This paper investigates how student engagement impacts learning in face-to-face and virtual classes. Our setting is an introductory financial accounting course taught by three instructors face-to-face in fall 2019 and virtually in fall 2021. Our proxy for student engagement is the accuracy of responses to in-class group polling questions. We measure learning using final exam scores. We hypothesize and find that higher levels of student engagement resulted in better scores on the final exam in both the face-to-face and virtual classes but that over the course of the semester engagement declined faster in the virtual classes than in the face-to-face classes. We also find that the decline in engagement contributed to lower final exam scores in the virtual classes relative to the face-to-face classes. Taken together, these findings highlight some of the challenges faced by educators when delivering content virtually.
Data Availability: Data are available from the authors upon request.