Journal of Diversity in Higher Education
2008, Vol. 1, No. 2, 95–107
Copyright 2008 by the National Association of Diversity Officers in Higher Education
1938-8926/08/$12.00 DOI: 10.1037/1938-8926.1.2.95
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Roles of Gender, Race, and SES in the College Choice Process
Among First-Generation and Nonfirst-Generation Students
Su-Je Cho
Cynthia Hudley
Fordham University
University of California, Santa Barbara
Soyoung Lee
Leasha Barry
City College of New York
University of West Florida
Melissa Kelly
Millsaps College
The authors examined students’ reports of their college choice process to understand
the influence of a set of psychological, personal, and institutional factors. The authors
also examined potential moderating influences of generational status, gender, race, and
SES on our variables of interest. A diverse sample of college freshmen (N ⫽ 1,339),
including 42% who were the first in their families to attend college, responded to a
self-reporting, Web-based survey. Findings indicate that psychosocial factors and
academic quality of the college were most influential for first-generation students as
compared to their nonfirst-generation peers in the college choice process. However,
gender, race, and SES moderated these influences in complex ways. For example,
females rated the psychological variables higher than males; Asian American and
African American first-generation students rated higher than their parents’ preferences
for which college to attend as compared to nonfirst-generation peers. First-generation
females, African American in particular, considered academic quality more important
than other groups. Our findings should be of value to counselors and other personnel
who facilitate students’ college choice process as well as college recruitment, retention,
and diversity enhancement programs.
Keywords: college choice, psychological factors, academic factors, personal factors, diversity
Knowledge about the college choice process
among high school students has grown substantially over the past decade, but knowledge about
college choice among students who are the first
in their families to attend college continues to
lag. We examine a number of institutional and
psychosocial factors that the literature suggests
are important for students’ satisfaction with
campus “fit” to determine if they are equally
significant for first-generation students and
across other relatively understudied groups. Unlike much of the earlier literature on firstgeneration students that focuses on specific ethnic minority student groups (e.g., Latinos, African Americans) or on underrepresented
students as a group, our research is broadly
focused on students of all ethnicities who are
now college freshmen. Understanding the
choice process has substantial personal implications, because a student’s satisfaction with his
or her choice of college plays a role in any
decision to leave that school or remain and
complete a college degree (Tinto, 1993). This
finding seems especially important for understanding the persistence of first-generation students, as college dropout rates among firstgeneration students are considerably higher
than among their nonfirst-generation peers. Approximately 25% of students drop out of college
Su-Je Cho, PhD, Graduate School of Education, Fordham
University; Cynthia Hudley, PhD, Gevirtz Graduate School
of Education, University of California, Santa Barbara; Soyoung Lee, PhD, Department of Childhood Education, City
College of New York; Leasha Barry, PhD, Division of
Teacher Education, University of West Florida; and Melissa
Kelly, PhD, Department of Psychology, Millsaps College.
Correspondence concerning this article should be addressed to Su-Je Cho, PhD, Graduate School of Education,
Fordham University, 113 West 60th Street, 1102, New
York, NY 10023. E-mail: [email protected]
95
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
96
CHO ET AL.
by the end of their first year, but the attrition
rate approaches 50% for first-generation students (Ishitani, 2003; NCES, 2001; National
Center for Education Statistics).
The college choice process has far-reaching
institutional consequences as well. The impact
of students’ college choice and college attrition
rates has become increasingly important for institutional planning in this era of scarce resources. Enrollment growth generally is not uniform for colleges and universities across the
nation (Kinzie, Palmer, Hayek, Hossler, Jacob,
& Cummings, 2004). Although colleges in
some states (e.g., Arizona and Florida) can look
forward to ample in-state enrollments for the
foreseeable future, institutions in other states
(e.g., West Virginia and Louisiana) will struggle to find state students to meet their enrollment goals over the next decade (Kinzie et al.,
2004). Consequently, schools in some areas
may have to more aggressively reach out to
recruit underrepresented students, including
those who are the first in their families to attend
college, as well as students who may attend
from out of state. However, aggressive recruitment efforts must be complemented by active
retention programs that maximize the “fit,” satisfaction, and ultimately the achievement of students, if colleges are to appropriately serve the
students that they have successfully recruited.
In developing such programs, staff and faculty
may have to develop knowledge and expertise
about student groups with whom they have had
limited prior contact and relationships. Thus, as a
starting point, a better understanding of the college
choice process among first-generation students,
and indeed among all students, has important implications for campus recruitment and retention
programs, programs tailored toward enhancing
and maintaining campus diversity, and policy decisions governing postsecondary institutions.
First in the Family
Although the literature presents a variety of
definitions, in our research we have relatively
restrictively defined first-generation students as
those who are the first in their immediate family
(i.e., parents, grandparents, and siblings) to attend a 4-yr college or university. Recent research on first-generation students has investigated minority high school students who are
planning for the transition to college (e.g., Stan-
ton-Salazar, 2001), first-generation students
who are now successful adults (e.g., Roberts &
Rosenwald, 2001), and comparisons of firstgeneration students and their nonfirst-generation peers who are enrolled in college (e.g.,
Massey, Charles, Lundy, & Fischer, 2003; Pascarella, Pierson, Wolniak, & Terenzini, 2004;
Pike & Kuh, 2005). This body of research consistently points out that first-generation college
students are more likely to be from a low SES
family, ethnic minority, and to speak a language
other than English at home (Bui, 2002; Terenzini, Springer, Pascarella, & Nora, 1995), relative to their nonfirst-generation peers. Students
who are the first in their families to attend
college are also less knowledgeable about college costs and application processes (YorkAnderson & Bowman, 1992), and they tend to
have a less rigorous academic preparation in
high school (Warburton, Bugarin, & Nunez,
2001). This lack of knowledge and preparation
may influence students’ decision making when
they are trying to determine which college will
be appropriate for their needs.
