Journal of Blacks in Higher Educationz6l.

Allport, G. 1954. The Nature of Prejudice. Reading, MA: Addison-Wesley.

Ancis, Julie R, William E Sedlacek, and Jonathan I Mohr. 2000. "Student perceptions of
campus cultural climate by race." Journal of Counseling and Development : JCD
78:180.

Antonio, A. L. 2001. "Diversity and the influence of friendship groups in college."
Review of Higher Education 25:63-89.

—. 2004. "When does race matter in college friendships? Exploring the role of race
within men's diverse and homogenous friendship groups." Review of Higher
Education 27:553-75.

Antonio, Anthony Lising, Mitchell J Chang, Kenji Hakuta, David A Kenny, and et al.
2004. "Effects of Racial Diversity on Complex Thinking in College Students."
Psychological Science 15:507.

Astin, A. W. 1993. "Diversity and multiculturalism on the campus: How are students
affected?" Change 23:44-49.

281

Blau, P. M. 1994. Structural contexts of opportunity. Chicago: University of Chicago
Press.

Bobo, Lawrence. 2001. "Racial Attitudes and Relations at the Close of the Twentieth
Century." in America Becoming: Racial Trends and Their Consequences, vol. 1,
edited by N. J. Smelser, William Julius Wilson, and Faith Mitchell. Washington
DC: National Academy Press.

Bobo, Lawrence and Vincent L. Hutchings. 1996. "Perceptions of Racial Group
Competition: Extending Blumer's Theory of Group Position to a Multiracial
Social Context." American Sociological Review 61:951-972.

Bogardus, Emory S. 1933. "A Social Distance Scale." Sociology and Social Research
17:265-271.

Bonilla-Silva, Eduardo. 1997. "Rethinking Racism: Toward a Structural Interpretation."
American Sociological Review 62:465-480.

—. 2001. White Supremacy and Racism in the Post-Civil Rights Era. Boulder, CO:
Lynne Reinner Publishers.

—. 2003. Racism without Racists: Color-Blind Racism and the Persistence of Racial
Inequality in the United States. Lanham, Maryland: Rowrnan & Littlefield.

Bosnjak, Michael and Tracy L Tuten. 2003. "Prepaid and Promised Incentives in Web
Surveys: An Experiment." Social Science Computer Review 21:208.

Braddock, J. H. 1985. "School Desegregation and Black Assimilation." Journal of Social
Issues 41 :9-22.

Braddock, J. H., R. L. Crain, and J. M. McPartland. 1984. "A longOterm view of school
desegregation: Some recent studies of graduates as adults." Phi Delta Kappan
66:259-264.

Cashin, Sheryll. 2004. The Failures of Integration: How Race and Class are
Undermining the American Dream. New York: Public Affairs.

282

 

Chang, M. J. 1996. "Racial diversity in higher education: Does a racially mixed student
population affect educational outcomes?" Dissertation Thesis, UCLA.

—. 1999. "Does racial diversity matter? The educational impact of a racially diverse
undergraduate population." Journal of College Student Development 40:377-395.

—. 2001. "The positive educational effects of racial diversity on campus." Pp. 175-186 in
Diversity challenged: Evidence on the impact of afiirmative action, edited by G.
Orfield. Cambridge, MA: Harvard Education Publishing Group.

Chang, Mitchell J ., Alexander W. Astin, and Dongbin Kim. 2004. "Cross-racial
interaction among undergraduates: Some consequences, causes, and patterns."
Research in Higher Education 45:529-553.

Chavous, Tabbye M. 2005. "An Intergroup Contact-Theory Framework for Evaluating
Racial Climate on Predominantly White College Campuses." American Journal of
Community Psychology 36:239.

Clark, Kenneth B and Michael Meyers. 1995. "Separate is never equal." Pp. A19 in New
York Times.

