The purpose of this study was to develop a prediction model for graduation rate at non-residential research universities. As well, this study investigated, described, and compared the student characteristics of non-residential and residential institutions. Making distinctions between significant predictor variables at non-residential research universities and significant predictor variables at residential institutions was also an aim. The researcher obtained data from the Integrated Postsecondary Data System. Student and institutional variables were analyzed using descriptive statistics, independent samples t-tests, analysis of variance, and regression analyses. Results indicated that student and institutional characteristics can be used to significantly predict graduation rate at nonresidential institutions with student variables yielding greater predictive power than institutional variables. As well, residential status was found to moderate the relationship between undergraduate enrollment and graduation rate. / Includes bibliography. / Dissertation (Ph.D.)--Florida Atlantic University, 2016. / FAU Electronic Theses and Dissertations Collection