Variables that Influence College Choice
In early work, Holland and Richards (1965)
found that high school students’ college choices
were heavily influenced by four major factors:
(1) intellectual emphasis of the prospective
campus, (2) practical concerns (e.g., closeness
to home, low cost), (3) advice of others (e.g.,
high school counselors, parental input), and (4)
perceived social climate of the campus. Although these variables continue to be significant
influences on the college choice process of high
school students today (Kinzie et al., 2004),
more recent research has revealed that these
variables predict differentially for students of
varying characteristics. For example, students
of high academic ability are more likely to
attend selective institutions or out-of-state institutions than students of lower ability (Braxton,
1990), but this is less consistently true for lowincome, rural, and female students. These students, regardless of their academic ability,
achievement, and expectations, are more likely
to choose less selective institutions and demonstrate greater sensitivity to institutional factors
such as size and location (Kim, 2004) as well as
individual perceptions such as perceived comfort on campus (Nora, 2004). Although aca-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
GENDER, RACE, AND SES IN THE COLLEGE CHOICE PROCESS
demic achievement in high school remains the
most important determinant of whether and
where students go to college, institutional characteristics and psychosocial factors are also significant influences in the college choice process.
Financial aid availability is the institutional
characteristic that has arguably been the most
extensively examined in the literature as a determinant of college choice. Yet, extant findings
remain inconclusive. A number of studies
(Hossler, Braxton & Coopersmith, 1989;
Hossler & Gallagher, 1987; Hossler, Schmit, &
Vesper, 1999; McDonough, 1997; McDonough
& Antonio, 1996) suggest that financial aid is a
critical factor for all students. Other work points
specifically to the importance of financial aid in
the college choice process for ethnic minority
students, many of whom are first-generation
college students (Kern, 2000). Conversely, Hurtado and Carter (1997) found that financial aid
had no bearing on students’ college choice. Kim
(2004) reported that African American and
Latino students, including first-generation students, are less influenced by the availability of
financial aid during their college choice process.
Rather, location of the college and the size of
the student population appear to be factors that
heavily influence their college choices.
These inconsistencies suggest that although
the availability of financial aid may have bearing on college choice, the process involves a
complex weighting and ranking of multiple individual preferences and institutional characteristics. However, the specific influence of psychosocial factors (e.g., perceived social relations on campus) remains perhaps the least
understood area of the college choice process.
Further, the college choice process and its
poorly understood balancing of institutional
factors and individual preferences and needs is
almost entirely opaque among first-generation
students who attend 4-yr institutions. Thus, our
research represents an effort to examine the
intersection of institutional, personal, and psychosocial variables that influence college choice
and how those variables may be moderated by
first-generation status.
Psychosocial Variables and College
Choice
Our study takes as a starting point Nora’s
(2004) three-stage model of college selection:
97
predisposition, search, and choice. We focus
explicitly on the final stage, the choice itself, in
which psychosocial, institutional, and personal
preference (e.g., location of the campus) factors
converge to produce a student’s decision to
apply and actually matriculate at a given institution. Thus, our interest here is in the perceptions of students who have completed their selections and actually enrolled in a college of
their choice. A relatively robust literature exists
on the influence of institutional and personal
variables on choice (Hearn, 1991; Hossler,
Braxton, & Coppersmith, 1989; Stage &
Hossler, 1988), finding that a perceived fit between personal and institutional characteristics
predicts more satisfactory college adjustment
and retention (Antonoff & Friedman, 1991;
Tinto, 1993). However, more recent research
suggests that psychosocial factors may be
equally, if not more, important in the choice
process (McDonough, Antonio, & Trent, 1997).
Work on the influence of psychosocial variables on college choice is less well developed
than in other (e.g., financial aid) domains (Nora,
2004). However, this literature suggests that
student perceptions of the campus social environment and their understanding of their own
psychological needs (e.g., values, interests) will
be fruitful areas of study to extend our knowledge of the college choice process, particularly
among first-generation students. The role of
psychosocial variables is captured by the construct habitus, or the individual’s internalized
representation of a given social environment
that guides decision-making and action in that
environment (Horvat, 2003). Habitus is integral
to the student’s perception of a fit between
psychosocial needs and the capacity of a campus environment to meet those needs (Nora,
2004) and is consistent with a goodness-of-fit
model from the developmental literature (Chess
& Thomas, 1991; Hudley, 1995). A goodnessof-fit model specifies individual adjustment as a
function of the match between individual characteristics and contextual demands. When a person’s behavior, interests, and motivation match
the perceived demand characteristics and benefits of an environment, adjustment in that environment should be optimal and levels of stress
minimal (French, Rogers, & Cogg, 1974; Lerner, Baker, & Lerner, 1985). Thus, understanding the intersection of psychosocial, institutional, and personal variables in the college
98
CHO ET AL.
choice process for first-generation students
should yield insights of theoretical and practical
significance for programs of recruitment and
retention, and ultimately for student success.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
The Current Study
This exploration of college choice addressed
two basic questions. (1) What psychosocial, institutional, and personal factors are most important to students in choosing to attend a particular
college? (2) How do these factors operate differentially across generational status, gender,
ethnic, and SES? We answered these questions
by testing several hypotheses that explored the
conjoint influences of psychosocial, institutional, and personal variables and examined
these variables at the intersection of generational status, ethnicity, gender, and social class.
We expected first-generation students as a
group to rate psychosocial and personal indicators of “fit” as more important in their college
choices than their nonfirst-generation peers
based on their underrepresented status on campus. However, we anticipated that these relationships would be moderated by race/ethnicity,
family income, and gender. We also hypothesized that participants would not differ by generation status in their ratings of financial considerations, and consistent with prior literature,
participants would not differ in rating an institution’s academic programs and reputation as
very important considerations in their choice of
college to attend.
