Content uploaded by Nathan Martin
Author content
All content in this area was uploaded by Nathan Martin on May 31, 2015
Content may be subject to copyright.
,QWHUUDFLDO)ULHQGVKLSV$FURVVWKH&ROOHJH<HDUV(YLGHQFH
IURPD/RQJLWXGLQDO&DVH6WXG\
1DWKDQ'0DUWLQ:LOOLDP7RELQ.HQQHWK,6SHQQHU
Journal of College Student Development, Volume 55, Number 7, October
2014, pp. 720-725 (Article)
3XEOLVKHGE\7KH-RKQV+RSNLQV8QLYHUVLW\3UHVV
DOI: 10.1353/csd.2014.0075
For additional information about this article
Access provided by Arizona State University (31 May 2015 18:29 GMT)
http://muse.jhu.edu/journals/csd/summary/v055/55.7.martin.html
720 Journal of College Student Development
Research in Brief
Interracial Friendships Across the College Years:
Evidence from a Longitudinal Case Study
Nathan D. Martin William Tobin Kenneth I. Spenner
Today, the student bodies of our leading colleges
and universities are more diverse than ever.
However, college students are increasingly self-
segregating by race or ethnicity (Saenz, Ngai,
& Hurtado, 2007). A burgeoning literature
documents the benefits of campus diversity and
shows that having friends of a different race
predicts greater acceptance and awareness of
other groups as well as higher levels of academic
self-confidence and learning outcomes (e.g.,
Antonio, 2004; Hu & Kuh, 2003). For many
young adults, the college years serve as the first
opportunity to interact with a large number of
peers from different backgrounds. Yet, in order
to fully realize the benefits of structural diversity
on campus, it is important to understand
how interracial friendships are formed and
maintained across the college years.
In this study we explored factors that
influence the degree to which students’
campus friends are of a different race or
ethnicity. We focused on relationships that
are more sustained and involve greater trust
than routine interactions on campus (e.g.,
Chang, Denson, Sáenz, & Misa, 2006), but
are less intimate than best or closest friends
(e.g., Antonio, 2004). Recent studies show
that high school experiences are associated
with having friendships that cross racial-ethnic
boundaries in college (Fischer, 2008), and that
having a different-race roommate can provide
opportunities to form interracial friendships
(Camargo, Stinebrickner, & Stinebrickner,
2010; Stearns, Buchmann, & Bonneau, 2009).
Additionally, Stearns and colleagues (2009)
found that fraternity or sorority membership
is associated with fewer interracial friendships
for White students. These studies make a
valuable contribution to our understanding
of interracial friendship formation, although
the existing literature has been limited by
focusing attention on the early college years or
by considering White and Black students only.
Using survey data from the Campus Life
& Learning Project—a prospective panel study
of students at a selective, private university in
the Southeastern United States—we extend
the existing literature by considering the
experiences of White, Black, Latino, and Asian
students, and by examining a broad range of
college activities. Our results indicate that
students’ friendships become less diverse from
the first to the fourth years, and that aspects of
the residential environment, interactions with
faculty, types of extracurricular participation,
and the presence of alcohol at social events
are significant predictors of having interracial
campus friendships.
DATA AND METHODS
e Campus Life & Learning Project (CLL)
followed the incoming classes of 2001 and
2002 with four survey waves administered in
Nathan D. Martin is Assistant Professor of Justice & Social Inquiry at Arizona State University. William Tobin is a
Research Scholar in the Kenan Institute for Ethics at Duke University. Kenneth I. Spenner is Professor of Sociology at
Duke University. e Campus Life and Learning data were collected by A. Y. Bryant, Claudia Buchmann, and Kenneth
I. Spenner (Principal Investigators), with support provided by the Andrew W. Mellon Foundation and Duke University.
O
◆
721
Research in Brief
the summer prior to matriculation and in the
spring of the first, second, and fourth college
years. Importantly, the CLL was not intended
to be representative of all postsecondary
institutions, but it can be considered as
characteristic of other elite private and many
highly rated public universities in terms of
admissions rate and yield, cost of attendance,
student–faculty ratio, SAT scores of incoming
students, and student retention rate.
Among all incoming students in the two
CLL cohorts (N = 3,254), about 60% were
White, 11% were Black, 8% were Latino, 15%
were Asian, and 7% were multiracial or some
other race. e CLL design randomly selected
one third of White students, two thirds of
Asian students, and one third of multiracial
students in each cohort, as well as all Black
and Latino students. e full sample included
1,536 students, and 79% of sample members
completed the precollege survey. Of these
precollege respondents, 77% also responded to
the first-year survey, 75% to the second-year
survey, and 67% to the fourth-year survey.
