Interesting result that someone ought to try replicating: "observing a demo is worse than having no demo at all! The problem is that you see a demo, and remember it in terms of your misconceptions. A week later, you think the demo showed you what you already believed. On some of the wrong answers that students gave in Mazur’s study, they actually said 'as shown in the demo.' The demo showed the opposite! The students literally remember it wrong. People remember models, not facts, said Mazur. By recording a prediction, you force yourself to remember when you guessed wrong." We see what we expect to see, which is why it is necessary to be conscious of our expectations as we observe - which in turn suggests a need for an interactive and engaged learning process. "Confused students are far more likely to actually understand. It’s better for students to be confused, because it means that they’re trying to make sense of it all." So, what does this say to the 'worked example' set? Here's more from ICER 2011.