Wikipedia talk:New pages patrol/Survey

== Survey needs fixing==

Question 15 needs fixing - if you select "other" which is a text entry, you cannot progress unless you also select another answer - this will warp the results. --Cameron Scott (talk) 12:23, 25 October 2011 (UTC)
 * I have the same problem, with both questions 15 and 20. I would assume it applies every place where "Other" is an option. I have aborted entry of my survey and will try again if this is fixed. --After Midnight 0001 13:49, 25 October 2011 (UTC)
 * Problem with number 14 for me in using the "other choice." Also, I got to the end of the survey and clicked "Done" and it seemed to reset all the fields rather than end the survey, as everything became blank and the survey did not progress.  Big problem!   Blue Rasberry    (talk)   13:53, 25 October 2011 (UTC)
 * Also - could the survey team please link to their Wikimedia accounts? There are names on the page but no usernames.  Thanks.   Blue Rasberry    (talk)   13:55, 25 October 2011 (UTC)
 * I'll try to fix those problems now. Okeyes (WMF) (talk) 15:23, 25 October 2011 (UTC)
 * Okay, the bugs should be fixed (sorry about that) and usernames are now provided. Okeyes (WMF) (talk) 15:26, 25 October 2011 (UTC)
 * Strange, everything worked perfectly  when I  did the pre-release test. Could it  be a browser issue? Kudpung กุดผึ้ง (talk) 15:44, 25 October 2011 (UTC)

Team names have been linked, but  apart from  bugs that  can be reported here (we have two  time zone shifts  watching this page), all  feedback  on  the poll  should be addressed according  to  the instructions on  the poll. --Kudpung กุดผึ้ง (talk) 15:44, 25 October 2011 (UTC)
 * No - I'm not willing to give you my email address, you either take it here or nowhere. --Cameron Scott (talk) 15:46, 25 October 2011 (UTC)
 * Then just don't fill that entry in :). Okeyes (WMF) (talk) 15:48, 25 October 2011 (UTC)


 * Then how would you get the meta feedback about the problems with the survey? As it currently stands, you can do the survey without providing an email but if you want to alert someone of a problem, you have to use your email address and email them. If only we had access to some sort of software where people could collaborate... we could call it Colloboratipedia! --Cameron Scott (talk) 15:50, 25 October 2011 (UTC)
 * You're welcome to drop Howie or myself a talkpage note too - our usernames are linked on this page - the email address is just there for ease of reference. Okeyes (WMF) (talk) 16:19, 25 October 2011 (UTC)

Does it apply to English Wikipedia only? That isn't clear. And rew D alby 19:51, 25 October 2011 (UTC)
 * It does; I'll leave a clarification on the survey proper. Okeyes (WMF) (talk) 20:03, 25 October 2011 (UTC)
 * Clarification left :). Okeyes (WMF) (talk) 20:04, 25 October 2011 (UTC)

Started completing, but I don't intend giving out, for example, even a false age. This is the same problem as with your previous surveys. Unless you are prepared to include a "don't wish to say" option against all the questions then include me out. Jan 1 naD (talk • contrib) 20:20, 25 October 2011 (UTC)
 * That's fair enough. Age is, however, one of the crucial aspects of the survey; since it aims to (a) gather data on new page patrollers and (b) see if the assumptions people have about patrollers are true, we'd be remiss not to specify age. One of the common beliefs, from which a lot of others stem, is that new page patrollers are all fairly young. Okeyes (WMF) (talk) 22:07, 25 October 2011 (UTC)
 * Provided you decline the offer to  be included in  the mailing  list, the poll is completely  anonymous. zAs Okeyes has said, a rough estimate of age is one of the key points of this survey, and age/maturity  appears to  be the major issue currently  concerning  new page patrolling. We  hope through this survey to  tailor the design of the new tools to  the requirements of less experienced users, and the information  received will  also help  the community  to  decide whether or not  NPP  should be a user right. Kudpung กุดผึ้ง (talk) 02:37, 26 October 2011 (UTC)
 * You can say this as much as you like, but if a reasonable number of users fail to complete the survey because they find it intrusive, that in itself devalues any results that may be claimed from the survey. Jan 1 naD  (talk • contrib) 10:04, 26 October 2011 (UTC)
 * This is true. However, we're currently standing at 469 responses, having only messaged around half of the users on our list - we were only expecting 450 responses to start with! I appreciate your concerns, and as said, an in-domain replacement is being worked on, but I don't think the "intrusiveness" of the survey is a serious concern to large numbers of users. The identifying data will all be stripped before any public release anyway. Okeyes (WMF) (talk) 10:07, 26 October 2011 (UTC)
 * (If there is a public release, of course). Okeyes (WMF) (talk) 10:08, 26 October 2011 (UTC)