Method
Participants
Participants were 1,539 incoming freshman
students attending four different universities,
public and private, located in both urban and
rural areas (i.e., one each public urban, private
urban, public rural, private rural) in the United
States. A total of 6,560 students were invited via
email to participate in a Web-based survey. Our
resultant sample of 1,539 represents a 25% response rate. Based on preliminary descriptive
analyses, we eliminated 200 respondents from
our sample. The sample size for Native Americans was too small to analyze meaningfully
(N ⫽ 5), and those who did not respond to the
item that requested first/nonfirst-generation sta-
tus were not included in these analyses (N ⫽
124). Respondents who identified as ethnically/
racially mixed (N ⫽ 71) were also not included
because we are not certain they represent a
coherent group. Students in this category reported membership in a number of different
ethnic and racial groups (e.g., Black/White;
White/Asian; Native American/Latino/White)
and thus may or may not have experiences that
generalize across the entire group.
Findings reported here are based on a final
sample of 1,339 students. Mean age for our
sample was 18.37 (SD ⫽ 1.41), and the gender
balance was 74% female and 26% male. Although females were somewhat overrepresented, our sample is consistent with the troubling trend of declining enrollment of males in
higher education, who represented 43% of the
postsecondary enrollment in 2005 (NCES,
2006). The racial/ethnic distribution of the sample was also relatively consistent with the national estimate of student enrollment (NCES,
2005). European American/White students
made up 68% of the sample and Latino/
Hispanic represented 13.7%. However, compared to the national estimate, Asian/Pacific Islander students were overrepresented in our
sample at 12.3%, and African American/Black
students underrepresented at 6%. The selfreported family income score was 2.70
(SD ⫽ 1.30) on a scale that ranged from 1 “less
than $34,999” to 5 “more than $200,000.” Finally, our sample contained 42% first-generation students (n ⫽ 556).
Measure
Participants responded to a self-report, Webbased survey comprising 39 substantive questions. Additional demographic questions included age, gender, racial group membership,
work status, first-generation status, languages
spoken, and family income. The 39 substantive
questions asked about students’ experiences in
the following three areas: (1) factors that affected participants’ college choices, (2) participants’ high school experiences, and (3) participants’ college experiences. The analyses undertaken for this study were explicitly concerned
with data on college choice as a function of
first-generation status and demographic variables. Questions pertaining to college choice
included: college rankings, college demo-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
GENDER, RACE, AND SES IN THE COLLEGE CHOICE PROCESS
graphic characteristics, financial considerations,
personal preferences, and perceptions of the
campus and community environment. Response
were recorded on a four-point scale ranging
from 1 “extremely unimportant” to 4 “extremely important.”
The individual questions on college choice
were either combined into three scales or used
individually for subsequent analyses. A psychosocial scale (␣ ⫽ .60) included three questions:
(1) perceived security and safety on campus, (2)
friendship with students or alumni, and (3) positive perceptions of social activities. An additional scale comprising psychosocial variables
scale explicitly examined the importance of racial/ethnic climate, given that first-generation
students tend to be disproportionately from underrepresented groups. This racial/ethnic scale
(␣ ⫽ .91) comprised two questions: (1) enough
members of my racial/ethnic group on campus
to feel comfortable and (2) enough members of
my racial/ethnic group in surrounding community. A scale measuring academic quality (␣ ⫽
.80) included five questions: (1) graduates get
into graduate/professional schools; (2) this
school has a reputation for strong academics;
(3) this school has academic support programs;
(4) this school has unique educational program;
and (5) graduates get good jobs. Finally, we
included three individual items tapping personal
family preference (my parents wanted me to
come here) and two items tapping economic
concerns (this college is affordable; financial
aid is available). A complete listing of survey
questions is available from the second author.
Procedure
Survey questions were posted on a secure,
password-protected site maintained by the information technology staff at the second author’s university. E-mail invitations were sent to
all incoming freshmen students in the first
month of the 2004 through 2005 academic year
on each of the four campuses. E-mail addresses
were obtained through cooperation with the institutional research offices at the respective institutions. Each invitation contained a unique
password for that particular student to use to
access the Web site, complete the informed
consent form, and complete the survey. Students were able to access the site for 6 weeks
after the initial invitation, and two follow up
99
Table 1
Participants’ Age, Gender, Family Income, and
Race/Ethnicity
First
generation
status
(n ⫽ 556)
Variable
Age: M (SD)
Gender
Female
Male
Family income: M (SD)
Race/ethnicity
Asian/Pacific islander
Black/African-American
Latino/Hispanic
White/European-American
18.31 (.91)
Nonfirst
generation
status
(n ⫽ 859)
18.41 (1.66)
73%
27%
2.14 (1.19)
74%
26%
3.08 (1.23)
15%
8%
24%
53%
10%
5%
7%
78%
reminders were sent to all freshmen students.
The survey took 20 to 30 minutes to complete in
its entirety, and participants were required to
complete the survey in a single sitting. Participants who completed the survey were entered in
a raffle for a $50 gift certificate at each of the
participating universities.
We analyzed our data to explore the conjoint
influences of psychosocial, institutional, and
personal variables with an omnibus multivariate
analyses (MANCOVA), and follow-up univariate analyses. Our three scales (psychological,
racial/ethnic, and academic quality) and three
individual variables (personal and economic
concerns) were the jointly dependent variables,
generational status, race and gender served as
our grouping variables, and family income was
a coviariate. We used family income as a covariate because of the substantial imbalance in
the highest and lowest categories of income. We
report the results of multivariate analyses using
the Roy’s largest root test criterion, because
Roy’s criterion has been found to be most consistent with post hoc tests to evaluate means
(Lutz, 2000; Gabriel, 1968).
Results
Before testing our hypotheses, we conducted
preliminary analyses to identify any significant
demographic differences between our firstgeneration students and their nonfirst-generation peers (see Table 1). T test results showed
no age difference between the first- and nonfirst-generation students. Chi-square tests re-
.07
1.94
1.98 (.80)
1.83 (.71)
2.45 (.99)
1.98 (.87)
2.06 (.85)
p ⬍ .01.
