I think an unasked question is whether we might consider self-explanation not just a means but an end in itself. In other words, having students able to explain their newly learned content is itself an aim of education. Not just a means to exam success.

Interesting thought Alistair. i just read over the 2013 paper and Dunlosky's newer 'study like a champ' chapter on self explanation and he is very positive and does recommend self explanation but he does say the research is just not there to be unreserved in his recommendation.

It's the assumption that the best thing (as tested in one way) is the best thing for all purposes, all the time, is crazy. If your goal is joined up, connected meaningful learning then do more than just retrieval of bits of info.

Perfect, obvious really yet overlooked.

Yes true. Also, the whole group think expanded with social media, mixed with the efficiency principle (just do what works best All The Time)

Yes Efrat's a great explainer.

#UKEd#EduSky Now place that in the context of Fiorella & Mayer's Select-Organise-Integrate model (as in their Learning As A Generative Activity). More than just telling-and-testing.

#UKEd#EduSky The process where individuals generate explanations to themselves while learning to make sense of new info or to integrate it with their existing knowledge. Chi's research highlighted that students tend to achieve a deeper understanding and better retention of the material.

ChatGPT gives a full definition. I asked it: Give me a definition of self-explanation as researched by Michelene Chi.

#UKEd#EduSky It involves deep info processing, demands meaning & as a result greatly aids retrieval: self-explanation. Mikki Chi has over a decade of research on it & it's considered part of the cog sci cannon. Yet I rarely hear of such strategies being reported here. Is there a reason?