COLLEGE IMPACT: A QUANTITATIVE STUDY OF INSTITUTIONAL PREDICTORS OF COMPLETION RATES AT FOUR-YEAR PRIVATE INSTITUTIONS
ABSTRACT LORI WRIGHT CROMWELL COLLEGE IMPACT: A QUANTITATIVE STUDY OF INSTITUTIONAL PREDICTORS OF COMPLETION RATES AT FOUR-YEAR PRIVATE INSTITUTIONS Under the direction of EDWARD BOUIE, Ed. D. The unique tension caused by the interdependent relationship between access, affordability, and sustainability adds an additional level of complexity for administrators in balancing societal educational needs, economic realities, and institutional success. Trends suggest postsecondary six-year degree completion remains stagnant despite most full time students receive some form of student aid. Despite higher education institutions spending more money competing for enrolled students, delay in the successful completion of program requirements remains a challenge and a priority across most institutions. This study targets persistent problems of college completion in the United States by examining the predictive role of institutionally controlled factors on institutional quality as defined by completion rates. Using the Integrated Postsecondary Education Data System (IPEDS), the sample included institutions with Carnegie classifications as four-year, private, non-profit colleges. Degree-granting with full-time, first-time undergraduate students, and eligible to participate in federal student aid programs were also considered. The study included 509 institutions (n=509), 35 independent variables including structural, institutional allocations, and institutional resources, dependent variable graduation rate, and control variables percent of Pell grant and the total amount of Pell grant awarded with data ranging from 1996-97 to the 2015-16 academic years. Descriptive statistics, two-step cluster analysis, principal components analysis, along with multiple regression analysis was used to determine which, if any, institutional factors selected as independent variables significantly predict graduation rates. The study confirmed differences exist between clusters. As a result, linear regression analysis was used to ultimately determine for institutions included in cluster one institutional components are not predictive of six-year graduation rates. Resources per capita seem to matter for this group, although not predictive of six-year graduation rates. Likewise, institutional component resources per capita did significantly predict graduation rates for institutions in cluster two and three. As a result, the findings of the study largely support college impact literature theorized conceptually in the between-college framework detailed by previous research models of student involvement. These models and study suggest student characteristics and experiences in the college/institutional environment influences those outcomes.
Cromwell, Lori Wright