Jessica P. Marini

Exploring the Utility of Advanced Placement Participation and Performance in College Admission Decisions

The current study evaluated the relationship between various operationalizations of the Advanced Placement® (AP) exam and course information with first-year grade point average (FYGPA) in college to better understand the role of AP in college admission decisions. In particular, the incremental validity of the different AP variables, above relevant demographic and academic variables, in predicting FYGPA was explored using hierarchical linear modeling. The AP variables of interest included the following: the number of AP exams the student took, the number of AP exams the student took and received a score of 3 or higher, the proportion of the number AP exams the student took out of the number AP courses offered at his or her high school, and his or her average AP score, highest AP score, and lowest AP score. Results showed that the AP predictor that most improved model fit was the average AP exam score. The final model that included multiple AP variables and most improved model fit included the average AP score, the number of AP exams the student took and received a score of 3 or higher, and the AP exam proportion (which had a negative relationship with FYGPA). These results are particularly relevant and timely for college admission and measurement professionals as AP course-taking information as opposed to AP exam score information tends to be more regularly factored into admission decisions if and when AP information is considered at all.

Identification of Multiple Nonreturner Profiles to Inform the Development of Targeted College Retention Interventions

Throughout the college retention literature, there is a recurring theme that students leave college for a variety of reasons making retention a difficult phenomenon to model. In the current study, cluster analysis techniques were employed to investigate whether multiple empirically based profiles of nonreturning students existed to more fully understand the types of students with particular characteristics that are related to leaving college. Based on over 18,000 students who left their initial institution after the first year, analyses supported three clusters, which were labeled as Affordability Issues, Unexpected Underperformers, and Underprepared and Facing Hurdles. Follow-up analyses were then conducted to determine whether students from each cluster had different higher education trajectories. Students in the Underprepared and Facing Hurdles cluster were most likely to drop out of higher education completely or transfer to a 2-year institution. Those students in the Affordability Issues cluster were most likely to transfer to a less expensive 4-year institution. Finally, the Unexpected Underperformers behaved somewhere in between the other two clusters with regard to dropout and transfer behavior. The implications of these findings in terms of developing more thoughtful and targeted retention interventions for these different types of students are discussed.