arXiv:2101.06755v3 [stat.AP] 25 Jul 2021 Are in-person lectures beneficial for all students? A Study of a Large Statistics Class Ellen S. Fireman, Zachary S. Donnini, Daniel J. Eck, and Michael B. Weissman * Abstract Over 1000 students over four semesters were given the option of taking an introductory statistics class either by in-person attendance in lectures, augmented by online recorded lectures, or by taking the same class without the in-person lectures. The all-online students did slightly better on computer-graded exams. The causal effect of choosing only online lectures was estimated by adjusting for poten- tial confounders using four methods. The four nearly identical point estimates remained positive but were small and not statistically sig- nificant. No statistically significant differences were found in prelimi- nary comparisons of effects on females/males, U.S./non-U.S. citizens, freshmen/non-freshman, and lower-scoring/higher-scoring math ACT groups. Keywords: online vs in-person lectures, statistics education, large online lectures, average treatment effect, causal inference * Ellen S. Fireman: Department of Statistics, University of Illinois Urbana-Champaign, 725 S. Wright St. room 101, Champaign, IL 61820, fireman@illinois.edu . Zachary S. Donnini: Yale University, zachary.donnini@yale.edu . Daniel J. Eck (corresponding au- thor): Department of Statistics, University of Illinois Urbana-Champaign, 725 S. Wright St. room 101, Champaign, IL 61820, dje13@illinois.edu . Michael B. Weissman: De- partment of Physics, University of Illinois Urbana-Champaign, 1110 West Green Street, Urbana, IL 61801, mbw@illinois.edu . Acknowledgements: We would like to thank Yuk- Tung Liu for essential technical assistance, Kurt Tuohy from the UIUC Atlas team for crucial timely help in obtaining anonymized covariates, and Karle Flanagan for informa- tive conversations about similar versions of another course. IRB approval was obtained for the project. 1