A paired-course comparison of supplemental instruction and traditional tutoring

Supplemental Instruction (SI) is a peer-assisted academic support program designed to increase the persistence of early college students in challenging introductory courses. While a considerable literature exists on the outcomes associated with SI, many of these studies have methodological limitations, including failing to account for self-selection bias and treating attendance as a binary variable. This uses a paired experimental design in which replicate sections of an introductory chemistry course taught by the same instructor in the same semester at a small liberal arts college in the United States were assigned to either traditional tutoring (“TT”) or SI. Students (n= 138) were also classified as under-represented or marginalized students (“URM”) or non-URM. Results indicate that both groups of students were significantly more likely to attend SI than TT sessions. After controlling for attendance, URM students in both years and non-URM students in one year earned significantly greater final exam grades in the SI treatment than in the TT treatment. The paired course experimental design provides strong evidence that these grade differences are driven by the content of the SI sessions and not the abilities or motivation of the students who attended. Thus, confirming that Supplemental Instruction is effective at improving exam grades, particularly for URM students, in part because it attracts higher attendance.