Date of Award


Degree Name

Doctor of Education


Counseling and Personnel

First Advisor

Dr. Edward L. Trembley

Second Advisor

Dr. Robert Brashear

Third Advisor

Dr. Robert Betz

Fourth Advisor

Dr. Avner Stern


In this study, the role of nonintellective variables as predictors of academic success in underachieving first year college students was explored. Traditional predictors of college grades, high school grade point average and standardized scores on the American College Testing examination (ACT), were also examined.

The purpose of this study was to assess whether college grades at the end of the first year would differ according to underachieving students' personality styles, expressed concerns, and behavioral correlates thereby offering the possibility that such nonintellective factors may be useful in the prediction of grades.

The sample was a randomly selected group of eighty-seven 18- and 19-year-old first year students entering Western Michigan University's Alpha program for underachieving students. Students completed the Millon Adolescent Personality Inventory-Guidance Form (MAPI-G) to provide measures of nonintellective variables. Pearson product moment correlations, step-wise regression analyses, factor analyses, and chi-square procedures were employed in statistical analyses.

High school grade point averages and ACT scores proved to be poor predictors of this sample's college GPAs with the highest relationship being R = .13.

The sample was dissimilar on the nonintellective measures employed in this study suggesting that underachievers may be heterogeneous on such variables.

The students in this sample also deviated significantly from the general population of adolescents in proportions of personality styles. The represented sample was more sociable and their personality styles were more rigid than the general populations of adolescents.

Several sets of predictive formulas were developed which predict these underachieving students' college grades in the R = .43 to .70 range.

No relationship existed between personality styles and college grades. Neither the styles considered as more adaptive nor the styles considered to be more successful for men or women were related to college grades. Rigidity of personality style did not appear to mediate college grades.

Results indicated a negative relationship between expressed concerns and college grades. Behavior problems also were negatively related to college grades. Behaviors demonstrating social nonconformity were uniformly negative in their relationship to college grades.

Theoretical and practical implications were proposed and recommendations for further research were offered.

Access Setting

Dissertation-Open Access