NBER Working Paper Analyzes Academic Performance and College Dropout

Federal and state policy makers have been devoting considerable attention to student success in recent years. Dissatisfaction with student success has helped fuel what broadly can be termed an “accountability” movement. Accreditors have also been paying attention to student success. Over the past five years, more than one-fourth of requests for follow-up related to Standard 8 (Student Admissions and Retention) from the Middle States Commission on Higher Education (MSCHE) have concerned the need to strengthen student enrollment and retention. In that context, evidence-based research can be valuable.

A National Bureau of Economic Research (NBER) Working Paper written by Todd Stinebrickner and Ralph Stinebrickner, both of the University of Western Ontario, adds insight into factors leading to student dropout. The paper notes that reasons for dropout shift after a student reaches the midpoint of his or her college studies.

To overcome limitations in administrative data, the authors constructed and applied a longitudinal survey of students who entered Berea College in 2000 and 2001. Each student was surveyed 12 times, beginning just prior to the start of his/her freshman year and then immediately after the student left the college.

Berea’s focus is on providing educational opportunities to students from a low socioeconomic background. Such students have, in general, a higher dropout rate than students from higher socioeconomic backgrounds. All students at Berea College receive a full tuition subsidy and a large subsidy for room and board, making access to funds an insignificant variable. With the financial variable effectively set aside, the study focused on academic and non-academic factors that drive students to abandon their studies.

To further understand the role of academic performance, the study also applied a scenario in which grade progression cutoffs e.g., a minimum GPA, were removed. In that scenario, all student departure would be voluntary. No student would be compelled to leave school. Based on that scenario, the paper found that students suffering from poor academic performance generally left school by choice.

The paper’s two major findings were:

  • 45% of dropout following the first year of college, 45% of the dropout following the second year of college, and 36% of the dropout following the third year of college is the result of a student’s learning about his/her academic performance (narrowing of the gap between a student’s expected and actual outcomes).
  • Poorly-performing students drop out, because they conclude that remaining in school is not “worthwhile.” That decision is based on decreased “enjoyability” and adjusted expectations of post-college earnings.

The authors suggest that their study has a number of implications for pre-college preparation and for institutions of higher education. Two points relevant to higher education are:

  • Colleges should provide prospective students with sufficient information to form a more accurate understanding of the degree of preparation and effort required to succeed in college. In turn, that information could increase student effort and result in an improved academic performance. However, the tradeoff in foregone leisure time might present a barrier to increasing student study time.
  • Colleges could provide poorly-performing students with counseling to reduce their loss of enjoyability. The challenge confronting counselors is that students often leave school without warning or when they are away from the campus (on break following their first year of school).

Aside from the possible avenues for improving student retention discussed in the paper, colleges and universities would also need to build-in adequate assessment mechanisms. To do so, colleges and universities could survey incoming students about their expectations concerning study time, lab time, and grades, as well as their level of confidence, among other things.

Those expectations could be compared with actual outcomes from preceding classes. Students having the largest expectations gaps relative to the actual outcomes from previous cohorts and those with low confidence could be targeted for early intervention and robust monitoring of their academic performance and satisfaction.

In the initial pilot study, one could compare the retention rates for the targeted students against those from preceding classes while controlling for academic performance, among other variables. One could also compare the outcomes of “at risk” students (based on the expectations gaps and level of confidence) who received intervention with those who did not. Such a study should mandate who participates and who does not, so as to avoid self-selection/self-initiative issues. The data from the pilot project could help determine the overall effectiveness of such a targeted student retention approach. If effective, it could be expanded to all incoming students who are considered “at risk” based on the incoming survey data.

Such an approach could satisfy accreditors’ demands for efforts to improve student retention. It could leverage the latest research to perhaps overcome some of the limitations of earlier interventions that have resulted in relatively low marginal returns on increasing student retention. Finally, a statistically significant improvement in retention among the targeted students would provide some concrete evidence that a college is increasing retention not just from increased selectivity, but also from the value it provides its students from its academic and student support activities.




Comments for NBER Working Paper Analyzes Academic Performance and College Dropout

Leave a comment

 By sharing your story, you agree to our Terms of Use.