The Journal of College and University Student Housing - Volume 43, No. 1 - 16

Lauren Schudde

who made different actual housing decisions,
to obtain a more trustworthy estimate of the
impact of living on campus.
In the second step, I ran a regression model
that controlled for all of the independent variables listed in Table 1, weighted by the estimated propensity scores. The regression analysis
eliminates any remaining bias between students who live on and off campus, while enabling a simple test for differential effects with
the inclusion of an interaction term for family
income. The regression model included the
demographic, educational, and institutional
information used in the propensity score
model, an indicator of campus residency, and
measures of stressors students faced during
college, as well as the interaction between
campus residency and family income. Because
stressors during the ﬁrst year of college occur
after students' housing decisions, these measures were not included in the regression predicting students' propensity to live on campus.
The inclusion of the stressors in the ﬁnal regression model, however, means that they are
controlled for in the estimation of the impact
of campus residency on persistence.

The lack of variation in effects
across educational background may
indicate that ﬁnancial constraints,
rather than just cultural mismatch,
contribute to the low impact of
campus residency among lowincome students.
16

RESULTS
Who Lives on Campus?
To describe the sample, Table 1 presents means
and standard deviations for all of the covariates. The average student in the sample comes
from a family with an income of $78,589
(Figure 1 presents the income distribution of
the sample). Almost 14 percent of the students
in the sample come from households in which
neither parent attended college, 27 percent
come from households in which at least one
parent attended college, but did not earn a
bachelor's degree, and almost 60 percent of
students have at least one parent who completed a Bachelor's degree or higher. Fifty-six
percent of the sample is female. Sixty-seven
percent identify as white and approximately 12
percent identify as black, 16 percent as Asian,
and 4 percent as another race. Eight percent of
students also identify as Hispanic. The average
SAT score is 1070 and the average high school
GPA a 3.2. Seventy percent of the sample
attends public colleges. The analytic sample
is representative of high school seniors who
graduated in 2004 and immediately enrolled
in a four-year college (NCES, 2009), making it
a more advantaged sample (and more likely to
persist) than a sample representative of 2004
ﬁrst-year college students in general.
As indicated in Table 1, the covariate means
of students living on campus and students
living off campus are signiﬁcantly different
on all measures, with the exception of gender,
which conﬁrms the need to match students
based on their propensity to live on campus.
Several indicators of advantage, like family
income and parental education, correspond to
living on campus. Students living on campus

THE JOURNAL OF COLLEGE AND UNIVERSITY STUDENT HOUSING

The Journal of College and University Student Housing - Volume 43, No. 1

Table of Contents for the Digital Edition of The Journal of College and University Student Housing - Volume 43, No. 1