Talk:Flow/New contributor survey

Comments

I concur that a survey of this kind could be useful.

1,000 new contributors (defined as users who have registered in the last 2 weeks, and made at least 5 edits during that period, and not been blocked) will be surveyed, split into two tranches - one for users who have contributed to a talk: namespace, and one for users who have not.

The target editor sample may possibly be too low and/or sample period too short. It would not include editors who have been blocked, but it would include those whose first edits were not constructive. It would also include editors whose only edits were article submissions to AfC, editors whose first and only edit is to make a request at en:Wikipedia:Requests for permissions/Confirmed, and editors whose first 5 edits were only to their user page. It would perhaps be interesting to know how many new accounts are actually created over a two week period.

I'm not sure about the legal/ethical issues of extracting email addresses and mailing users direct. It would perhaps also be interesting to know how many new accounts actually provide an email address, and how many have enabled the "Email this User" feature.

Kudpung (talk) 22:35, 14 November 2013 (UTC)

In order:

1. 1,000 responses is probably enough to make vague assertions about incoming users. I would love to go higher (and hopefully I'll be able to); 1,000 is set as a minimum bound because I can't guarantee how many users will actually fall into the sample category. As an example, if I run a query right now to see how many people have made more than 5 edits, registered in the last two weeks and got an email address (that's excluding "not been blocked" and any namespace restrictions) I get 2,704 names. I suspect the actual number, accounting for namespace restrictions and blocks, will be a lot lower.
2. It will include users whose first edits aren't constructive because that's very hard to automatically identify; undo/revert/so on are not distinctly tracked things. It will exclude users who, for example, contributed by way of a deleted page (that's excluded from user_editcount).
3. There will be namespace restrictions, solving for the other concerns about edit count.
4. 97,000 and "about half", re how many accounts are created and how many provide an email address. Okeyes (WMF) (talk) 22:55, 14 November 2013 (UTC)