Cohen, E., J. A. Bianchini, R. Cossey, N. C. Holthuis, C. C. Morphew, and J. A.
Whitcomb. 1997. "What did students learn? 1982-1994." in Working for equity in
heterogenous classrooms, edited by E. G. Cohen and R. A. Lotan. New York:
Teachers College Press.

Cowan, Gloria. 2005. "Interracial Interactions at Racially Diverse University Campuses."
The Journal of Social Psychology 145:49.

Curtin, Richard, Stanley Presser, and Eleanor Singer. 2000. "The Effects of Response
Rate Changes on the Index of Consumer Sentiment." 64 4.

Duster, Troy. 1991. "Understanding Self-Segregation on Campus." The Chronicle of
Higher EducationzBl-2.

Farley, Reynolds and Howard Schuman. 1978. "Chocolate City, Vanilla Suburbs: Will

the Trend toward Racially Separate Communities Continue?" Social Science
Research 7:319-344.

283

Feagin, Joe. 2000. Racist America: Roots, Current Realities and Future Reparations.
New York: Routledge.

Feagin, Joe and Eileen O'Brien. 2003. White Men on Race: Power, Privilege, and the
Shaping of Cultural Consciousness. Boston: Beacon Press.

F eldman, DC. 1981. "The Multiple Socialization of Organization Members." Academy of
Management Review 6:309-318.

Feldman, N. A. and T. M. Newcomb. 1969. The Impact of College on Students, vol. I: An
Analysis of Four Decades of Research. San Francisco: Jossey-Bass.

Groves, R. M. 2006. "Nonresponse Rates and Nonresponse Bias in Household Surveys."
Public Opinion Quarterly 70:646-675.

Gurin, P. 1999. "The Compelling Need for Diversity in Higher Education."

Gurin, P., E.L. Dey, S. Hurtado, and G. Gurin. 2002. "Diversity and Higher Education:
Theory and Impact on Educational Outcomes." Harvard Educational Review
72:330-366.

Gurin, Patricia Y, Eric L Dey, Gerald Gurin, and Sylvia Hurtado. 2003. "How Does
Racial/Ethnic Diversity Promote Education?" Western Journal of Black Studies
27:20.

Gurin, Patricia Y, Biren A Nagda, and Gretchen E Lopez. 2004. "The Benefits of
Diversity in Education for Democratic Citizenship." Journal of Social Issues
60:17-34.

Hallinan, M. T. and R. A. Williams. 1989. "Interracial Friendship Choices in Secondary
Schools." American Sociological Review 54:67-78.

Heberlein, T. A. and R. Baumgartner. 1978. "Factors affecting response rates to mailed
surveys: A quantitative analysis of the published literature." American
Sociological Review 43.

Heerwegh, Dirk. 2005. "Effects of Personal Salutations in E-mail Invitations to
Participate in a Web Survey." Public Opinion Quarterly 69:588.

284

Heerwegh, Dirk, Tim Vanhove, Koen Matthijs, and Geert Loosveldt. 2005. "The effect of
personalization on response rates and data quality in web surveys." International
Journal of Social Research Methodology: Theory and Practice 8:85-99.

Herring, Cedric and Charles Amissah. 1997. "Advance and Retreat: Racially Based
Attitudes and Public Policy." Pp. 121-143 in Racial Attitudes in the 1990s:
Continuity and Change, edited by S. A. Tuch and J. K. Martin. Westport: Praeger.

Hu, S. and G. Kuh. 2003. "Diversity experiences and college student learning and
personal development." Journal of College Student Development 44:320-334.

Hughes, Michelle, Rick Anderson, Julie Harms Cannon, Eduardo Perez, and Helen A.
Moore. 1998. "Campus Racial Climate Policies: The View From The Bottom
Up." Race, Gender & Class 5:139.

Hurtado, S. 1992. "The campus racial climate: Contexts of climate." The Journal of
Higher Education 63:539-569.

