This paper presents the results from a large, randomized, controlled experiment conducted in the introductory statistics course at Brigham Young University. The purpose of the study was to assess the impact of multimedia lectures on student learning and attitudes. A randomized complete block design was implemented to evaluate the treatment that had two levels: multimedia versus overhead transparencies. Data was collected over four semesters on 5,603 students. Several student characteristics were measured and controlled for in the analyses. Our findings indicate that the multimedia lectures did not improve student learning or attitudes compared to the control group. However, our research also indicates that large, randomized, controlled experiments can be implemented in educational research. We advocate their use as the standard method of evaluation for educational innovations.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education