For this study, the analytic sample (n = 996)
is restricted to precollege respondents who
completed at least one in-college survey and
excludes multiracial students.
e dependent variable in the analysis
to follow is from an item collected in each
in-college wave that asked students: “What
best describes your friends on campus?”; with
response choices: all or nearly all your race,
mostly your race, half your race and half not your
race, mostly not your race, and all or nearly all
not your race (see Table 1). In general, Latino
TABLE 1.
Responses (% n) to “What best describes your friends on campus?”
Group n
All or
Nearly All
Your Race
Mostly
Your Race
Half Your
Race and
Half Not
Your Race
Mostly Not
Your Race
All or
Nearly All
Not
Your Race
White
First Year 393 17.3 34.9 20.6 14.8 12.5
Second Year 378 34.1 43.9 15.3 3.2 3.4
Fourth Year 339 26.8 48.1 14.8 3.5 6.8
Black
First Year 162 14.8 21.0 22.2 21.0 21.0
Second Year 146 32.4 31.1 16.9 12.2 7.4
Fourth Year 139 37.4 28.1 14.4 10.8 9.4
Latino
First Year 144 9.0 12.5 31.3 31.3 16.0
Second Year 142 9.9 23.2 21.1 20.4 25.4
Fourth Year 122 17.2 18.0 18.9 22.1 23.8
Asian
First Year 161 4.4 13.7 29.2 30.4 22.4
Second Year 148 10.1 21.6 25.0 25.0 18.2
Fourth Year 131 14.5 17.6 25.2 26.7 16.0
Note. Percentages may not sum to 100% due to rounding.
722 Journal of College Student Development
Research in Brief
and Asian students were more likely to have
friends from a different race than were White
or Black students. Across all groups, campus
friendship networks become less diverse as
students continued along their college career
and especially between the first and second
years. For example, the percent of White and
Black students reporting mostly same-race
friends increases between the first and second
years from 52% to 78% and from 36% to
64%, respectively.
One time-invariant independent variable
collected from the precollege survey describes
the percent of students’ closest friends from a
different race or ethnicity. Other variables were
collected in each in-college survey, including:
opportunities to interact with students from
different backgrounds in the campus residence
(responses from never to always), having
a roommate of a different race (yes or no),
hours per week spent interacting with faculty
outside of class, reports of being treated by
instructors more as a group representative than
as an individual (from never to always), and
membership in various extracurricular activities
(member or nonmember). Additionally, students
were asked to describe how important alcohol
was to their enjoyment of campus life (from not
at all important to extremely important), and how
often alcohol was present at the campus social
events they attended (from never to always).
With multiple observations for each
student, our strategy is to treat within-unit
variation as a random effect or disturbance
to account for unmeasured characteristics
that may influence the propensity to form
interracial friendships. Given the unbalanced
panel and measurement of the dependent
variable, we estimated generalized linear
latent and mixed models with an ordinal
logit link function. We addressed longitudinal
dependence across observations by including
a student-specific random intercept in the
proportional odds model:
logit{Pr(y
ij
> s | X
ij
, ζ
1j
)} = βX
ij
+ ζ
1j
–
κ
s
In this equation, the log odds of the ith student
in the jth year being in a higher response
category of the outcome variable (y) than s is
a function of measured student characteristics
(βX
ij
) and an overall intercept (ζ
1j
) that varies
across students. All analyses were conducted
with Stata (version 12.0).
RESULTS
Table 2 presents models of the effects of selected
college contacts and activities on interracial
campus friendships. Due to different structural
opportunities to interact with students from
different backgrounds by racial/ethnic group,
our strategy was to conduct analysis separately
for White, Black, Latino, and Asian students.
An indicator for year captures how friendship
networks change as students continue along
the college career. For White, Black, and
Asian students campus friendships become
significantly less diverse after the first year.
More precisely, the odds of reporting a higher
category of the dependent variable (e.g., all or
nearly all not your race versus mostly not your
race) decline from the first to the second year
by 12% for White students and 27% for Black
and Asian students.
For White, Black, and Asian students, the
proportion of precollege friends of a different
race or ethnicity has a positive effect on
interracial campus friendships. Additionally,
an alternative specification that includes a
dummy variable for living in a segregated
neighborhood (i.e., nearly all non-White)
during high school yields a negative coefficient
for Black students (OR = .59; z = –2.03).
ese results suggest that the propensity to
engage in interracial friendships while in
college can be encouraged by opportunities
during the high school years, and that the
qualities that enable a student to cooperate
O
◆
723
Research in Brief
TABLE 2.
