I teach math, but I think this point applies to most or all subjects: Early in the module when discussing 'learnativity,' we are told that case studies and other student interactions with material are important for a number of reasons, not the least of which it gives them more ways to remember the myriad things thrown at them. While this is true, it's only a good idea if they understand the principles behind the interaction. In other words, a student who only knows, for example, how to apply the Pythagorean Theorem to one situation but cannot tackle a question written in a different way hasn't actually learned anything. A recent study of standardized testing showed that students accustomed to a certain kind of test do not do well if presented with a test, say, from another organization. In other words, they know how to answer questions written a certain way, largely because they practice questions only in the one format. My point, then: do not rely only on student interaction with material. If they can answer that question but cannot transfer the knowledge, they are no better off than they were before.