—. 2001. "Linking diversity with educational purpose: How diversity impacts the
classroom environment and student development." Pp. 187-203 in Diversity
challenged: Legal crisis and new evidence, edited by G. Orfield. Cambridge, MA:
Harvard Publishing Group.

—. 2003. Preparing college students for a diverse democracy. Ann Arbor, MI:
University of Michigan, Center for the Study of Higher and Postsecondary
Education.

Hurtado, 8., EL. Dey, P. Gurin, and G. Gurin. 2003. "College environments, diversity,
and student learning." Pp. 145-190 in Higher Education: Handbook of theory and
research, edited by J. C. Smart. UK: Kluwer Academic Publishers.

Hurtado, Sylvia. 2005. "The Next Generation of Diversity and Intergroup Relations
Research." The Journal of Social Issues 61:595.

Ishii-Kuntz, Masako. 2000. "Diversity within Asian American Families." Pp. 274-292 in
Handbook of Family Diversity, edited by D. H. Demo, K. R. Allen, and M. Fine.
New York: Oxford University Press.

285

James, Jeannine M. and Richard Bolstein. 1990. "The Effect of Monetary Incentives and
Follow-Up Mailings on." Public Opinion Quarterly 54:346.

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.

Johnson, Monica Kirkpatrick and Margaret Mooney Marini. 1998. "Bridging the Racial
Divide in the United States: The Effect of Gender. " Social Psychology Quarterly
61:247-258.

Joyner, Kara and Grace Kao. 2000. "School Racial Composition and Adolescent Racial
Homophily." Social Science Quarterly 81:810-825.

Kaplowitz, Michael D., Timothy D. Hadlock, and Ralph Levine. 2004. "A Comparison of
Web and Mail Survey Response Rates." Public Opinion Quarterly 68:94-10].

Kerwin, Christine, Joseph G. Ponterotto, Barbara L. Jackson, and Abigail Harris. 1993.
"Racial identity in biracial children." Journal of Counseling Psychologz 40:221-
231.

Koch, Will. 1999. "Integration and de facto segregation in campus housing: An analysis
of campus housing policy." College Student Affairs Journal 18:35.

Koehler, Gretchen Elizabeth. 2001. "Racial residential segregation in a university
setting." 3036211 Thesis, University of South Carolina, United States -- South
Carolina.

Krysan, Maria. 2000. "Prejudice, Politics, and Public Opinion: Understanding the
Sources of Racial Policy Attitudes." Annual Review of Sociology 26:135-168.

Kubitschek, Warren N. and Maureen T. Hallinan. 1998. "Tracking and Students'
Friendships." Social Psychology Quarterly 61 :1-15.

Lange, Nancy Q. 2006. "First-Year Students Share Their Expectations For Their First
Semester at MSU." Department of Residence Life, Michigan State University,
East Lansing, MI.

286

 

—. 2007. "F irst-Year Students’ Reflections on Fall Semester 2006." Department of
Residence Life, Michigan State University, East Lansing, MI.

McClelland, Katherine and Erika Linnander. 2006. "The Role of Contact and Information
in Racial Attitude Change among White College Students*." Sociological Inquiry
76:81.

McPherson, Miller, Lynn Smith-Lovin, and James M. Cook. 2001. "Birds of a Feather:
Homophily in Social Networks." Annual Review of Sociology 27:415-444.

Merton, R. K. 1949. Social Theory and Social Structure. New York: Free Press.

Milem, J. F. 1994. "College, students, and racial understanding." Thought and Action
9:5 1-92.

Milem, Jeffrey F., Paul D. Umbach, and Christopher T. H. Liang. 2004. "Exploring the
Perpetuation Hypothesis: The Role of Colleges and Universities in Desegregating
Society." Journal of College Student Development 45:688-700.

Moody, James. 2001. "Race, school integration, and friendship segregation in America."
The American Journal of Sociology 107:679.