2.43 (.90)
2.14 (.82)
2.26 (.81)
*
3.06
2.25 (.86)
2.15 (.83)
2.44 (.84)
2.38 (1.00)
2.27 (.87)
2.33 (.94)
2.51 (.86)
2.52 (.97)
Our expectation that ratings of psychosocial
and personal factors would differ but differences
would be moderated by race/ethnicity, family income, and gender was supported. Our omnibus
test revealed an interaction of first-generation status and gender, F(2, 1191) ⫽ 3.57, p ⬍ .01 that
was partially shown in differences on the psychosocial scale, F(2, 1195) ⫽ 2.76, p ⫽ .06. Males
differed more strongly by first-generation status
than females, who rated this scale higher overall
(see Table 2).
Recall that we included a psychosocial scale
assessing campus racial/ethnic climate, and this
was also marginally significant in follow up
univariate analyses, F(6, 1195) ⫽ 1.94, p ⫽ .07
of our omnibus interaction of first-generation
status and race/ethnicity. It is noteworthy that
our covariate, family income, represented a particularly strong influence on this scale (F(2,
1195) 6.26, p ⬍ .01), suggesting that the effects
of first-generation status and family income on
perceived ethnic climate are quite confounded
in this sample. However, when holding income
constant, multiple comparison post hoc tests
revealed that all African American students
rated this scale significantly higher than their
peers of other races. Latino first-generation students perceived this scale as significantly more
important than their nonfirst-generation peers
and all Asian and White students (see Table 3).
F
p
Psychosocial and Personal Variables
White M
(SD)
vealed no differences by gender, but did reveal
significant differences in family income categories. First-generation students were underrepresented at the highest category of income (4.8%
for first-generation vs. 12.8% for nonfirstgeneration) and overrepresented at the lowest
category (39.6% for first-generation vs. 13% for
nonfirst-generation), ␹2[4, 1243] ⫽ 170.17, p ⬍
.001. We also found significantly more ethnic
minority students in the first-generation sample
(47%) compared to the nonfirst-generation sample (22%), ␹2[3, 1339] ⫽ 100.55, p ⬍ .001.
Latino M
(SD)
.06
A-A M
(SD)
2.76
Asian M
(SD)
2.88 (.59)
2.57 (.63)
White M
(SD)
2.77 (.65)
2.51 (.64)
Latino M
(SD)
First generation
Not first generation
A-A M
(SD)
p
Asian M
(SD)
F
Scale or item
Female
Nonfirst-generation status (n ⫽ 859)
Male
First-generation status (n ⫽ 556)
Generation status
Table 3
Perceived Parental Preference and Racial and Ethnic Climate Scales by First Generation Status and Race/Ethnicity (N ⫽ 1,339)
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Table 2
Psychosocial Scale by Gender (N ⫽ 1,339)
.006*
CHO ET AL.
Importance of
parent
preferences
Racial/ethnic
climate
scale
100
.01*
2.77
2.68 (1.15)
2.85 (1.04)
3.00 (1.00)
3.46 (.87)
4.00 (.00)
3.44 (.93)
***
p ⬍ .001.
Academic Quality
African-American.
a
Academic
quality
Male
Female
Financial
considerations
Male
Female
Financial
availability
Male
Female
*
p ⫽ .01.
**
p ⬍ .05.
3.08 (.97)
3.27 (.94)
3.61 (.63)
3.61 (.79)
3.20 (1.03)
3.77 (.65)
2.76 (1.05)
3.35 (.91)
Asian M
(SD)
Item or scale
101
Overall, our results suggest that measures of the
perceived psychosocial quality of a campus affect college choice substantially. However, as
one might expect, racial climate is more important to some groups of underrepresented students. Gender differences in perceived school
climate reflect women’s greater interest in physical safety and social acceptance. A secure and
welcoming feeling on campus seems to be important to both females and first-generation
males. All African American students and
Latino and Asian first-generation students appear to weigh parental input in their college
choice process as more important than their
peers of other races. Acceptance of racial diversity was more important to all African American respondents and Latino first-generation students in comparison to their peers of other
races.
We also observed an omnibus interaction of
first-generation status and race/ethnicity, F(6,
1195) ⫽ 7.56, p ⬍ .001, and the follow up
univariate analysis revealed a significant effect
for students’ perceptions of their parents’ preferences, F(6, 1195) ⫽ 3.06, p ⬍ .01. Asian- and
African American first-generation students differed most strongly from their nonfirstgeneration peers in their reliance on their parents’ preferences for which college to attend,
followed by Latino first-generation students,
with relatively little difference for White students (see Table 3).
2.27 (1.04)
2.92 (1.16)
.03*
2.33
2.68 (.93)
2.77 (.88)
2.88 (.93)
3.11 (.97)
3.50 (.71)
3.00 (1.13)
2.69 (.86)
2.91 (.85)
2.60 (1.17)
3.50 (.65)
2.64 (.99)
3.04 (.86)
3.00 (.98)
2.96 (.91)
2.38 (.98)
2.96 (.87)
.000***
4.34
2.87 (.66)
3.00 (.61)
3.00 (.58)
3.29 (.63)
3.67 (.47)
3.31 (.64)
3.04 (.57)
3.22 (.58)
3.00 (.59)
3.63 (.45)
2.92 (.64)
3.35 (.54)
3.43 (.48)
3.33 (.55)
2.58 (.69)
3.06 (.63)
White M
(SD)
Latino M
(SD)
A-Aa M
(SD)
Nonfirst-generation status (n ⫽ 859)
Asian M
(SD)
White M
(SD)
Latino M
(SD)
A-A M
(SD)
a
First-generation status (n ⫽ 556)
Table 4
Academic Quality and Financial Scales by First Generation Status, Gender, and Race (N ⫽ 1,339)
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
F
p
GENDER, RACE, AND SES IN THE COLLEGE CHOICE PROCESS
We also hypothesized that first-generation
participants would not differ from nonfirstgeneration participants in their ratings of academic quality scale. However, contrary to our
expectations, a significant omnibus three-way
interaction between first-generation status, gender, and race, F(6, 1195) ⫽ 4.93, p ⬍ .001
remained significant in follow-up univariate
analyses based on the academic scale, F(6,
1195) ⫽ 4.34, p ⬍ .001. Compared to other
groups, first-generation females, African American in particular, considered the academic scale
more important in their choice of college. Surprisingly, Asian males overall rated academic
quality lower in importance compared to their
peers. However, it is important to note that,
irrespective of first-generation status, gender, or
race, all students rated academic quality as the
102
CHO ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
top influence in their college choice process (see
Table 4). As with perceived ethnic climate, our
covariate, family income, represented a particularly strong influence on ratings of academic
quality, F(2, 1195) ⫽ 76.57, p ⬍ .001. Firstgeneration students in the lowest income category (less than $35,000) rated academic quality
more important than their more affluent peers.
