On Fri, Aug 19, 2011 at 8:25 PM, Slim Virgin slimvirgin@gmail.com wrote:
James, that stat about 30 percent is astonishing. Do you have a link?
Sarah
Hey Sarah,
I forget if we put it on Meta in our research documentation yet, but that number came out of the summer research work that I'm a part of, so I will clarify the exact nature of it for you...
We looked at the top English editors as defined by the "List of Wikipedians by number of edits". Out of the top 1,000, we searched through their first contributions to see how many were reverted for vandalism. We did this using a regex from an earlier paper,[1] and which works by searching revision comments for the various edit summaries that suggest a reversion of vandalism (like "rvv" or the automated ones). It's not perfect but the error rate is known to be pretty minimal.
Out of that group of top 1,000 by edit count and excluding bots, 30% began by making edits that the community reverted for vandalism at that time (so take into account changing standards of course).
That's just one rough measurement, but I think it communicates something important when thinking about how to treat new Wikipedians: we all start as clueless newbies, and some of us even start as outright vandals.