The Journal of College and University Student Housing - Volume 46, No. 1 - 88

Jonathan Manz

Seven analyses of covariance (ANCOVA)
were used (one for each SCM value) to determine if there were differences in SCM values
between RAs and OSLs, while controlling for
pre-college leadership traits (RQ2). The covariate for the analysis was the pre-test score
related to the dependent variable SCM score.
The RA/OSL dummy variable served as the independent variable in the models.
ANCOVA has two assumptions: The independent variable and covariate are independent, and there is a homogeneity of regression
(Howell, 2010). However, this study used survey
data, and, as Keppel and Wickens (2004) noted,
results from non-experimental designs can still
be accurately interpreted, even when the first
assumption is not met. Despite the models
for controversy with civility, commitment, and
common purpose not meeting the first assumption after conducting an Analysis of Variance
(ANOVA), these models were still interpreted
due to the non-experimental nature of this
study. The second assumption was met after the
ANCOVA model was altered to examine interaction effects, and no statistical significance was
found between the pre-test variable and dummy
variable for each equation, which demonstrated
the homogeneity of regression.
The dummy variable (RA/OSL) was of
primary interest in this analysis due to the fact
that its significance determined if the two types
of student leaders differed in SCM values.
There were significant differences between the
scores of RAs and OSLs for the SCM values of
controversy with civility, consciousness of self,
collaboration, congruence, common purpose,
and commitment. Citizenship was the only
non-significant value. Table 3 presents the
ANCOVA data findings. When combining this
88

. . . instead of working towards a
common purpose, RAs are in fact
working towards diverse purposes.

information with the mean scores for RAs and
OSLs (see Table 4), one can infer that when
controlling for the pre-test score (covariate),
OSLs had significantly higher scores than RAs
for six of the SCM values.
Effect size was determined by the partial
Eta square statistic in addition to Cohen's d for
all instances when the dummy variable was
significant. The partial Eta square statistic was
found to be trivial for all seven SCM values,
as shown in Table 3. Similarly, results from
the Cohen's d analysis demonstrated a trivial
effect size for all models (see Table 4). These
numbers demonstrate that while the data from
the ANCOVA results may be statistically significant, they were not practically significant,
meaning that the magnitude of the difference
in socially responsible leadership between RAs
and OSLs was minimal.

DISCUSSION
The Input-Environment-Outcome (I-E-O) model
was the conceptual framework guiding this
study, with leadership development serving
as the outcome under review. The SCM was
used to measure leadership development outcomes. Only three of the seven SCM values
were significant in predicting the type of
student leader (RQ1). Differences in outcomes
between RAs and OSLs were minimal (RQ2).
Peer helpers were chosen to serve as the OSL

THE JOURNAL OF COLLEGE AND UNIVERSITY STUDENT HOUSING

The Journal of College and University Student Housing - Volume 46, No. 1

Table of Contents for the Digital Edition of The Journal of College and University Student Housing - Volume 46, No. 1