Financial Considerations
In testing our hypothesis concerning financial
considerations, our omnibus three way interaction between first-generation status, gender, and
race remained significant in follow up univariate analyses because of respondents’ perceptions of college affordability, F(6, 1195)
⫽ 2.33, p ⫽ .03. African American firstgeneration females and African American nonfirst-generation males rated the college cost
variable highest of all groups in our sample.
Finally, Asian students showed the greatest gender differences in their ratings of college cost,
with Asian nonfirst-generation males rating this
scale considerably lower than their peers (see
Table 4).
A similar three way univariate effect of firstgeneration, gender, and race emerged for financial aid availability, F(6, 1195) ⫽ 2.33, p ⫽ .01.
Univariate analyses revealed that African
American first-generation females, African
American nonfirst-generation males, and all
Latino first-generation participants rated financial aid availability more highly in the college
choice process than other groups (see Table 4).
Asian males again rated financial aid availability considerably lower than their peers. As
would be expected, our covariate, family income, was the most powerful influence on perceived affordability, F(2, 1195) ⫽ 21.26, p ⬍
.001 and financial aid availability, F(2, 1195) ⫽
129.21, p ⬍ .001 of all variables under consideration.
Discussion
We designed this study to understand how
psychosocial, personal, and institutional factors
jointly influence the college choice process and
how ethnicity, gender, and family income might
moderate those relationships. Our findings demonstrated that some psychosocial factors are
indeed important considerations for students in
the process of selecting a college, although not
as important as financial and academic factors.
When choosing a college, first-generation students and female students, in particular, were
most sensitive to psychosocial characteristics
that included perceived safety, positive social
climate, and having friends present on campus.
Our finding is quite consistent with and expands
on findings from a national survey of high
school seniors who planned to attend college a
decade ago (Market Intelligence for Higher Education, 1997). A total of 57% of all students in
that study rated campus safety as “very important,” second only to the quality of the facilities
(e.g., labs, dorms, technology) offered by the
campus. Among female students, that total rose
to 72%, suggesting that differences between
male and female students in their concerns
about campus safety are a relatively stable phenomenon. However, findings reported here also
suggest that among first-generation students,
psychosocial characteristics are even more important than for the general college student
body. Similarly, a decade ago 58% of all students cited a wide variety of social activities as
important in choosing a college, again suggesting that psychosocial variables are important
and persistent factors in college choice. Prior
qualitative research with females (Horvat,
1997) that examined the idea of a campus “fit”
similarly found that for females in particular,
social and psychological variables can be critically important, especially in the latter stages of
the choice process. Finally, our finding that all
African American students and Latino firstgeneration students reported that the ethnic
make up of the campus and the community was
important in their choice process is consistent
with survey findings. A decade ago, 77% of all
students (88% African American and 74%
White) reported that an ethnically diverse student body was an important factor in their college choice. Again, our findings also clarify that
first-generation status relates differentially to
the importance of perceived ethnic climate according to one’s ethnic/racial background.
Overall, our findings support the importance
of habitus, or the fit between a student’s social
and psychological needs and the perceived campus milieu, in the college choice process. This
study extends previous findings on habitus to
the population of first-generation college students. It is notable that African American stu-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
GENDER, RACE, AND SES IN THE COLLEGE CHOICE PROCESS
dents rated the scale specific to racial/ethnic
climate more important than other groups of
students, regardless of generation status. Latino
students differed significantly according to generational status, and Asian (especially males)
rated the ethnic climate scale relatively unimportant. On most college, campuses Latino and
African American students in particular are severely underrepresented. A wealth of data point
to the stress and isolation many underrepresented students feel on campuses when they are
a small minority (Hurtado, Milem, ClaytonPedersen, & Allen, 1998; Watson et al., 2002).
Thus, the importance of ethnic climate relative
to other considerations may be especially acute
for underrepresented students.
One interpretation of findings concerning the
psychosocial scales becomes visible when we
use the lens of the psychological literature on
stigma (Major & O’Brien, 2005), or perceived
devaluation based on personal characteristics
(e.g., race, SES). Current social psychological
research on stigma concludes that members of
socially stigmatized groups are aware that they
are devalued by the dominant group, recognize
the negative stereotypes of their group from a
young age (Hudley & Graham, 2001), and often
perceive that their identity is being threatened
by virtue of their membership in a stigmatized
group (e.g., ethnic minorities, low income).
Members of stigmatized groups often cope by
identifying more closely with their own group
(Allport, 1954) as a way to seek social support
and a sense of belonging that supports their self
esteem. As a result, first-generation or ethnic
minority students who are underrepresented and
most likely to be stigmatized as incompetent in
the academic domain (e.g., African Americans,
Latinos) (Hudley & Graham, 2001) may be
more vigilant for social opportunities that provide a sense of belonging and connectedness
when they are selecting a college. They seek
campuses where they will find friends, members
of their own group, and supportive social activities. Our scale may be tapping into students’
desires to protect themselves against threats to
their identity and feelings of devaluation.
An alternative interpretation of our results
specific to the racial/ethnic scale is that among
our students who identify as Latino or Asian,
more may be from families who have immigrated to the United States relatively more recently than their African American peers. Ogbu
103
proposed a theory of minority school performance (Ogbu, 1978; Ogbu & Simons, 1998)
that focuses specifically on immigrant status.
