Performance gaps in science are well documented, and an examination of underlying mechanisms that lead to underperformance and attrition of women and underrepresented minorities (URM) may offer highly targeted means to promote such students. Determining factors that influence academic performance may provide a basis for improved pedagogy and policy development at the university level. We examined the impact of class size on students in 17 biology courses at four universities. Although the female students underperformed on high-stakes exams compared with the men as class size increased, the women received higher scores than the men on nonexam assessments. The URM students underperformed across grade measures compared with the majority students regardless of class size, suggesting that other characteristics of the education environment affect learning. Student enrollment is expected to increase precipitously in the next decade, underscoring the need to prioritize individual student potential rather than yield to budget constraints when considering equitable pedagogy and caps on classroom sizes.
Bibliographical noteFunding Information:
Administrative data were obtained from 17 lower-division biology courses taken by 1836 students in fall 2016 (minimum class size n = 40, maximum n = 239; figure 1). To establish a collaborative research group, we solicited participation through an existing professional network from biology instructors who teach majors or nonmajors from a diverse range of institutions, and we received data from California State University, Chico; Cornell University; the University of Minnesota, Twin Cities; and the University of Puget Sound. The network was sustained through a Research Coordination Network funded by the National Science Foundation (RCN–UBE Incubator: Equity and Diversity in Undergraduate STEM; award #1729935). We compared (a) pooled exam grades, (b) pooled assessments of student knowledge other than exams (hereafter “non-exam grades”; e.g., discussion sections, laboratories, online activities, written assignments, low-stakes quizzes, as well as active-learning, in-class activities), and (c) final course grades, which reflect cumulative performance in all aspects of the course. We present analyses with transformed z-scores (a measure of how many standard deviations a value is from the class section’s mean score) for ease of interpretation.
- Behavioral science