Archive for May, 2012

PLoS scientific paper gets also published on Wikipedia


Wikipedia is awesome, and we all rely on it on a day-to-day basis. For basic statistical topics, I usually find it very reliable. For more advanced topics, however, coverage is often sketchy or even non-existent. The article on ABC is exceedingly short; the one on the Wang-Landau algorithm is rather odd; the on on quantitative comparative linguistics is essentially a long list.

I often get annoyed when people complain about Wikipedia, for two reasons: 1. Although imperfect, it is pretty good, and often better than other more widely accepted general knowledge encyclopedias; 2. If you find a problem, you should fix it rather than complain.

Although I am far from being the most active contributor on Wikipedia, I edit it quite a lot, and am an administrator of Wikimedia Commons. But I have hardly done any contributions in Statistics, my area of expertise. The reason is simple: when I do Statistics, it is for my job, do I either do research or prepare lectures. Editing Wikipedia is a hobby, so I want it to be separate from my job.

But this is sub-optimal: Wikipedia could be improved if more specialists were to participate. Also, and I should try not to be too depressed by this, anything I write on Wikipedia will probably be read thousands time more than the scientific papers I write. One could thus argue that it would be an efficient use of my time for me and other statisticians to edit Wikipedia on Statistics topics. One could even go further and state that it should be one of our duties.

There are several obstacles here. The first obstacle is that Wikipedia’s model is that it can be edited by anyone; other encyclopedias which attempt to have articles written by specialists are nowhere near as successful (e.g. Citizendium). Some kind of assurance that an expert’s edits are given more importance than a layman’s could incentivize more experts to participate. I actually believe that in the large majority of cases, an expert’s contribution would be recognized as such and that no one would try to deconstruct it (most exceptions would be in topics subject to hot debate either in the scientific community or in the media). A second obstacle is that editing Wikipedia is not recognized in the career path of scientists: I might want scientists in general to participate in Wikipedia; but I won’t do it, because I really need to finish these three papers and grant proposals (a kind of NIMBY issue). A third obstacle is that scientists simply cannot be bothered to edit Wikipedia.

PLoS Computational Biology has started an initiative to address these obstacles. In a nutshell: scientists write a review paper. It gets published in the journal, and the scientists get a publication on their CV. At the same time, the paper gets copied over to Wikipedia. The initial version is marked as being written by specialists, but anyone can edit it (e.g. improve the wording, add relevant links, etc.). Everyone wins. The first such article is on Circular permutation in proteins.

Obviously, scientists should also edit Wikipedia even when they don’t get a publication out of it. But such initiatives will help get scientists involved. Explaining scientific topics to the general public is one of the duties of scientists, and Wikipedia is one of the best tools to do this.