[16:35:50] lol [20:24:50] hmm maybe I should add this channel to my auto-join list [20:30:44] probably [20:30:55] guillom: watching Chocolat, I am going to the village in which it is set next week [20:31:11] I've never watched it [20:32:41] moi aussi, mais ma soeur insiste que je voir :) [20:43:49] hi howief [20:44:26] her werdna [20:44:28] hey werdna [20:44:54] hey howief just wanted to let you know i had a very long conversation with one of our former research subjects about flagged revs [20:45:06] mostly because i think he was excited to get someone from wikimedia on the phone and was a chatter [20:45:15] cool [20:45:15] but i have his contact information if that can in any way help [20:45:21] yeah definitely [20:45:34] he's waiting on a gift certificate from our study [20:45:38] do you think he'd mind if i gave him a call? [20:45:39] howief: I gave you some names to spot check, to make sure I've picked up what you're after [20:45:42] so probably best to wait until after he gets that to contact hi [20:45:47] feel free to spot check as many or as few as you like [20:45:49] don't want to milk him too much [20:45:49] ok [20:45:50] ;) [20:45:54] true [20:45:58] thanks! [20:46:10] werdna: i'll take a look at the names [20:46:14] thanks for putting that together [20:46:24] do you think we can do a first batch of say 10,000 at random? [20:46:31] sure, that can be done [20:46:34] awesome [20:46:45] i'm waiting for some final feedback on the outbound email [20:46:55] but should have everything to you by late tonight/early tomorrow [20:46:55] sure, I'm in no rush [20:47:05] although perhaps you are :) [20:47:10] haha [20:47:21] some people here are anxious to get it out :) [20:47:35] so all you'll need from me then is the outbound email (with link to the survey), correct? [20:47:53] I will probably do some automated spot-checking to make sure the query didn't break or anything [20:48:03] yes, that is enough [20:48:09] cool [20:48:12] thanks again for all your help [20:48:17] I will give you a list of all the users that I send it to [20:48:22] cool [20:48:26] perhaps we could add the username to the URL? [20:48:59] is there any privacy issue with that? [20:49:08] would that be helpful? [20:49:14] I don't think so [20:49:24] Ask Mike for confirmation if you like [20:49:25] well [20:49:40] it would be helpful if the survey could somehow make use of the information [20:49:45] but I don't see that it's a release of personally-identifiable information [20:49:47] and i'm not sure limesurvey can [20:49:58] you could ask for the user's username in the survey [20:50:50] yeah [20:50:56] that might be a better way to handle it [20:51:24] I wonder if you could prefill the field through a URL parameter [20:51:44] you could then remove some of the questions if they're redundant to analysing the user's contribution history [20:51:50] we might be able to, but i don't think i'm going to have the time this afternoon to figure out how to do that with limesurvey :( [20:52:10] nod [20:53:08] do you want me to restrict the people by date of last edit? [20:53:15] how so? [20:53:21] e.g. "We want to ask you about your editing experiences in 2003" is kinda useless [20:53:31] ah [20:53:41] that's a good call [20:53:49] so maybe we should add a lower bound on last edit [20:54:01] agreed [20:54:22] how about last edit within the past 3 years? [20:54:29] 2006 [20:54:54] I would personally put it at this year [20:55:03] lots has changed since 2006, speaking as a contributor since 2005 [20:55:18] unless you're going for diverse data [20:55:28] e.g. how much has people's experience with Wikipedia changed since 2005? [20:55:30] well, if we put it at this year, then we only have 9 months worth of users [20:55:36] but i see your point. [20:55:58] but really, IMO people who last edited over a year ago probably don't remember much about it [20:56:05] ok [20:56:12] well we can always do another batch [20:56:31] so yeah, let's just do this past year [20:56:40] if we don't get a good response rate, then we'll increase the time period [20:56:44] I don't mean to railroad you into a decision [20:56:48] I'd like to know what you think :) [20:56:52] no -- it's a good point [20:57:21] i just want to make sure we get enough responses for this to be meaningful [20:58:12] i'm expecting repsonse rates to be on the low end since we're targeting users who by definition have low activity [20:58:16] but if we do this in batches, there's no reason not to do what you're suggesting [20:59:02] hold on, gotta go have dinner [20:59:10] I'm doing some analysis on the distribution of users by last edit [20:59:21] will tell you what I find [20:59:36] interesting [21:00:00] i may want to follow up with you after this project about user editing analysis :)