Class Size Does Matter and its Effects on Student Performance Varies by Gender and Ethnicity

Presented are estimates of the effects of class size on course grades in a one semester course in introductory economics at a medium size land grant institution in the Southwest. Demonstrated is that the grades of Hispanic undergraduates and white males are lowered by nontrivial amounts when class sizes rise above 49 students. The grades of white females are unaffected. These findings are consistent with research on outcomes in higher education showing that females, compared to males, and whites, compared to Hispanics are more likely to earn bachelor's degrees. The data were collected in 19 sections of varying sizes over a 10 semester period in which the author, using identical instructional and evaluation methods, was the only course instructor. Used was an econometric model that allowed for incorporation of sample selection bias as well as for the estimation of nonlinear class size effects moderated by gender and ethnicity