Research talk:Surviving new editor/Work log/2014-03-10

Monday, March 10th

Back to hacking on this today. The first thing I did was extend my data gathering re. survival months so that I can compare enwiki to other large wikis.

In the meantime, I want to start thinking about how to explore our sensitivity to parameters. There's really three chunks of user-time that we're looking at:

activation period $t_{1}$

A short period after registration to check to see if an editor will edit at all.

trial period $t_{2}$

The time between the activation and survival period where an editor is presumed to be testing out Wikipedia and Wikipedians are testing out the editor. I assume that this is the time when non-surviving editors tend to shove off and when surviving editors decide to stick around.

survival period $t_{3}$

The time after the trial period where we expect surviving editors to show some activity.

Since I've already explored activation periods in new editor, there's really two parameters that I want to explore: 1. The duration of the trial period 2. The duration of the survival period.

I have this dataset that, for each user, contains the number of edits that an editor performed in months 2-12 and year 1 and 2. This will let me play around with different boundaries for trial periods and survival periods.

Testing: trial and survival periods

I'll gather a dataset of survival based on the following periods in order to compare the effects of trial period duration and survival period duration.

Key: trial survival

$t_{2}$	$t_{3}$	0	1	2	3	4	5	6	7	8	9	10	11	year 1	year 2
1	1
2
3
4
5
6
1	2
2
3
4
5
6
1	3
2
3
4
5
6
1	4
2
3
4
5
6
1	5
2
3
4
5
6
1	6
2
3
4
5
6

--Halfak (WMF) (talk) 22:27, 10 March 2014 (UTC)