Research talk:Onboarding new Wikipedians/Rollout/Work log/2014-03-14

Friday, March 14th

I have data for the new sample today. Here's the count of users I was able to get my hands on for each of the wikis:

wiki	eswiki	frwiki	zhwiki	ruwiki	dewiki	ptwiki	itwiki	fawiki	nlwiki	plwiki	viwiki	svwiki	ukwiki	kowiki	huwiki	hewiki
n (before)	2000	2000	2000	2000	2000	2000	2000	1695	1186	1176	920	811	584	525	568	585
n (after)	2000	2000	2000	2000	2000	2000	2000	1697	1121	1055	1239	781	655	671	595	541

Here's the updated figures:

Delta productive editor proportion. The difference in the proportions of productive new editors before and after the deployment of mw:Extension:GettingStarted is plotted with 95% CI error bars.

Difference in activation rates. The difference in the proportions of new editors before and after the deployment of mw:Extension:GettingStarted is plotted with 95% CI error bars.

Article editing activation rate. The difference in the proportions of new editors (main NS only) before and after the deployment of mw:Extension:GettingStarted is plotted with 95% CI error bars.

No meaningful changes here. The observed differences seem to get closer to zero and the error bars got appropriately narrower due to the increased observations for most wikis. You can tell which wikis didn't have more registrations to sample from the size of the error bars. --Halfak (WMF) (talk) 14:48, 14 March 2014 (UTC)

---

I just had another idea for a metric that might be meaningful here -- "returning editor". I define a returning editor as an editor who came back to make at least one edit after their first edit session. In this case, I allow editors who register and make no edits, but return to make an edit over an hour later to be considered "returning".

Returning editor rate. The difference in the proportions of returning editors before and after the deployment of mw:Extension:GettingStarted is plotted with 95% CI error bars.

Nothing new to say here. It looks like returning editor, productive editor and article editor all correlate strongly.

                 returning_editor day_editor main_day_editor day_productive
returning_editor             1.00       0.39            0.36           0.36
day_editor                   0.39       1.00            0.83           0.73
main_day_editor              0.36       0.83            1.00           0.88
day_productive               0.36       0.73            0.88           1.00

I take it back. It looks like returning editor is measuring something else. I'll have to consider that for expanding the Editor Engagement Vital Signs project. Anyway, on to comparison of scalar values. --Halfak (WMF) (talk) 15:19, 14 March 2014 (UTC)