Previous research has shown that participation in programs designed to connect students to campus can increase student success and retention. However, very little of this research has controlled for selections bias. The purpose of this study was to use propensity score matching to control for selection bias while analyzing three campus connection programs for their relationship to academic performance, measured by GPA after the first and second year and measured by retention to the second and third year. The campus connection programs analyzed were membership in a Greek organization, participation in a learning community, and enrollment in an academic skills course.

Data for this study were gathered from a land-grant institution in the Midwest. Participants were first-year students directly from high school in the first-year cohort of 2004, 2005, 2006, and 2007. Propensity score matching was done using background variables including measures of academic ability, financial aid information, and college in which the student's academic major is housed. After propensity score matching was completed, the background variables were again included as control variables in calculating least-squared and logit regression for the analysis.

Results showed measurable differences in the success and retention rates due to the campus connection programs. Results indicated that membership in a Greek organization had a positive impact on retention to the second and third year, but no statistically significant relationship to GPA after the first or second year. Participation in a learning community showed modest gains in GPA after the first and second year and similarly modest gains in retention to the second and third year. Finally, enrollment in an academic skills course had a modest, negative relationship to GPA after the first and second year and no statistically significant relationship to retention to the second and third year. These mixed results demonstrate the complex effects of campus connection programs.

This study provides valuable information to faculty and staff who are responsible for implementing programs designed to increase student success and retention. Recommendations include encouraging Greek chapters to align programming more closely with the academic support services offered on campus and conducting assessments of the highest performing learning communities to identify the most impactful practices. These practices should then be replicated in other learning communities on campus. Additionally, it was recommended to conduct assessments of the various aspects of the academic skills course including instructor preparation and training and the degree to which students integrate course material into their academic experiences.

Students self-select into many campus programs and this study can also serve as a model of how to control for selection bias in assessing these programs. Additional suggestions for background variables are provided as well as implications for future research.