Please forgive me: the effects of course repetition policies on college graduation rates

• In Event: Testing and Retention

Abstract

An overwhelming majority of colleges and universities have adopted course repetition policies that allow students to repeat coursework to improve their GPA.  However, it is theoretically ambiguous whether more generous course repetition policies improve graduation rates.  Allowing a student to repeat a low grade may open new academic pathways, for example by providing the opportunity to achieve the grade in an introductory level course required to pursue a major, or by creating a pathway to improve overall GPA and avoid academic probation.  Alternatively, the safety net of a course repetition and grade forgiveness policy might incentivize students to put in less effort, knowing that they can retake the class if they receive a low enough grade. Research on the impact of such policies on the academic success of students is limited and has thus far focused on data from individual institutions (Sovero & Griffith, 2024; Jiang et al., 2023).

 

In this paper, we use data from the Integrated Postsecondary Data System (IPEDS) paired with information on course repetition policies for 4-year public universities in the U.S. To date, we have hand collected data on repetition and forgiveness policies at roughly 50% of 4-year public universities in the U.S. between 2006 and 2016. We find that course repetition policies differ across several dimensions. The first dimension is how the grades of the initial and subsequent attempt are used to calculate GPA.  The most generous policy is referred to as grade forgiveness, which removes the initial grade from GPA calculations. The second dimension is the letter grade threshold, which determines which classes are repeatable.

 

We estimate a differences-in-differences regression model of graduation rates as a function of institution and year fixed effects, time-varying institutional characteristics, and the variation in grade forgiveness policies and grade thresholds that we observe over our sample period.  Because of the staggered adoption of grade forgiveness, we use the Sun and Abraham (2021) interaction-weighted estimator. We find that graduation rates decrease in the range of .5 to 1 percentage points after the introduction of grade forgiveness, with the largest impacts on the 6-year graduation rates. This suggests an increase in moral hazard/risk-taking in response to the availability of grade forgiveness. We do not find any gender differences in our estimates, but we find that graduation rates decline the most for white students compared to other races/ethnicities. To examine the impact of the letter grade threshold for grade forgiveness on graduation outcomes, we use a synthetic control estimator, which allows for valid inference under a small number of treated units (Abadie et al, 2010). Our initial results suggest that grade forgiveness policies with more generous letter grade thresholds for repeating coursework improve graduation outcomes.

 

Author

• Pete Pridanonda, University of California, Riverside
  Presenting Author