Taking a Scientific Approach to Science Education, Part II—Changing Teaching

In Part I of “Taking a Scientifıc Approach to Science Education” (Microbe, April 2015, p. 152) we focused on the research on learning and teaching in undergraduate science. Our focus here is to address why these improved teaching methods are not the norm in college and university science classes, and what it will take to achieve widespread adoption. In Part I we discussed the cognitive psychology research on the nature of expertise and how it is acquired.We thenmapped those research fındings onto the learning and teaching of science, and presented examples of research in university science courses. In those examples, learning outcomes were compared between classes using conventional lecture methods and those that implemented the “practice of expert thinking with feedback” that cognitive psychology has shown is essential for developing expertise. These examples illustrate the large increases in learning that result with the research-based teaching methods and substantial reductions in failure rates. In addition to examples described in Part I, there is a vast literature of similar studies across the science and engineering disciplines. These results are summarized in a meta-analysis of 225 papers by Scott Freeman and coauthors (Fig.1). They compared failure rates and performance on identical or near-identical exams for courses that were taught using traditional lecture methods or incorporating “active learning” methods (our “authentic practice and feedback”). On average, in active learning courses, the failure rates are 35% lower, the exam scores are 0.47 standard deviations higher, and scores on concept inventory tests are 0.9 standard deviations higher. The larger difference on the latter type of test is to be expected, as concept inventories are carefully developed research instruments that specifıcally target the extent to which students learned to think like experts in the discipline. As discussed in Part I, active learning with feedback has been found to be necessary for developing expert thinking. According to Freeman and his collaborators, these benefıts of active learning methods over lecture instruction are consistent across all fıelds of science and engineering and all course levels and study designs. They also suggest that “more is better,” in that those studies with results that were well above the average also relied more extensively on active learning methods.