Using Regression Analysis to Find Predictors of Performance in a Pharmacotherapy Course

Examine variables internal and external to the classroom and their relationship with final course outcomes within a Doctor of Pharmacy program using exploratory linear regression analysis.

Data were collected from students enrolled in a required Pharmacotherapy I course in fall 2023. An exploratory linear regression model was generated using 8 variables. The variables included in the final model were derived from Doctor of Pharmacy admissions data, course learning analytics, and a voluntary survey administered in the course about student study habits.

A complete data set was collected from 57 of 98 students (58%). Correlations among variables revealed that final course grade was positively associated with sense of belonging (r = 0.544), grade point average at the time of matriculation (r = 0.537), spending 15 or more hours on cocurriculars in a month (r = 0.421), and average practice quiz score (r = 0.214). Final course grade was negatively correlated with an early warning and intervention policy being triggered (r = -0.652), progression committee attendance (as related to other courses) (r = -0.553), attending office hours to study for an exam (r = -0.401) (open to all but required for students who trigger the early warning and intervention policy), and use of recorded lectures (r = -0.377). The revised regression model was statistically significant (F[8, 48] = 16.763, p <.001).

A combination of factors matter in determining final course grades, including those directly connected to the class and those outside of it. A similar approach could be replicated by others interested in exploring influences on learning outcomes in their own courses.