Jun 22, 2020
Discovery learning is an approach that I was trying out at around the time I was in the 20's of episode numbers of this podcast. I tried out the idea of Maths Circles, running a few of my own and attending a course about them in Notre Dame University in the United States. I also tried running a Self-Organised Learning Environment or SOLE, modelled on the work of Sugata Mitra, famous for his "hole-in-the-wall" experiments in India.
Since then, I have discovered the reasons why these sorts of approaches don't work, and I've been discussing this recently on the podcast. I also discussed the Sugata Mitra's apparent dishonesty in the episode on Hope in the Wall.
In this episode, we look at the history of the idea of discovery learning. First suggested in the 1960's by Jerome Bruner, it has since gone through several rounds of re-branding and repeated research.
The article in question is called Should There Be a Three Strikes Rule Against Discovery Learning? The Case for Guided Methods of Instruction by Richard Mayer. Specifically, it shows that pure discovery learning methods (where students are mostly left to their own devices) have abundant evidence that they are not as good as guided discovery methods (where the teacher provides feedback and hints during problem solving), though it seems that there is also some evidence that guided discovery methods are better than expository approaches (i.e. explaining everything in detail beforehand).
The defining features of the human mind that seem to be the cause of this are two: the limitations of working memory capacity on the one hand; and the human desire to avoid thinking where possible on the other. Discovery learning appears to overload working memory, whereas expository approaches might result in the students not actually thinking. In a way, it is two sides of the admonition "they need to engage with the material." Discovery learning focuses on engagement at the expense of the material, and expository methods focus on the material often at the expense of engagement, but it appears that guided discovery can avoid both of these traps with greater reliability, at least in some cases.
Enjoy the episode.