Newsroom, Michigan State University. 2007, "MSU Facts", Retrieved April 15, 2007,
2007 (http://newsroom.rnsu.edu/sn_av/1 84L/pagehtm).

Nora, Amaury and Alberto F Cabrera. 1996. "The role of perceptions of prejudice and
discrimination on the adjustment of minority students to college." The Journal of
Higher Education 67:1 19.

O'Neill, James M. 2003. "Students, Officials Differ on Views of Diversity at
Pennsylvania Colleges." Knight Ridder Tribune Business News: 1 .

Orfield, Gary and Chungmei Lee. 2004. Brown at 50: King's Dream or Plessy's
Nightmare. Cambridge, MA: The Civil Rights Project at Harvard University.

Orfield, Gary and Dean Whitla. 2001. "Diversity and Legal Education: Student
Experiences in Leading Law Schools." Pp. 143-174 in Diversity Challenged:

287

Evidence on the Impact of Aflirmative Action, edited by G. Orfield and M.
Kurlaender. Cambridge, MA: Harvard Educational Review.

Park, Robert E. 1924. "The Concept of Social Distance As Applied to the Study of Racial
Relations." Journal of Applied Sociology 8:339-344.

Pascarella, E. and P. Terenzini. 1991. How college aflects students: Findings and insights
from twenty years of research. San Francisco: Jossey-Bass.

Pascarella, Ernest T., Patrick T. Terenzini, and Gregory S. Blimling. 1994. "The Impact
of Residential Life on Students." Pp. 22-52 in Realizing the Educational Potential
of Residence Halls, edited by C. C. Schoeder and P. Mable. San Francisco:
Jossey-Bass.

Pettigrew, T. 1998. "Intergroup contact theory." Annual Review of Psychology 49:65-85.

Pettigrew, T. and Linda R. Tropp. 2000. "Does Intergroup Contact Reduce Prejudice?
Recent Meta-Analytic Findings." Pp. 93-114 in Reducing Prejudice and

Discrimination, edited by S. Oskamp. Mahwah, NJ: Lawrence Erlbaum
Associates.

Pinsker, Beth. 1999. "The Quarrel: Yale Five Get Their Day in Federal Court." Pp. 4 in
Forward, vol. CII.

Pogrebin, Letty Cottin. 1987. Among Friends: Who We Like, Why We Like Them, and
What We Do With Them. New York: McGraw-Hill Book Company.

Porter, Stephen R. and Paul D. Umbach. 2006. "Student survey response rates across
institutions: Why do they vary?" Research in Higher Education 47:229-247.

Powers, Daniel A. and Christopher G. Ellison. 1995. "Interracial Contact and Black
Racial Attitudes: The Contact Hypothesis and Selectivity Bias." Social Forces
74:205-226.

Rios, Denice A. 1994. "Segregated Student Housing-By Choice." The Hispanic Outlook
in Higher Education 5:1.

288

 

Rubin, Lillian B. 1994. Families on the Fault Line: America's Working Class Speaks
about the Family, the Economy, Race, and Ethnicity. New York: Harper Collins.

Saenz, V.B., H.N. Ngai, and S. Hurtado. 2007. "Factors influencing positive interactions
across race for Afi'ican American, Asian American, Latino, and White Students."
Research in Higher Education 48: 1-38.

Safron, Deborah J. and Clifford L. Broman. 1997 . "Racial Differences in Perceptions of
Discrimination." National Journal of Sociology 1 1:61-79.

Sax, L. J., S. K. Gilmartin, and A. N. Bryant. 2003. "Assessing response rates and

nonresponsive bias in web and paper surveys." Research on Higher Education
44:409-432.

Schelling, Tomas C. 1971. "Dynamic Models of Segregation." Journal of Mathematical
Sociology 1:148-186.

—. 1972. "A Process of Residential Segregation: Neighborhood Tipping." in Racial
Discrimination in Economic Life, edited by A. H. Pascal and D. C. Heath.

