Although this study is good for grabbing headlines, the analysis seems a little bit shallow. For one thing, the focus is on STEM (Science, Technology, Mathematics and Engineering) disciplines, As someone who teaches at the college level in both a STEM field and a traditional humanities field, I am well aware that different areas require different methods. For instance, if one is teaching the basics of computational cognitive modeling, then some interactive segments are necessary. However, things work entriely differently if one is teaching, for instance, the history of the philosophy of mind. Another issue I have with the study is (as best I can tell -- I cannot access the original paper) that they do not control for lecturer effectiveness. To put it simply, we all know that some people are better at lecturing than others. That being said, even when teaching say, Cartesian Dualism, there are steps that can be taken to make lecture classes better. For instance, it is widely known that most humans have an attention span of between 10 to 20 minutes. So, it is simple enough to give everyone a break every twelve minutes, or so and tell a story, or some historical anecdote. Similarly, the Socratic approach, asking for input from students throughout the class and then encouraging discussion, can also make lectures much more effective and enjoyable. These are some of the things I do. That being said, I have known people who just drone on in a monotone, in lecture classes. Folks such as that can be utterly tedious. My point here is that unless the effectiveness of the teachers is taken into account, this study cannot be trusted.