Robin Ryder pointed out to me this new experiment run by PLoS since March 2012, namely the introduction of a new article type, “called “Topic Pages” and written in the style of a Wikipedia article“. Not only this terrific idea gives more credence to Wikipedia biology pages, at least in their early stage, but also “the paper contains direct links to Wikipedia pages for background“. Now note that PLoS keeps a wiki separate from Wikipedia. I wonder about the development of a similar interface for statistics, maybe as a renaissance of the former StatProb wiki initiated by John Kimmel two years ago. And mostly abandoned for the past months…

When looking around the site I came upon a page on ABC written by Mikael Sunnåker et al.! A very nice survey of the existing debates around ABC, including uncertainties on the validity of the ABC approximation to the Bayes factor. Ad mentioning the original version of Donald Rubin (1984, AoS). As well as of Peter Diggle and Richard Gratton (1984, JRSS Series B). (I have a lingering feeling I may have seen this paper earlier as a referee and that I sadly missed the connection with this wiki page, hence refereed it as a “classical” submission… However, I just cannot remember whether or not this happened, nor can I find any trace in my past reviews! Which may hint at a weakness of this solution, by the way, namely that referees are less eager to review surveys than novel research articles…) To reinforce the above point, compare this page on ABC with the page on ABC produced by Wikipedia!