According to this theory, immigrant students
are less likely to be concerned with campus
ethnic climate than are students who belong to
what he labels castelike groups because of their
orientation toward schooling and American society. Immigrant students see education as an
instrumental means to credentials for success in
ways that are not possible in their home country
and are more likely to see discrimination as a
temporary problem of adjustment. Castelike
groups, which include low-income, disadvantaged African Americans and Latinos, have visible evidence that discrimination is a permanent
feature of life that sometimes negates the benefits of an education for members of their communities. Thus, castelike students are more
likely to require that schools demonstrate their
commitment to diversity concerns, including a
policy of recruiting adequate proportions of
members of their racial/ethnic groups.
We also found significant interactions in our
academic quality and financial scales, suggesting several important considerations. Perhaps
most importantly, nonfirst-generation males
rated academic quality considerations lower in
importance than any other generation/gender
groups, whereas first-generation females rated
the scale as the most important of our six scales
or variables in the college choice process, even
more so than financial considerations. Our scale
asked students to rate the importance of available academic coursework and support programs while in college as well as the value of
the academic preparation for future career and
graduate school plans. We also found that all
females, as well as African American and
Latino students were much more influenced
than their male peers and Asian and White
students by the cost of college and financial aid
considerations in making a college choice. Although our data are somewhat consistent with
previous findings that academic and financial
variables are critical factors for all students
(Hossler, Braxton, & Coopersmith, 1989;
Hossler & Gallagher, 1987; Hossler, Schmit, &
Vesper, 1999; McDonough, 1997; McDonough
& Antonio, 1996), we also found that firstgeneration female students and underrepresented ethnic minority students are especially
sensitive to the value of their college choice for
104
CHO ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
their college success. We speculate that these
students are the most likely to lack resources or
family and social networks outside of college
that can improve their access to the opportunity
structure (Stanton-Salazar, 2001); thus, they are
aware that they are more completely dependent
on the outcome of a college career to improve
their life chances.
Implications for Higher Education
Programming
Our investigation adds complexity to the notion
of what it means to be “first-generation,” providing evidence that first-generation students are not
a homogeneous group, even in their concerns
about how to finance a college education. This
understanding has obvious implications for college recruitment and retention efforts. Clearly, a
one-size-fits-all strategy to recruit first-generation
students is likely to be ineffective. All students
were very interested in the quality of the academic
offerings and supports, as well as the potential for
future benefits they might derive from higher education. Thus, recruitment programs that focus on
matching prospective students’ interests and future aspirations to available university programs
and supports will undoubtedly be more successful
in retaining students than programs focusing more
generally on increasing admission and matriculation rates. As well, universities might consider
more carefully the interpersonal quality of their
campuses and plan strategically for students to
develop social networks within their ethnic/racial/
gender groups as well as strong, positive crossgroup ties. Recall that the social psychological
literature makes a strong case for campuses to
support gathering in groups that share common
backgrounds, cultures, and values, and this is especially true to emotionally support members of
groups who are typically stigmatized in this society. However, the development of cross group ties
can be critical for developing the kind of intellectually challenging, rich experiences that should be
the hallmark of a university education and that
will support positive intergroup experiences into
adulthood (Schofield, 2001).
Limitations and Suggestions for Future
Research
Our study examined the experiences of firstgeneration students in a comparative frame-
work, which limits the attention we have paid to
within-group variability for the sake of analyzing between group differences. Our analyses
have revealed the variability inherent in the
label “first-generation;” it will be necessary as a
next step to examine that variability through
other lenses (e.g., ethnic, region of the country).
For example, an investigation into how SES,
gender, and high school achievement for a single ethnic group affect first-generation students’
choice of college and college adjustment will
allow us to more closely consider how best to
match the environment on college campuses
with particular students’ academic and psychosocial needs.
The response rate for our web survey was
25%, which is within the range reported for
previous Web surveys. For example, Fricker
and Rand (2002) in a meta-analysis reported
that the response rates of six studies published
between 1998 and 2001 ranged from 8% to
44%. Porter and Whitcomb (2003) examined
students’ perceptions of college and reported a
response rate of 14.8% for their Web based
survey. Some research suggests that survey response rates are a function of the length of time
to complete the survey. Kittleson and Brown
(2005) suggest that increased response rates of
Web surveys can be gained if the time required
to complete the survey is 3 to 5 minutes or less;
our survey required 15 to 20 minutes to complete. Future researchers may design a survey
that balances relevant questions, a reasonable
number of surveys per year, and a reasonable
response rate.
Our response rate may also have suffered
because of incorrect or infrequently used email
addresses (Sax, Gilmartin, & Bryant, 2003) or
spam filters. Registrar’s Offices at the four universities from which we collected the data for
this study issued the official campus student
emails. However, we had no way of identifying
how often the E-mails were checked, and students may have other primary E-mail accounts.
Further, spam filters may have been too sensitive to include our E-mail invitation on some
students’ personal computers. Future researchers might take advantage of multiple means of
advertisement for the survey including campus
newspapers, campus radio station announcements, banners in residence halls, participation
in popular campus blogs, and classroom announcements.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
GENDER, RACE, AND SES IN THE COLLEGE CHOICE PROCESS
A number of researchers have pointed out
several advantages of Web surveys. Surveys are
self-administered and anonymous that provides
a high level of privacy to the respondent, increasing the likelihood of honest responses.
Web surveys are also free of potential bias from
an interviewer or test administrator (e.g., McDaniel & Gates, 2005).
Another limitation of our study is the use of
the three single-item measures (i.e., personal
family preference) with three multiple-item
measures (i.e., the Academic Scale). We took
strides to balance between using multiple items
for each construct that we investigated and
keeping the number of items as little as possible
to increase response rates. Our initial analysis
results revealed that reliability alphas for some
scales were relatively low, and thus we decided
to use three single items instead. It has been
commonly known that multiple items provide a
more discriminating response scale than one
item, and are inherently more reliable by allowing for calculation of coefficient alpha than a
single item only when common methods bias in
predictor and criterion is taken into consideration. Bergkvist and Rossiter (2007), however,
argued that multiple items are not necessary if
attribute is concrete (by expert judgment). They
went on to argue that additional items put into a
scale run the risk of tapping into another predictive attribute. Indeed, the authors compared
multiple-versus single-item measures. Using
correlations and regressions, they found that
single-item measures demonstrate having as
good psychometric prosperities as multipleitem measures. If single-item measures are as
strong as multiple-item measures, future researchers should consider increasing the use of
single items for Web surveys since the number
of items on the survey is one of the factors that
determine response rates. However, before this,
a through investigation on multiple-versus single-item measures of dependent and independent variables of a study should proceed.
