Citation
Peterfreund, A., Rath, K., Xenos, S., & Bayliss, F. (2007). The impact of supplemental instruction on students in STEM courses: Results from San Francisco State University. Journal of College Student Retention: Research, Theory & Practice, 9(4), 487-503.
Highlights
- The study’s objective was to examine the impact of Supplemental Instruction in college science, technology, engineering, and mathematics (STEM) courses on the course grades earned at San Francisco State University.
- The study compares outcomes for students who enrolled in a Supplemental Instruction class paired with a STEM course to those who enrolled only in the STEM course. The study used administrative records for students who enrolled in at least one STEM course that offered optional Supplemental Instruction from fall 1992 to spring 2005.
- When analyzing outcomes within gender, the authors did not conduct any statistical testing to estimate differences in the STEM course grades of students who participated in Supplemental Instruction and those who did not.
- The quality of causal evidence presented in this study is low because the study does not adequately control for baseline measures. Although the study did not conduct statistical testing to estimate differences, the low evidence rating means that we would not be confident that any estimated effects would be attributable to Supplemental Instruction; other factors are likely to have contributed.
Intervention Examined
Supplemental Instruction
Features of the Intervention
Supplemental Instruction at San Francisco State University is an optional intervention offered for 22 STEM courses—such as biology, chemistry, math, and physics—with the highest risk of low student performance. The Supplemental Instruction classes are led by peer facilitators for 90 minutes every week, where students are encouraged to work cooperatively on supplemental course materials. In addition to covering course-specific content, the Supplemental Instruction classes also included information on study and learning skills.
Features of the Study
The study used data from San Francisco State University administrative records. Institutional records of grades, courses taken, and demographic information were collected for about 12,000 students who enrolled in at least one of the following courses from fall 1992 to spring 2005: General Chemistry I, Introduction to Biology I, Introductory Statistics, Calculus I, or Calculus II. This set of foundational classes was used to identify the sample because students who are majoring in programs sponsored by the National Institutes of Health are generally required to take at least one of these courses.
The authors compared mean course grades for students who enrolled in one of 15 Supplemental Instruction classes paired with a STEM course. In addition to the five courses used to select the sample, the analyses included an additional 10 Supplemental Instruction courses that had cumulative enrollments of at least 40 students from spring 1999 to spring 2005. The authors did not conduct statistical testing to measure differences in outcomes between treatment and comparison groups within gender.
Findings
- When analyzing outcomes within gender, the authors did not conduct any statistical testing to estimate differences in the STEM course grades of students who participated in Supplemental Instruction and those who did not.
Considerations for Interpreting the Findings
The analyses are based on unadjusted means and do not account for any of the control variables in the protocol. The treatment and control groups are self-selected and the authors noted baseline differences between the groups on measures of prior academic achievement (SAT scores and high school grade point average) for 6 of the 15 STEM courses, although these analyses were not presented separately by gender.
Causal Evidence Rating
The quality of causal evidence presented in this study is low because the study did not adequately control for baseline measures. Although the study did not conduct statistical testing to estimate differences, the low evidence rating means that we would not be confident that any estimated effects would be attributable to Supplemental Instruction; other factors are likely to have contributed.