As much as it seems a pleasing result from a Wikipedia perspective, my quick summary of it is that there were a lot of uncontrolled variables and very small sample sizes. My quick summary of the experiment: 5 students attended 7 hours of traditional lectures about cells. Another 5 students were told to read Wikipedia articles about cells at their own pace, although there was some additional activity taking place as part of a process including the identification of other Wikipedia articles to read and then a collaborative activity where they drew diagrams of cells and shared them with others. Both groups were tested before and after about their knowledge of cells. The Wikipedia group learned more. To me, the differences between the two learning approaches has a lot of uncontrolled variables. For example, did the content of the lectures cover the same material as the Wikipedia articles (I presume there was no limit to how widely the students roamed by clicking links within Wikipedia articles, which may make it difficult to assess the breadth of knowledge to which the Wikipedia group was exposed). Was the quality of the content as good? What was the quality of the traditional lecturing, a lively entertaining lecturer or one droning along? I am sure that we would all think that makes a difference. There was no mention of the researchers interacting with the traditional lecture group, yet it seems there were interactions between the researchers with the Wikipedia group, which appeared to be trying to ensure they were reading related articles and to summarise their knowledge with shared diagrams. To me, the learning approaches were so different that it seems hard to argue that it was Wikipedia that made the difference. It may just as easily have been a boring set of lectures, the preparation and sharing of the diagrams, or the researcher interaction (being part of a research experiment may have been more fun/interesting than attending yet another set of lectures). To me, the need here is to control the variables a whole lot more. Maybe assign one group to read from a traditional textbook for N hours vs read Wikipedia for N hours. That would be a fairer test of Wikipedia's content (although not controlling for the novelty value). If I have a general criticism of research of educational methods in general, it is the use of regular teachers using traditional methods compared against excited researchers using a new method, which the researchers hope/expect to confirm their hypothesis - the presence of the researcher in the process brings bias. Sometimes there is an attempt to control this by first training a regular teacher to teach using the new method, but there is still the problem that the teacher involved is usually a volunteer, who is probably keen to try this new method, so again there is bias. Unless these studies involve enough students using randomly-chosen run-of-the-mill teachers over enough time, it is very difficult to judge if the new method has any real potential. I don't think most people would argue against Wikipedia being a useful self-learning resource, but I don't think this paper would persuade me that it can or should replace more formal educational processes. I note that since the students were tested on their knowledge of cells after the Wikipedia reading, they had a motivation that most self-learners do not! I think anyone involved in formal education will know the motivational power of exams :-) Sent from my iPad