Date of Award

1999

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Psychology

Abstract

The 1990 national census report revealed that Indiana ranked 47 th in the nation in regard to individuals over the age of 25 with bachelor degrees. Forty-nine percent of the 1995–96 Indiana high school graduates matriculated into Indiana higherEducation institutions—56 percent in-state and out-of-state (Indiana Commission for HigherEducation Report, 1998). However, the African-American student persistence rate is 24 percent, regarding those who attain a bachelor degree within four years compared to 51 percent for all students. However, if extended to six years, the percentage increases to 42.2 percent compared to 57.5 percent for all students. This study determined certain retention factors that can be used as predictors that impacted persistence for African-American students and which were not the same predictors that impacted persistence for White students at Indiana State University. The subjects in this study included all freshmen students who completed the 1996 and 1997 Student Information Questionnaires during the preceding summers of fall 1996 and 1997, respectively. The logistic regression process was used to identify retention factors that were significant for each cohort and ethnic group by eliminating those variables that did not meet the .05 level of significance. This process resulted in a persistence equation with an odds ratio which determined the odds of each group's persistence based on the significant variable. The predictor variables that impacted persistence for all students in both the 1996 and 1997 cohorts were also the predictor variables that impacted persistence for White students in both cohorts. The predictor variables that impacted persistence for African-American students in the 1996 and 1997 cohorts were different from those variables that impacted persistence for White students and students overall. The logistic regression process tested for model fit to the data and correctly predicted 72% to 99% of the overall cases.

Share

COinS