Q10, "What user rights do you have on the English Wikipedia?" needs fixing too. It requires an answer, but does not include an answer for "none". jonkerz♠ 06:50, 26 October 2011 (UTC)
 * Now fixed; thanks for bringing it to my attention :). Okeyes (WMF) (talk) 10:07, 26 October 2011 (UTC)

Demographics
I'm curious what my gender, age, and location on this planet have to do with my experiences as a new-page patroller? Moreover, I'm curious as to why those make up the required questions, but the lesser identifying items such as when I edit are not required. JPG-GR (talk) 05:50, 26 October 2011 (UTC)
 * These are among the most  crucial  questions. NPP is a magnet for the least  mature and least  experienced of users. The developers of new tools expect to  be able to  address these issues from  the returned information. Location  is important  because there are times when the USA (for example) is asleep (such  as now), and copyvios, attack pages, spam, and blatant  vandalism  are arriving  from  other parts of the English (or semi-English)  speaking  world.  I'm  curious why  you  asked this question  on  the bug  reporting  page. I  hope this helps. --Kudpung กุดผึ้ง (talk) 07:40, 26 October 2011 (UTC)
 * As said above - the basic identifying information is centred around one of the most common myths around New Page Patrollers - that they're 14 year old boys editing from High School ;). Since the objective of the survey is to find out if some of our assumptions need to be re-evaluated before we go ahead and finish designing the new Special:NewPages interface, it's important that was we find out that sort of info. The rest are non-mandatory not because we don't want to know the answers, but because we tried to have as few mandatory questions as possible. I'd actually contest Kudpung's statement - we're finding that users are neither particularly young nor particularly inexperienced. Okeyes (WMF) (talk) 07:42, 26 October 2011 (UTC)
 * I took part in one of the early drafts of this. The location bit is partly in there because I want to find out how our patrolling resources vary around the clock and the week. Are they broadly in line with the flow of new articles, or is there a great imbalance with times when we are falling over each other and other times when stuff slips through. I'm fairly sure that we are in the latter situation, but it would be nice to check and have evidence for that. The sorts of changes we need could be different in the two scenarios, it would also be useful if we could tell patrollers whose experience comes from patrolling at either our most undermanned or overmanned times that their experience is not the same as if they were in a different timezone. Even if all we could say was "don't worry if you can't keep up at that time of day,its when we have our highest ratio of new articles to patrollers. A few hours later our patrollers in x will log in and catch up".  Ϣere Spiel  Chequers  18:04, 26 October 2011 (UTC)


 * Empirical experience tends to  show that  while it  is somewhat quieter when theUSA  is asleep, there are nevertheless many  articles arriving  from  my  opposite end of the globe in  Asia, and fewer patrolers,  and even fewer admins, hence it's quite lonely  out  here. The demographics are vital  to  this survey, and it  remains to  be seen if theories will be confirmed. — Preceding unsigned comment added by Kudpung (talk • contribs) 18:49, 26 October 2011 (UTC)


 * I'm willing to bet that the majority of NPPers will be male, age 18-30, have at least undergraduate education, be single and not have any children, just like the majority of Wikipedia editors are. I wish the survey had been more precise as to what age people are in, I'd really rather narrow the age range down -- is it late 20's or early 20's who are more likely to edit, etc.