Finally, we must acknowledge another limitation of our data. Despite substantial advances
on recruitment and retention of underrepresented groups, many 4-yr colleges still enroll
very small numbers of certain ethnic groups
(e.g., African Americans – 1.2 million; Hispanic – .77 million; Native Americans – .09
million; Whites – 6.8 million) (Cook & Cordova, 2007). Our data have significantly unbal-
105
anced cell sizes (for which we compensated
statistically), and our sample size of African
American males in particular was especially
small. Further, our number of Native American
students was so small as to be inappropriate for
analyses. If we are to understand the choices
and experiences of all students, we must develop strategies to ensure that all voices are
heard. In particular, examining within group
variability will require substantially more diverse pools of respondents.
We set out to answer two questions. Our data
make clear that students’ understanding of the
institutional character, or habitus, of a given
university played a part in their college choice
process, and the influence of habitus varied, as
we might expect, by first generation status as
well as by gender. Today, the main source of
modern success is education, and the children of
those who never entered college can dream of
going to college as a first step toward improving
their life chances. Although first generation and
minority students are actively recruited and
congratulated for getting into college, we understand too little about these students’ process
of college choice and the impact of choice on
their adjustment after enrollment. Continued attention to college choice for first generation
students can only improve equal access and
opportunity for all students.
References
Allport, G. (1954). The nature of prejudice. Boston:
Addison Wesley.
Antonoff, S., & Friedman, M. (1991). College match.
Alexandria, VA: Octameron Press.
Bergkvist, L., & Rossiter, J. R. (2007). The predictive
validity of multiple-item versus single-item measures of the same constructs. Journal of Marketing
Research, XLIV, 175–184.
Broxton, J. M. (1990). How students choose colleges.
In D. Hessler & J. Bean & Associates (Eds.), The
strategic management of college enrollments (pp.
57– 67). San Francisco: Jossey-Bass.
Bui, K. V. (2002). First-generation college students
at a four-year university: Background characteristics, reasons for purposing higher education, and
first-year experiences. College Student Journal, 36, 3–11.
Chess, S., & Thomas, A. (1991). Temperament and
the concept of goodness of fit. In J. Strelau & A.
Angleitner (Eds.), Explorations in temperament:
International perspectives on theory and measurement (pp. 15–28). NY: Plenum Press.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
106
CHO ET AL.
Cook, B., & Cordóva, D. (2007). Minorities in higher
education: Twenty-second annual status report :
2007 Supplement. Washington DC: American
Council on Education.
French, J., Rodgers, W., & Cogg, S. (1974). Adjustment as person-environment fit. In G. Coelho, D.
Hamburg, & J. Adams (Eds.), Coping and adaptation (pp. 316 –333). New York: Basic Books.
Fricker, R., & Rand, R. S. (2002). Advantages and
disadvantages of internet research surveys: Evidence from the literature. Field Methods, 14, 347–
367.
Gabriel, K. R. (1968). Simultaneous test procedures
in multivariate analysis of variance. Biometrika, 55, 489 –504.
Hearn, J. C. (1991). Academic and nonacademic influences on the college destinations of 1980 high
school graduates. Sociology of Education, 64,
158 –171.
Holland, J. L., & Richards, J. M. (1965). A factor
analysis of student “explanation” of their choice
of a college. American College Testing Research
Reports, 8, pp. 23.
Horvat, E. M. (1997, March). Structure, standpoint
and practices: The construction and meaning of
the boundaries of blackness for African American
female high school seniors in the college choice
process. Paper presented at the Annual Meeting of
the American Educational Research Association,
Chicago. (ERIC Document Reproduction Service
No. ED407884).
Horvat, E. M. (2003). The interactive effects of race
and class in educational research: Theoretical insights from the work of Pierre Bourdieu. Penn
GSE Perspectives on Urban Education, 2, 1–25.
Hossler, D., Braxton, J., & Coppersmith, G. (1989).
Understanding college choice. In J. C. Smart (Ed.),
Higher education: Handbook of theory and research (pp. 231–288). New York: Agathon.
Hossler, D., & Gallagher, K. S. (1987). Studying
student college choice: a three-phase model and
the implications for policymakers. College and
University 2(3):207–221.
Hossler, D., Schmit, J., & Vesper, N. (1999). Going
to college: Social, economic and educational factors’ influence on decisions students make. Baltimore: Johns Hopkins University Press.
Hudley, C. (1995). Assessing the impact of separate
schooling for African-American male adolescents.
Journal of Early Adolescence, 15, 38 –57.
Hudley, C., & Graham, S. (2001). Stereotypes of
achievement striving among early adolescents. Social Psychology of Education: An International
Journal, 5, 201–224.
Hurtado, S., & Carter, D. F. (1997). Effects of college
transition and perceptions of the campus racial
climate on Latino college students’ sense of belonging. Sociology of Education, 70, 324 –345.
Hurtado, S., Milem, J., Clayton-Pedersen, A., &
Allen, W. (1998). Enhancing campus climates for
racial/ethnic diversity: Educational policy and
practice. The Review of Higher Education, 21,
279 –302.
Ishitani, T. (2003). A longitudinal approach to assessing attrition behavior among first-generation students: Time-varying effects of pre-college characteristics. Research in Higher Education, 44, 433–
449.
Kern, C. W. K. (2000). College choice influences:
Urban high school students respond. Community
College Journal of Research and Practice, 24,
487– 494.
