Diverse Teams Tend to do Good Work in Wikipedia (but Jacks of All Trades Don't)

Abstract

We define network-based indicators of diversity for Wikipedia teams and users. A team of Wikipedia contributors is diverse to the extent that its members edit different articles. An individual contributor is a “jack of all trades” to the extent that she edits articles that are rarely co-edited by the same other users. For both indicators of team and individual diversity we propose a model-based normalization in which we compare observed values to expected values in a random graph model that preserves expected degrees of users and articles. Using data on all articles in the English-language edition of Wikipedia, we show that diverse teams tend to write high-quality articles, but articles written by teams containing jack of all trades contributors tend to be of lower quality. These findings are robust to several alternative explanations for article quality. We also show that the proposed model-based normalization of network indicators outperforms an ad-hoc normalization via cosine similarity.