Research talk:Teahouse long term new editor retention/Work log/2015-11-02

Monday, November 2, 2015

Setting up HostBot to collect more data that we can combine with the 14K+ users in the October 2014 dataset to gain a little more statistical power so that we can perform more in-depth analysis.

• same 5+ invite criteria as before
• no "expert" vs. "peer" invite template condition this time (this shouldn't matter)
• 50% control sample this time. Should still yield between 100 and 150 invitees per day, and an equal number of controls.

The first batch was sent out today, to 109 newcomers. We'll be running this for at least a month.

See the code changes captured in these commits [1], [2] Jmorgan (WMF) (talk) 19:15, 2 November 2015 (UTC)