Derek Muller has created some great videos for science education. He posts them at Veritasium on YouTube. He summarizes one of his studies in the clip that appears below.
In this study, Muller had students complete a quiz about the force that acts on a basketball once it is thrown upwards. He then showed them a 10-minute video (actually one of a variety of 10-minute videos) in which a narrator presents the truth about this force. After viewing the video, the students completed the same quiz again and were asked what they thought about the 10-minute clip. Students seem to have liked the video, characterizing it as 'clear' and 'easy to understand'. However, their results on the post-video quiz weren't much better than on the pre-video quiz. It turns out that the students didn't learn much at all from the video even though they thought highly of it.
Muller decided to make a video that followed a different strategy. Instead of presenting a straightforward explanation of the force that operates on the ball, Muller's new clip showed one actor giving a popular but false account of this force. This was followed by another actor's correction of the first one's erroneous claims.
The pre- and post-video quizzes indicated that students learned more from this second clip. That's an interesting result -- students learned more by seeing an erroneous view (perhaps their own) exposed and corrected than by viewing a straightforward, correct explanation.
What strikes me as even more interesting is that the students didn't think as highly of the second video. They thought it was 'confusing'. The better learning experience was correlated with a more negative affect.
I guess this shouldn't be so surprising. Learning challenges preconceptions. Challenge and correction of an initial opinion are unsettling, breeding confusion as one realizes that one doesn't know what one thought one knew. Apart from the sense of embarrassment that can follow the exposure of an error in one's thinking (even if no one else knows about it), correction of an error goes against inertia, requiring more effort on the part of the learner in order to change her/his thinking.