Kim, D. (2004). The effect of financial aid on students’ college choice: Differences by racial
groups. Research in Higher Education, 45, 43–70.
Kinzie, J., Palmer, M., Hayek, J., Hossler, D., Jacob, S.,
& Cummings, H. (2004). Fifty years of college
choice: Social, political and institutional influences
on the decision-making process. New Agenda Series,
5. Lumina Foundation for Education: Indianapolis.
Kittleson, M., & Brown, S. (2005). E-mail versus
web survey response rates among health educational professionals. American Journal of Health
Studies, 20(1), 7–14.
Lerner, J. V., Baker, N., & Lerner, R. M. (1985). A
person-context goodness of fit model of adjustment. In P. Kendall (Ed.), Advances in cognitive
behavioral research and therapy (Vol. 4) (pp.
111–135). Orlando: Academic Press.
Lutz, G. J. (2000). RoyTable: A program for generating tables of critical values for Roy’s largest root
criterion. Educational and Psychological Measurement, 60, 644 – 647.
Major, B., & O’Brien, L. (2005). The social psychology of stigma. Annual Review of Psychology, 56,
393– 421.
Market Intelligence For Higher Education. (1997). Importance of race, ethnicity, and other forms of diversity in college choice. Student Poll (Vol. 2). Baltimore: Arts and Sciences Group. Downloaded from
http://www.artsci.com/admin/pdf/ACF319D.pdf on
3/4/2006.
Massey, D. S., Charles, C. Z., Lundy, G., & Fischer,
M. J. (2003). The source of the river: The social
origins of freshmen at America’s selective colleges
and universities. Princeton: Princeton University
Press.
McDaniel, C., & Gates, R. (2005). Marketing Research Essentials (5th ed.) Hoboken, NJ: Wiley.
McDonough, P. (1997). Choosing colleges: How social class and schools structure opportunity. Albany: SUNY Press.
McDonough, P., Antonio, A., & Trent, J. (1997).
Black students, Black colleges: An African American college-choice model. Journal for a Just and
Caring Education, 3, 9 –36.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
GENDER, RACE, AND SES IN THE COLLEGE CHOICE PROCESS
McDonough, P., & Antonio, A. L. (1996). Ethnic and
racial differences in selectivity of college choice.
Paper presented at the Annual Meeting of the
American Educational Research Association. New
York.
National Center for Education Statistics (NCES).
(2001). Students whose parents did not go to college: Postsecondary access, persistence and attainment (NCES 2001–126). Washington DC:
U.S. Department of Education.
National Center for Education Statistics (NCES).
(2006). Digest of education statistics, 2005 (NCES
2006 – 030). Washington DC: U.S. Department of
Education.
National Center for Education Statistics. (2005). The
condition of education 2005 (NCES 2005– 094)
Washington, DC: U.S. Department of Education,
Nora, A. (2004). The role of habitus and cultural
capital in choosing a college, transitioning from
high school to higher education, and persisting in
college among minority and nonminority students.
Journal of Hispanic Higher Education, 3, 180 –
208.
Ogbu, J. (1978). Minority education and caste: The
American system in cross-cultural perspective.
New York: Academic press.
Ogbu, J., & Simons, H. (1979). Voluntary and involuntary minorities: A cultural-ecological theory of
school performance with some implications for
education. Anthropology and Education Quarterly, 29, 155–188.
Pascarella, E., Pierson, C., Wolniak, G., & Terenzini, P. (2004). First-generation college students: Additional evidence on college experiences and outcomes. Journal of Higher Education, 75, 249 –284.
Pike, G., & Kuh, G. (2005). First- and secondgeneration college students: A comparison of their
engagement and intellectual development. Journal
of Higher Education, 76, 276–300.
Porter, S. R. & Whitcomb, M. E. (2003). The impact
of contact type on web survey response rate. Public Opinion Quarterly, 67, 579 –588.
Roberts, J. S., & Rosenwald, G. C. (2001). Ever
upward and no turning back: Social mobility and
identity formation among first-generation col-
107
lege students. In D. P. McAdams, J. Ruthellen,
& L. Amia (Eds.), Turns in the road: Narrative
studies of lives in transition (pp. 91–119). Washington, DC: American Psychological Association
Sax, L., Gilmartine, S., & Bryant, A. (2003). Assessing response rates and nonresponse bias in web
and paper surveys. Research in Higher Education, 44, 409 – 432.
Schofield, J. (2001). Maximizing the benefits of
student diversity: Lessons from school desegregation research. In G. Orfield (Ed.), Diversity
challenged: Evidence on the impact of affirmative action (pp. 99 –109). Harvard Educational
Publishing Group.
Stage, F., & Hossler, D. (1988). Differences in family
influences on college attendance plans for male
and female ninth graders. Research in Higher Education, 30, 301–315.
Stanton-Salazar, R. (2001). Manufacturing hope and
despair: The school and kin support networks
ofU.S.-Mexican youth. New York: Teachers College Press.
Terenzini, P. T., Springer, L., Pascarella, E. T., &
Nora, A. (1995). Academic and out-of-class class
influences on students’ intellectual orientations.
The Review of Higher Education, 19, 23– 44.
Tinto, V. (1993). Leaving college: Rethinking the
causes and cures of student attrition (2nd ed.).
Chicago: University of Chicago Press.
Warburton, E., Bugarin, R., & Nunez, A. (2001).
Bridging the gap: Academic preparation and postsecondary success of first-generation students. Education Statistics Quarterly, 3, 73–77.
Watson, L., Terrell, M., Wright, D., Bonner, F.,
Cuyjet, M., Gold, J., et al. (2002). How minority
students experience college: Implications for
planning and policy. Sterling VA: Stylus Publishing.
York-Anderson, D. C., & Bowman, S. L. (1991).
Assessing college knowledge of first-generation
and second-generation college students. Journal of
College Student Development, 32, 116 –122.
Received July 20, 2007
Revision received November 13, 2007
Accepted November 19, 2007 䡲
E-Mail Notification of Your Latest Issue Online!
Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!