Generalized Linear Latent and Mixed Ordinal Logit Models Predicting
Interracial Campus Friendships
Model 1
White
Model 2
Black
Model 3
Latino
Model 4
Asian
College Year 0.88 * 0.73 *** 0.97 0.73 ***
(–2.14) (–3.74) (–0.27) (–3.73)
High School Years
Percent of closest friends from a
different racial/ethnic background
1.01 * 1.02 *** 0.98 *** 1.02 ***
(1.98) (5.47) (–3.23) (3.92)
College Contacts and Activities
Opportunities in college residence to
interact with students from different
backgrounds
1.42 *** 1.22 * 1.45 *** 1.28 *
(5.75) (2.23) (3.46) (2.34)
Any roommate from a different racial/
ethnic background
2.50 *** 2.27 *** 0.85 2.48 ***
(5.62) (3.66) (–0.38) (3.72)
No roommate 1.01 2.85 *** 0.92 3.57 ***
(0.06) (3.62) (–0.16) (3.74)
Hours per week spent interacting or
meeting with faculty outside of class
1.07 * 1.02 1.18 *** 1.09
(2.19) (0.55) (3.30) (1.50)
Viewed stereotypically by instructors in
the past year
0.98 0.82 * 0.91 1.01
(–0.32) (–2.17) (–0.75) (0.11)
Fraternity/sorority membership 0.61 *** 0.80 0.43 ** 2.20 *
(–3.43) (–0.76) (–2.99) (2.34)
Cultural or ethnic club membership 1.42 0.46 *** 0.96 0.45 ***
(1.20) (–3.89) (–0.12) (–3.28)
Community service club membership 1.36 * 0.85 0.78 1.13
(1.98) (–0.78) (–0.91) (0.47)
Sports team membership 1.18 1.62 * 1.05 2.02 **
(1.20) (2.16) (0.19) (2.58)
Other club or group membership 0.98 0.67 * 1.92 * 1.27
(–0.14) (–2.01) (2.40) (0.95)
Presence of alcohol at campus social
events student attends
0.86 ** 1.04 0.92 1.18 **
(–2.81) (0.43) (–0.97) (2.06)
Importance of alcohol to student’s
enjoyment of college life
0.82 *** 0.92 1.11 1.35 **
(–3.38) (–0.88) (0.96) (2.64)
Observations 1,110 449 408 440
Students 454 190 173 179
Note. Displaying odds–ratios (z scores in parentheses).
* p < .05, two-tailed. ** p < .01. *** p < .001.
724 Journal of College Student Development
Research in Brief
and compete in an inclusive setting may be
considered as a skill or capacity that is acquired
experientially and developed over time.
Features of the residential environment
are relatively strong predictors of interracial
friendships for students of all racial/ethnic
groups. For White, Black, and Asian students,
having a different-race roommate is asso-
ciated with having more interracial campus
friendships. Additionally, for Black and Asian
students, not having a roommate is associated
with more diverse friendship networks;
a plausible interpretation of this effect is
that students without roommates are more
likely to seek out friends in their residence
hall and thus come into contact with more
students of different backgrounds (Stearns
et al., 2009, p. 190).
Previous studies have noted the many
benefits of faculty interactions on college
student development. To add to this literature,
we find that interacting with faculty outside of
class has a positive effect on friendship diversity
for White and Latino students. Conversely,
when Black students perceive that instructors
treat them stereotypically, they are less likely
to have interracial friendships. Favorable
interactions with faculty can encourage
students’ personal growth and reaffirm their
shared status as a member of the campus
community. As a result, students may place less
emphasis on peers’ racial/ethnic background
when building social relationships.
Students at this university were highly
involved in campus life, although patterns of
extracurricular membership varied considerably
by race and ethnicity. In the first year, 43%
of White and 42% of Latino students were
members of a fraternity or sorority, compared
to 12% of Asian and 6% of Black students.
About 60% of Black, 50% of Asian and
36% of Latino students were members of a
cultural or ethnic club, compared to 3% of
White students. Activities that emphasize
common goals are associated with having
more interracial campus friendships, namely
community service clubs for White students
and intramural or varsity sports teams for Black
and Asian students. However, other activities
tend to encourage same-race friendships:
fraternity or sorority membership had a
negative effect on interracial friendships for
White and Latino students, but a positive
effect for Asian students (who comprise
only 6% of fraternity or sorority members).
Similarly, membership in a cultural or ethnic
club was associated with fewer interracial
friendships for Black and Asian students.
