Wikipedia talk:Wikipedia Signpost/2011-10-31/Recent research

Just wanted to say thank you for yet another interesting review of scholarly work on Wikipedia. This is always very appreciated! --Piotr Konieczny aka Prokonsul Piotrus&#124; talk to me 21:55, 1 November 2011 (UTC)
 * Yep. Good job.  Keep it up. - Peregrine Fisher (talk) 04:50, 2 November 2011 (UTC)

85% attrition
85% is a very high proportion. However I think you need to differentiate sharply here between newbies and the longer term users, and a sample of a million editors is by definition going to be mostly editors with quite low edit counts - As of September we [http://stats.wikimedia.org/EN/TablesWikipediaEN.htm#activitylevels didn't even have three quarters of a million EN wikipedia editors with >=10 edits.

Attrition rates amongst our longer term users are an order of magnitude lower than that. I've done some research on our administrators in the English wikipedia and if we were losing anything close to 85% per five months then almost all the 700 or so active ones would be drawn from the couple of hundred appointed in the last 27 months.

New editors are a very different proposition, yes when someone does their first edit or edits they rarely continue frequently editing. But when we surveyed "former" editors to ask why they'd left the most common response was "I've not left yet". It is easy for the more active editors to lose site of this, but the vast majority of Wikipedians are readers not editors. People can be avid and contented readers of this site and only once every few months do they spot something they want to correct.  Ϣere Spiel  Chequers  08:22, 3 November 2011 (UTC)