 * Just for the record, I did not include any permissions other than my adminship in my response. I would think the reason is fairly obvious - if I was to mention that I was also a Checkuser and Oversighter, it would be incredibly simple to identify my responses.  Given the very small pool of users with one or the other or both of these permissions, I'd suggest they not be included in the future.  Risker (talk) 23:41, 9 November 2011 (UTC)

Possible Problem
I just completed the survey but when I press the Done button it just returns to the top of the page. I haven't missed any questions and I was forced to press the Exit this Survey button. Should I take the survey again? Mo ainm ~Talk  09:20, 26 October 2011 (UTC)
 * No need; it saves your data, but for some reason the survey program occasionally chokes on the "done" button. We're taking all issues into account when we review the data (and what we'll do for surveys in the future) but we can't fix it at the moment. I have jotted off an email to the company that maintains the service. Okeyes (WMF) (talk) 09:25, 26 October 2011 (UTC)
 * Grand thanks. Mo ainm  ~Talk  09:27, 26 October 2011 (UTC)
 * And thank you for filling out the survey! Your answers are genuinely appreciated :). Okeyes (WMF) (talk) 09:29, 26 October 2011 (UTC)

Username errors in invitations
The bot has sent the invitations to several users who don't exist or have never edited any pages. For example, User talk:SkÃ¤pperÃ¶d (instead of User talk:Skäpperöd). The error seems to be occurring with usernames that contain non-English letters or the "@" symbol, or have a space followed by a number at the end, and a few others (User talk:Snek01 673, User talk:User talk:Darthedit and User talk:NoisyJiÒnx). The earlier invitations seem to have skipped the incorrect usernames, but they have not been sent to the correct talk pages. At least one (User talk:RRuk/Intro) has gone to a subpage but not to the user's main talk page. Most of the errors are also in the list at New pages patrol/patrollers (is this the source of the list?). It also looks like users who are blocked or have been blocked recently have not received the invitation (was this intended?) Peter E. James (talk) 20:09, 26 October 2011 (UTC)
 * Both questions are true, yup :). I'm going to go through the contribs of the bot in a tick, delete any misplaced messages and replace them with actual messages on the relevant users' talkpages - just working on summat else at the moment. Thanks for bringing it up and reminding me, though! Okeyes (WMF) (talk) 20:15, 26 October 2011 (UTC)

Peter, thank you for this excellent feedback. New pages patrol/patrollers provided roughly half the targets. and was compiled by a bot. The entire target list  was checked several  times both  with  regex and manually, but  with  this number of items on  a list, some errors will  of course occur. from A-  D, the messages were delivered by  User:MessageDeliveryBot, several  hundred more were delivered semi-automatically  by  AWB, and the remainder by User:EdwardsBot (probably  one of the best)  that  was designed by  User:MZMcBride. Blocked users were deliberately filtered out. Although there are several  reason  for blocking, due to  the fact  that many of them  are probably for socking  and vandalism, it  was thought  more expedient  to  leave them  out  of the survey rather than expose the poll to  possible silly  responses that  would contaminate the accuracy  of the data. About 40 such  user names were removed. --Kudpung กุดผึ้ง (talk) 02:11, 27 October 2011 (UTC)
 * Righto; now all fixed :). Okeyes (WMF) (talk) 02:46, 27 October 2011 (UTC)


 * Although it says "done" for me, I don't see where/when the invitation ever came ... ( talk→  BWilkins   ←track ) 12:07, 5 November 2011 (UTC)
 * Sent. Kudpung กุดผึ้ง (talk) 01:31, 7 November 2011 (UTC)

