Research talk:Wikipedia Editors Survey 2011 April

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

Logged in Users and Gender Gap[edit]

This is only a survey of a subsection of wikipedia contributors since it was only available to logged-in users. There has been much publicized discussion as to who contributes to wikipedia. Media outlets such as opinion writers for NYTimes and Huffington Post used conclusions from these surveys to discuss a gender bias in wikipedia content. It would be interesting to see the difference in demographics between contributors that log in and those that do not. A more serious study need to identify if the gender gap is related to all wikipedia contributors (as is being assumed by the news outlets) or just those who actively engage in the wikipedia community. It is important to determine if wikipedia is only a consensus encyclopedia of males or if it is a true consensus. Does wikipedia screw articles to favor male rather than female understanding? If so, the wikipedia foundation needs to actively recruit and retain female contributors. Articles that may not be of importance to the average potential male contributor may be of vital importance to the average potential female contributor. A gender gap would ensure deletion of much of this content, which might cause the formation of a cycle of alienating both potential female readership and contributors from wikipedia.

Great points all, but a study of non logged-in users presents several complications. For instance, I am not aware of a way of presenting a survey banner to non-logged in editors without presenting it to readers as well. Definitely something worth exploring though. Akhanna 00:43, 21 January 2012 (UTC)Reply[reply]
I'm a little confused about the description of the survey as limited to logged-in users given that 2% of the survey respondents reported being unregistered users (Q3). How is that possible if the survey was only available to logged-in users? Thanks in advance for help understanding this!


Yikes. This needs one. Findings are rather interesting though. sonia 10:04, 30 August 2011 (UTC)Reply[reply]

Why is there nothing about admins?[edit]

Editors aren't going away, they are leveling off, and to the extent they could be construed as declining they will be around for decades. Admins are going to be gone in seven years at their current rate of attrition. Why is there nothing in this survey about administrators? 03:36, 12 October 2011 (UTC)Reply[reply]

This is a great point, and definitely something we will look into for the future. Akhanna 02:42, 20 December 2011 (UTC)Reply[reply]

Survey questions and feedback[edit]

There are what I would consider three serious flaws in the questionnaire for the November survey:

  1. There is no feedback item where the subject can comment on the problems they have experienced with the questionnaire. This makes it unlikely that most of the problems will be reported. I took a while to find this page to do so and have no idea how appropriate it is to use for this purpose, but I dont have an indefinite amount of time available to find out for sure.
Several people have pointed this out. At the start of the page, we included an email address for reporting any issues. In the future, we will also include a link to the Wiki page. Presently, we are collecting survey feedback at Research_talk:Wikipedia_Editors_Survey_November_2011.
  1. Several of the sections insisted on a fixed number of answers when there were not enough applicable answers to choose from. As there was no option, I selected the least bad answers, but this will skew the results, making the validity of the whole survey questionable. The available answers must be selected so that there are always valid options.
This issue was limited to Q23/ Q24, and Q17. We will redesign Q23/24 for the next iteration, though Q17 needs to stay the same for longitudnal analysis.
  1. Many questions had black/white answers for grey areas, and there were hardly ever options of "dont know" or "insufficient information to make a choice"
We'll look into this for the next iteration, but it would be helpful if you could tell me which specific questions had this issue. The questionnaire is available at Research:Wikipedia_Editors_Survey_November_2011/en.
  • The impression I get is that the people who set up the survey either think they do not make mistakes, they do not care about accuracy, or they have not done the job properly. None of these possibilities fill me with confidence.
The survey has been the result of hours of work by WMF employees, as well as people from the community who have helped us with suggestions as well as with translating the survey in 16 languages. On several measures, including performance and response rates, this survey has done a lot better than the previous one. We have internally tested the survey as much as we can, while working under deadlines. Some errors are inevitable, but I can assure you that this has been a learning process, and the next iteration will be better for it. I have personally put in a considerable amount of time into answering as many queries as I can. Trying to capture the essence of the global Wikipedia community is no small undertaking, and we can always use constructive advice.
  • I would recommend on all future surveys to put in a box at the end where the person who has taken the time and trouble to take part in the survey can comment on things like this, or at the very least provide a link or email address to send comments.
Noted above.
  • It would be nice to get some feedback on this even if just to indicate that it has gained the attention of an appropriate person. Cheers, Pbsouthwood 06:33, 15 December 2011 (UTC)Reply[reply]
Acknowledged. As I mentioned above, there is a dedicated feedback email and Wiki page. In the future, we will advertise both. Thanks. Akhanna 04:19, 20 December 2011 (UTC)Reply[reply]

Survey population[edit]

I would like the report to briefly describe the target population and survey population. I believe the data came from Research:Wikipedia Editors Survey 2011, or did it also include Research:Wikipedia Editors Survey November 2011? I assume that the target population was all editors on all Wikipedia projects (survey was announced via CentralNotice to Wikipedias), but which period was the notice deployed and which period was the survey open, and what fraction of the target population made it into the survey population. In what ways was the survey population uneven. e.g. did English Wikipedia have statistically more survey respondents than German Wikipedia? John Vandenberg (talk) 03:06, 29 August 2012 (UTC)Reply[reply]


I'd like to propose more detailed responses to "Do you have children?" in future surveys. Here's my first-draft suggestion:

Do you have children?
  • No, I am too young to have children.
  • No, I am an adult, but I don't have any children.
  • Yes, I have children, stepchildren, or grandchildren living primarily or entirely with me.
  • Yes, I have children, but they mostly or entirely live somewhere else.
How old are the children, stepchildren, or grandchildren living primarily or entirely with you? (Choose all that apply.)
  • Babies or very young children
  • Children old enough to go to school
  • Teenagers
  • Young adults
  • Older adults

What I'd like to know is how few of our editors are active while having young children in the house. "I have children" doesn't tell you whether that's "I have children, and I edit when I can, which usually means for 45 minutes during his afternoon nap" or "I have children, but this has no effect on my editing except when the grandkids come see Grandma on Christmas Day". WhatamIdoing (talk) 23:20, 27 October 2012 (UTC)Reply[reply]

Next survey[edit]

When will the next Wikipedia survey happen? 13:54, 11 January 2013 (UTC)Reply[reply]

Actually, two subsequent ones have already happened since this (the April 2011 one). One ran in November/December 2011, and the most recent one on October/November 2012, for which we are currently preparing the publication of the results. Regards, Tbayer (WMF) (talk) 18:54, 11 January 2013 (UTC)Reply[reply]

Results on income question (D6)[edit]

The questionnaire includes a question on monthly income (D6) but the dataset doesn't include it. Columns jump from D5b to D7b. What did happen to this question? Thank you. --MarioGom (talk) 20:02, 4 March 2018 (UTC)Reply[reply]