Schuman, Howard and Charlotte Steeh. 1996. "The complexity of racial attitudes in
America." Pp. 455-469 in Origins and destinies : immigration, race, and ethnicity
in America, edited by S. Pedraza and R. G. Rumbaut. Belmont, CA: Wadsworth.

Sears, David O. et al. 1997. "Is it Really Racism? The Origins of White Americans'
Opposition to Race-Targeted Policies." Public Opinion Quarterly 61:16-53.

Shea, Christopher. 1993. "Does Student Housing Encourage Racial Separatism on
Campuses?" The Chronicle of Higher EducationzA26, A27.

Siegel, Adam J. 1997. "Is Special-Interest Housing The Answer? The Yale Controversy."
Pp. 33 in The New York Jewish Week, vol. 210.

Sigelman, Lee, Timothy Bledsoe, Susan Welch, and Michael W. Combs. 1996. "Making
contact? Black-White Social Interaction in an Urban Setting." American Journal
of Sociology 101 :1306-32.

289

 

Slavin, R. E. 1995. "Cooperative learning groups and intergroup relations." in Handbook
of research on multicultural education, edited by J. A. Banks and C. A. M. Banks.
New York: MacMillan.

Slavin, R. E. and R. Cooper. 1997. "Improving Intergroup Relations: Lessons Learned
from Cooperative Learning Programs." Journal of Social Issues 55:647-663.

Smith, T. W. 2002. "Measuring inter-racial friendships." Social Science Research
31:576-593.

Staples, Brent. 1993. "Apartheid on campus, continued." Pp. 410 in New York Times.

—. 1996. "Jim Crow at Cornell?" Pp. 4.14 in New York Times.

Stephan, Cookie White. 1992. "Mixed-Heritage Individuals: Ethnic Identity and Trait
Characteristics." in Racially Mixed People in America, edited by M. P. P. Root.
Newbury Park: Sage.

Stepick, Alex. 1997. Pride Against Prejudice: Haitians in the United States, Edited by N.
Foner. MA: Allyn & Bacon.

Suarez-Balcazar, Yolanda, Lucia Orellana-Damacela, Nelson Portillo, Jean M Rowan,
and Chelsea Andrews-Guillen. 2003. "Experiences of Differential Treatment
Among College Students of Color." The Journal of Higher Education 74:428.

Tatum, Beverly Daniel. 1997. "Why Are All the Black Kids Sitting Together in the
Cafeteria? " And Other Conversations About Race. New York: Perseus Books
Group.

Tyson, Will. 2004. "Roommate and residence hall composition effects on interracial
friendships among first-year college students." Dissertation Thesis, Sociology,
Duke University.

US Census Bureau, Census Summary File. 2002, "Census 2000. PHC-T-16. Ranking
Tables for Places of 100,000 or More Population by Race Alone, Race Alone or

in Combination, Two or More Races, and Hispanic or Latino Population",
Retrieved (http://www.censusgov/population/cen2000/phc-t16/tab30;pdf).

290

 

Van Kenhove, P., K. Wijnen, and K. De Wulf. 2002. "The influence of topic involvement
on mail-survey response behavior." Psychology & Marketing 19:293-301.

Van Laar, C., S. Levin, S. Sinclair, and J. Sidanius. 2005. "The effect of university
roommate contact on ethnic attitudes and behavior." Journal of Experimental
Social Psychology 41 :329-345.

Waters, Mary C. 1999. Black Identities: West Indian Immigrant Dreams and American
Realities. New York: Russel Sage Foundation.

Wolfe, C. T. and S. J. Spencer. 1996. "Stereotypes and prejudice: Their overt and subtle
influence in the classroom." American Behavioral Scientist 40:176-185.

Zipf, G. K. 1949. Human Behavior and the Principle of Least Effort. Menlo Park, CA:
Addison-Wesley.

291