Another entry in the great debate, I suppose, though it comes down in part to what you mean by “on their own”:
Following an introduction on the general topic by their teachers, the school children were given a workbook of geometric tasks that they had to solve on paper and using a computer over four school periods. Calculating the surface area of Gran Canaria was just one of the real-world, free-form assignments the students had to tackle. The workbook material included explanations and examples of various problem-solving approaches. The teachers took a back seat during the session but were on hand to answer questions from the children, who worked in pairs.
“We expected students who were weaker at math to benefit more from a greater degree of guidance through the module,” reports Reiss. “But we didn’t see a significant difference between these and stronger students.”
This isn’t actually surprising to me — it reminds me in some ways of the NCAT Emporium Model, which has also seen spectacular success with traditionally low-performing students.
But there’s a lot of ingredients here that are essential — well-defined tasks, immediate feedback, required course meeting times, availability of teachers for assistance when online resources and peer resources fail, measurable outcomes.
This is why I’m watching the MOOC space with interest, but I am not sold on its direct application to higher ed’s pressing problems yet. The MOOC’s adaption into remedial use (for example) is going to require some additional contrivance. And I think we’re headed that way, but not there yet.