In the MOOC-study that I'm currently reading one of the first things that stroke me - building on my last comment - that the design of the environments can limit research. Of course it is lower hanging fruit to get data from say Linkedin and combine this with the course outcome. However it is a narrow way of thinking about skill building as we like to have those skills pay off in a narrow job related content. The data would be more interesting (as I would have done the study) to combine the outcome with online social data. In order to do this one needs more account linking. Say my course on forensic accounting increased the amount of tweets, linkage in this topic. To see how people put their skill to use one should not only confine themselves to the job environment. It might be that people take up tasks like cheat sheets, blogs, community work etc. These are all off springs of the course and create a notice-able but hard to combine data stream. In order for MOOCs to present this kind of work it also needs a portfolio space not only for papers but for such things as thoughts and comments. Because like cases and open questions they produce answers where the gained knowledge is presented. Only for now the structure obscures the quality of the comments. We can count how many, but not the quality in a comprehensive overview. This is a feature I would want to have created myself in order to capture what I see that is mostly hidden.