Appearance of the survey invitation
Long version: I was excited for a moment when I saw the survey invitation -- I thought someone had given me a barnstar. Then I read the message and found that the neat star was just there to look neat and didn't mean anything. In the future, let's just make the post look more like a normal post instead of dressing it up for no apparent reason. If the reason for the "shiny" is, "Well, we have to draw attention to the post to make sure that people see it..." then we run the risk of multiple groups starting to use the same reason since everyone wants "their" message to be read and the resulting war of escalation will eventually make our Wikipedia user talk pages look very much like a 12-year-old's MySpace page with glitter and spangles and animated gifs and shiny things everywhere. I'm not saying "don't give awards", don't misunderstand what I'm trying to say, I just think that "shinies" should only be used when there's a real and valid purpose for the shinies other than "We want to draw attention to this post." Short version: If the reason for a "shiny" on a post is to draw attention to it, then we risk a war of escalation as multiple groups try to draw attention to their posts, making our user talk pages look like a 12-year-old's MySpace page. Banaticus (talk) 23:24, 9 November 2011 (UTC)
 * Thank you. The point has already  been made and taken above. And FWIW, almost  immediately  after the invitations were sent  out, I  realised that  perhaps I  should have used a different  image. Nevertheless, I  do  not  believe that  great  damage to  the project  has been made by  my  minor faux pas - this is a major survey, and if it  has helped to  draw attention  to it, so  much  the better.  Let us not  forget  however, that  like all static  'non  message' messages, it  can, like talkbacks, project  spam,  and the Signpost, be removed uncontentiously when it  has done its job, a bit  like taking  the decorations of the Christmas tree and and getting  rid of it, leaving  the living room  uncluttered once again. You  nevertheless raise a peripheral point: a great  many  user pages look very much like a 12-year-old's MySpace page with glitter and spangles and animated gifs and shiny things everywhere, and Wikilove banners also do  little to  help  our younger editors to  understand that  Wikipedia is not  MySpace,  and even less to  convince our mature readers and users that  they  have come to  a serious project that  should have, above all,  a reputation  for quality. In  pre Wiki  times, I  used to  buy  the Britannica, but  I  probably  would not  have done so  if its presentation  had lent the impression  that  it  may  have been created by  kids. --Kudpung กุดผึ้ง (talk) 23:54, 9 November 2011 (UTC)
 * "this is a major survey, and if it has helped to  draw attention  to it, so  much  the better"  Yes, but everyone thinks that their message is important, that if attention is drawn to it then so much the better.  If people and projects don't think their messages are important, they usually don't bother posting the messages.  Also, the survey invitation template was apparently substituted, so unless somene wants to create a bot to dig around through userspace to find all versions of this message in any possible archives, that NPP barnstar's going to be displayed forever.
 * Two things: you  may  not  know the difference between a template, transclusion, and substitution; and you  are perfectly  free to  remove anything  from  your talk  page except  a block  notice, but  messages should be signed ;) Kudpung กุดผึ้ง (talk) 23:44, 29 November 2011 (UTC)
 * I'd looked at it in the edit pane and it was a table -- neither a transclusion nor a template. I figured it had been substituted when posted.  I know I could've removed it (and probably would've when I archive my talk page in a month), I just don't want the next survey in a couple years to have another star and colored background and whatever else people think it'll take to make it flashy enough that people notice it. :p  Anyway, thanks for taking the liberty, Kudpung. :) Banaticus (talk) 00:09, 30 November 2011 (UTC)

Survey results
Hey guys! The results of this survey should be out within the next few days; thank you everyone who participated. The data will be used to further investigate how we can help improve new page patrol on a technical front - hopefully it'll turn it from a job everyone hates to, at least, a job everyone can tolerate :). Okeyes (WMF) (talk) 01:45, 19 November 2011 (UTC)
 * Okay; looks like we're waiting on legal. Terribly sorry about this, guys - we've got everything written up, but just need to confirm whether we can release the full dataset/under what restrictions/etc, etc. Okeyes (WMF) (talk) 21:26, 24 November 2011 (UTC)
 * I don't see what's taking so long -- as long as it's released with all names hashed to something completely unrecognizable (like turning Banaticus into, say, g*&^uklg786LU), it shouldn't really be a problem, right? But then I'm not a lawyer. :) Banaticus (talk) 22:54, 29 November 2011 (UTC)
 * You should worry -  even the organiser has not  been made privy to the results ;) --Kudpung กุดผึ้ง (talk) 23:48, 29 November 2011 (UTC)
 * Basically it's a question of "whether the clause we added to the survey actually gives us permission to release the data". Now, you'd think it would, but it needs clarification - I'm poking the lawyers again to find out what the heck is going on. Okeyes (WMF) (talk) 10:51, 30 November 2011 (UTC)

Research project
As this is an official WMF research project, please refer future enquiries to  discussions at the offical  project  page. A full report  will  be published when the raw results of the survey  and additional  data have been analyzed by  professional statisticians. This NPP talk  page will  shortly  be archived. --Kudpung กุดผึ้ง (talk) 04:06, 1 December 2011 (UTC)
 * I am sad to note that - after reading that page - this seems to be the work of a groups of unskilled amateurs. And I am not pointing at you, the proponent, as it was indeed a good idea, but looks like a mess of a implementation. - Nabla (talk) 03:35, 8 December 2011 (UTC)
 * Please feel free to post your concerns at the linked page. Discussion here is now closed and I have no longer have personal influence over this project. All I do, and can do, is update the pages with any information that comes my way.--Kudpung กุดผึ้ง (talk) 10:07, 8 December 2011 (UTC)
 * I did. Thank you. - Nabla (talk) 16:32, 8 December 2011 (UTC)