graphic illustrating results of the meta-analysis

Meta-analysis comparing the effectiveness of active learning versus traditional lecture in introductory STEM classes.
Recently I had a chance to hear Sarah Eddy (now an Assistant Professor in the Florida International University Department of Biological Sciences) speak about the meta-analysis she and her colleagues, led by Scott Freeman at the University of Washington, completed of 225 studies comparing the effectiveness of active learning(1) vs. traditional lecture in introductory STEM classes(2). They considered studies which compared student learning outcomes in classrooms incorporating at least some active learning versus those traditional classrooms focused on lecturing. These studies each compared sections of the same introductory STEM course at the same university, as measured by examinations or concept inventories and also by looking at the percentage of students receiving a D, an F, or withdrawing from the course (the DFW rate).

The results of the meta-analysis, summarized in the figure above, are remarkable. The left panel shows the difference between student performance on examinations and concept inventories for those in active learning classrooms relative to traditional lecture sections, normalized by the standard deviation of the distribution of scores. Overall, averaged over subjects and studies, we see that student performance in active learning sections increased by approximately half of a standard deviation over performance in traditional sections. The right hand panel demonstrates that the failure rate of students in active learning classrooms was about 10% less than those in traditional classrooms – an effect which, for this sample of students, implies that students in a traditional lecture classroom were 1.5 times more likely to fail than those in an active learning classroom. Finally, it is worth noting that these results are replicated over many studies across a wide variety of STEM disciplines – the results are shown for individual subjects, the number of studies in each subject is noted below the data points, and the estimated uncertainty for each subject is also displayed.


Not surprisingly, these results have prompted many to call for an end to lecturing (only)! These results illustrate why many, perhaps most, of introductory MSU STEM education classes incorporate some active learning regularly. It also explains the focus of the STEM Alliance and affiliated activities on this and related topics.

Perhaps most important, these results illustrate that pedagogical practice matters – and that how we teach, and not just what we teach, is important. This lesson transcends STEM to all subject areas taught at MSU.


(1) Active learning can be defined as “a process whereby students engage in activities, such as reading, writing, discussion, or problem solving that promote analysis, synthesis, and evaluation of class content.”

(2) Freeman, Scott, et al. "Active learning increases student performance in science, engineering, and mathematics." Proceedings of the National Academy of Sciences 111.23 (2014): 8410-8415.

Feedback and suggestions, especially from the MSU community, welcome: email sekhar@msu.edu.

R. Sekhar Chivukula is the Associate Provost for Undergraduate Education and Dean of Undergraduate Studies at Michigan State University.