It's hard enough trying to evaluate regular education because learning is not easily measured; if indeed we can even hope to do so. We can of course test to see what people have remembered or whether they can explain concepts in a clear and logical manner but how do we know what they have actually learnt? So if evaluating traditional education is tough how hard is it to evaluate a MOOC? A bit like nailing jelly to the wall as a colleague once said, though I can't remember when.
Via Alastair Creelman