For many students, alcohol featured
pro minently in their relationship to campus
social life. We find that the routine presence
of alcohol at campus social events can serve as
an obstacle to forming interracial friendships.
In line with patterns for fraternity or sorority
membership, during the first year 77% of
White and 88% of Latino students reported
that alcohol was often present at the social
events they attended, compared to 39% of
Black and Asian students. Additionally, 24%
of White and 28% of Latino students reported
that alcohol was important to their enjoyment
of campus life, compared to 7% of Black and
Asian students. A more prominent role of
alcohol to campus life is associated with fewer
interracial friendships for White students but
more interracial friendships for Asian students.
DISCUSSION
e students in this study were part of a diverse
campus community, yet more than three
fifths of students reported that their campus
friends were mostly from their same race by
the fourth year. e trend towards increasingly
segregated social networks is discouraging,
given the benefits of interracial friendships and
interactions for a range of student development
outcomes. At the same time, it is important
O
◆
725
Research in Brief
to note that active involvement in cultural
or ethnic clubs can play an important role in
the development of a student’s racial/ethnic
identity (Cross & Vandiver, 2001). Further,
marginalized student populations may rely on
same-race friendships as a coping strategy in
response to experiences of discrimination and
perceptions of a hostile environment (Beasley,
2011). Our finding that Black students who
reported stereotypical treatment by instructors
had fewer interracial friendships is consistent
with this explanation.
What can be done to address the decline
in friendship diversity across the college
years? Our results suggest several policies that
could promote interracial friendships and
interactions on campus. Among the most
feasible would be to ensure that residence halls
contain substantial diversity across buildings
and floors. Instituting random roommate
assignment after the first year would likely
result in more frequent interracial interactions
in the residential environment. Additionally,
the university could provide more support for
students to meet with faculty outside of class
and to participate in activities that emphasize
common goals, such as community service
organizations and sports teams. Currently,
the most popular extracurricular activities
on campus reinforce homophilous social
networks. Finally, given the significant racial/
ethnic differences in the role of alcohol to
campus life, curbing campus drinking overall
or supporting alternative social events that
do not include alcohol could result in more
diverse friendships on campus.
Admissions committees at selective uni ver-
sities consider a range of characteristics beyond
academic merit, including under represented
minority status, athletic ability, and alumni
connections. To fully realize the benefits of a
diverse student body may require extending
preferences to students who demonstrate
a commitment to diversity as well. There
are a number of advantages to considering
students’ propensity to seek out interracial
friendships in admissions. An ability to work
cooperatively in an inclusive setting is a skill
that is directly transferable to the college years
and beyond. Further, providing a signal to
students, parents, and guidance counselors that
a commitment to diversity is a valued resource
could provide incentives for voluntary, creative
ways to promote integration at the primary
and secondary school levels.
Correspondence concerning this article should be addressed
to Nathan Martin, School of Social Transformation,
Arizona State University, PO Box 876403, Tempe, AZ
85287; ndm@asu.edu
REFERENCES
Antonio, A. L. (2004). e influence of friendship groups on
intellectual self-confidence and educational aspirations in
college. Journal of Higher Education, 75, 446-471.
Beasley, M. A. (2011). Opting out: Losing the potential of America’s
young Black elite. Chicago, IL: University of Chicago Press.
Camargo, B., Stinebrickner, R., & Stinebrickner, T. (2010).
Interracial friendships in college. Journal of Labor Economics,
28, 861-892.
Chang, M. J., Denson, N., Sáenz, V., & Misa, K. (2006). e
educational benefits of sustaining cross-racial interaction among
undergraduates. Journal of Higher Education, 77, 430-355.
Cross, W. E., Jr., & Vandiver, B. J. (2001). Nigrescence theory
and measurement. In J. G. Ponterotto et al. (Eds.), Handbook
of multicultural counseling (pp. 371393). ousand Oaks,
CA: SAGE.
Fischer, M. J. (2008). Does campus diversity promote friendship
diversity? A look at interracial friendships in college. Social
Science Quarterly, 89, 631-655.
Hu, S., & Kuh, G. D. (2003). Diversity experiences and college
student learning and personal development. Journal of College
Student Development, 44, 320-334.
Saenz, V. B., Ngai, H. N., & Hurtado, S. (2007). Factors
influencing positive interactions across race for African
American, Asian American, Latino, and White college
students. Research in Higher Education, 48, 1-38.
Stearns, E., Buchmann, C., & Bonneau, K. (2009). Interracial
friendships in the transition to college: Do birds of a feather
flock together once they leave the nest? Sociology of Education,
82, 173-195.
