Research:How does the type of work that newcomers do affect their retention?

From Meta, a Wikimedia project coordination wiki

This question set is among the high level research priorities for the Wikimedia Summer of Research 2011. The full list is here, and this is categorized as "RQ3".


Topical clustering[edit]

(RQ3.1) Which topics of participation (e.g. science, the arts, breaking news vs history) are the most common entry vectors for new editors, and how does retention change based on the topic they are interested in?

(RQ3.2) Do new users' first edits tend to cluster around a small number of topics? How does this differ from the editing pattern of veteran editors? Does more tight clustering increase new editor retention, and can we use existing or new technology to facilitate it (i.e., SuggestBot2.0)?

(RQ3.3) What topics are we lacking and how could that integrate with the article feedback tool or expert recruitment?

Cohort differences[edit]

(RQ3.4) How is newbie work in 0-9 edits and 10-100 edits different?

(RQ3.5) How have the first 10 edits of a new registered editor changed over time?

(RQ3.6) How is retention affected by the amount of work that newbies do initially? [1]

(RQ3.7) Does the length of time between first IP edit and account registration correlate with editor retention?

Outside mainspace[edit]

(RQ3.8) Does registering an email address suggest a better expected retention rate?

(RQ3.9) Does having a confirmed email address predict a good-faith editor?

(RQ3.10) Does participating in a Wikiproject (either as an organizer or a member) affect editor retention?

(RQ3.11) Does creating and maintaining a user page affect editor retention?

(RQ3.12) Are certain user page features (userboxes, Babel templates, etc.) more or less indicative of an active editor?

(RQ3.13) Can userspace activity be better supported and funnelled towards mainspace editing?

(RQ3.14) Are new editors that frequently use edit summaries more likely to be retained?

(RQ3.15) Has the percentage of new users editing in good faith changed over time?

Further basic cohort questions[edit]

(RQ3.16) How many apparently good faith users (qual) create new articles each month? How many of those articles get deleted? Does that effect editors survival?

(RQ3.17) How many new highly active good editors begin their careers by creating an article? By vandalizing? By spamming? By fixing a typo?

(RQ3.18) Who are our new highly active editors? This could be a short essay that summarizes research from the other questions by a qual.


  1. Panciera et al., Wikipedians are born, not made., that, on average, editors will perform edit most frequently immediately after joining. They also showed that editors who performed 4-5 edits on their second day after registering, they had a %13 chance of making more than 250 edits in total.