I recently ran across two unrelated but interesting posts about issues of bias on Wikipedia. It should hardly come as a shock that a huge community generated blob like Wikipedia has weirdnesses that reflect the properties of that community. That particular blob plays an important role in the world (I know look up stuff on Wikipedia all the time), so a better understanding those biases certainly has significant value.
TechCrunch reports on a piece on SomethingAwful.com about WikiGroaning. The idea is simple – compare the write-up on Wikipedia of things that are of arguably important in the “big sense” (say, Aristotle) to things that are well known now but arguably ephemeral in the long view (say, Oprah). Gee, now that’s probably not the balance an editorial board would have planned for, eh? This isn’t entirely surprising, but still disappointing. The pieces I’ve linked to have numerous pairs to compare, although the fact that they’ve brought those pairs to the attention of a wider audience may lead to at least some of the imbalances being addressed. Many of the pairs show a definite “nerd bias” (e.g., Lizard vs. Dragon), but others (like the aforementioned Aristotle vs. Oprah) are arguably a more general “current events and pop culture bias”.
And because none of us have any real work to do, some silly people created WikiGroaning.com, where one can type in pairs of words or phrases and they’ll compute the “nerd points” for each based on the contents and edit history of the appropriate entries on Wikipedia. I did a comparison of Jon Stewart and Edward Murrow and, big surprise, Stewart racks up way more nerd points than Murrow (113,720 vs. 14,380). Ditto when comparing Stewart to Walter Cronkite (14,462), and Stephen Colbert has even more nerd points than Stewart (141,964).
As is pointed out multiple times in the discussion of the TechCrunch piece, Wikipedia is still largely driven by the early adopters, and those are (not surprisingly) still nerds to a large degree. The real question will be whether the demographics of Wikipedia contributors will become more “mainstream” over time. If it does, some of the bias goes away of its own accord. If the bias, however, works to prevent that mainstreaming we could get a feedback loop that traps Wikipedia in this nerdly space. I don’t honestly consider that terribly likely ’cause the kids are growing up with it, but we’ll have to wait and see. The pop culture bias is perhaps more chronic, as most people in most times tend to dwell inordinately on the people and events of their day. At some level, though, I don’t really mind Jon Stewart having some huge entry as long as Edward R. Murrow has gotten his due.
The other (and arguably more important in the long run) issue is raised by John Naughton, who reports on serious gaps in the Wikipedia entry on “Spreadsheet”. Naughton is picking up on a post by Dan Bricklin (a key developer of VisiCalc – the first spreadsheet I had any contact with), who discusses his concerns at some length and raises important questions about how a community project builds “neutral” content in a way that incorporates the experience of heavily invested experts like Bricklin in the process. Important stuff and worth the read.No tag for this post.