Grants:IdeaLab/Survey women who don't contribute
What is the problem you're trying to solve?
We have repeatedly surveyed Wikimedians and asked women within the community about their experiences, but the problem is that we have already attracted and (so far) retained those women, so they don't give us much insight into the women that we're missing.
We currently have no (or few) answers to questions like these:
- What perception do women who don't contribute have of people who do contribute to Wikipedia? Do they identify with this perception?
- What perception do they have of the experience of contributing? What influenced this perception (eg. media articles, accounts from friends, reading on-wiki)
- What reasons do they have for not editing/contributing? (note: partially answered by UNU-MERIT)
- What changes/incentives would help them decide to edit/contribute?
For women who used to edit, but no longer do, we could also ask them:
- What originally motivated them to edit/contribute?
- What was their experience of contributing? What positive aspects did they experience? What negative aspects did they experience?
- What sorts of contribution did they do? Eg. editing pages, discussing on talk pages, participating in wikiprojects, being an admin, etc etc
- What was their perception of the community as a member of it? What positive/negative experiences did they have?
- What made them stop editing?
- Is there anything that could have been done differently, that would have prevented them from leaving?
- Is there anything that could be done now, to bring them back?
- Would they recommend that other women contribute? Why or why not?
Breaking the survey into "never contributed" and "ex contributors" would also help us separate ideas to attract women and in the first place, those to retain women who have already joined us. Both ends of this pipeline are important to increasing the overall number of women contributors, but existing efforts often tend toward the "attraction" stage, or are aimed to help newer contributors, ignoring the problems that may be faced by more experienced editors. (This pattern is mirrored in STEM fields, such as the technology industry.)
Knowing the answers to these questions would help us usefully target other ideas/programs/etc to solve gendergap related problems through both attracting non-contributors, and retaining existing ones.
What is your solution?
A research study, with questions like those above. I think to be really useful, it would be good to have an online survey with multiple choice/short answers, supported by in-depth interviews.
For ex-contributors, we could also ask them for their wiki username, and analyse their actual contributions (eg. number of edits, types of edits, how long were they involved, etc).
Since (by definition) we can't find our survey respondents in our own community, we could seek respondents through:
- Women's technical mailing lists (eg. Systers)
- Women's studies departments at universities
- Women-centric forums (Mumsnet or things like that)
- Mainstream and online media targetted to women (eg. the Mary Sue)
- Social media (especially asking people to signal boost outside Wikimedian circles)
- Project manager Calls for data regarding the Gender Gap go unanswered, because this data is either unavailable or too time consuming to compile. Surveying should go some way. QEDK (talk) 04:43, 6 March 2015 (UTC)
- Can identify and administer questionnaires to potential respondents.→Masssly (talk)
- Volunteer Data analysis, Repondent profiling, questionnaire preparation etc... Ludopedia (talk) 19:50, 28 March 2015 (UTC)
- Advisor I come from a country where a woman's voice is crushed by the male dominated society in the name of...uhm, social manners. Yes, India is quiet modern now! But that my dear sir is only till a girl reaches a certain age. After which every single family in this country thinks of nothing but the girl's (arranged) marriage. In my country, marriage is mostly the one word answer to breaking the woman's high independent dreams. As a result, she has to do away with her aspirations and get involved in the housework. As a woman, I would like to put forward my ideas and opinions on breaking away with this practice. Well, not exactly the practice of marriage, but the practice of sacrificing ones wishes in the name of marriage. I know a lot of people who had to do this sacrifice. But I have a first hand example of my mother, who wanted to do community service and help the needy but was made to marry by my grandparents. My father though allowed her to continue with her dreams and now she is working with the widows of the martyrs.
- Collecting data is always a good first step. I could persuade my female non-editor friends to answer a survey like this. The fact that they are more likely to be culturally similar to me need not be a problem, since we don't need a random sample of non-contributors, but a sample of women (and girls) who could be contributors but aren't. I suggest some more objective questions (Do you know any programming languages? Can you write HTML? How often do you read Wikipedia? What is your highest level of education?). If editors could self-declare as female or male without posting the fact publically, we could compare their editing patterns (How long are they active? What do they edit just before they leave? Do they get posts from specific bots just before they leave?) Actually, given the accuracy of algorithms that guess your data from your browsing history, we could probably make a first stab at this without asking people to self-declare. HLHJ (talk) 22:44, 8 March 2015 (UTC)
- Regarding questions about education and technical experience: this could potentially be a different survey, I think, as you'd want to compare it against 1) women who *do* edit, and 2) men who do/don't edit. You could then use the data to answer questions like "do women who don't edit have less technical proficiency than men who don't edit", or whatever. Just surveying non-editing women wouldn't give us these answers. --Skud (WMF) (talk) 00:18, 10 March 2015 (UTC)
- Too few women? Just ask them. And ask them why not. More potential survey questions:
- Do you read Wikipedia? (How often?)
- Do you like Wikipedia? (How much?)
- Do you know people who edit wikis? (How many?)
- Do you think that wiki editors are doing a good thing? (Image of editors)
- Does your partner edit wikis?
- Do you like it that your partner edits wikis?
- In some partnerships, there may be some "conventional role assignment" going on. Male is hobby Wikipedian, female conciously or unconsciously steers clear of that domain, takes care of things "more important", "less nerdy" or whatever. Rainer (talk) 00:04, 11 March 2015 (UTC)
- Just a note that we should avoid heteronormativity in any questions like this -- not all women have male partners. It might be worthwhile to just include "partner/spouse" as one of the options among "do you know people who edit", and address the gender norm questions separately. --Skud (WMF) (talk) 23:52, 15 March 2015 (UTC)
- Endorse, specially in regard to improving editor retention. Langus-TxT (talk) 03:55, 13 March 2015 (UTC)
- Sounds most useful so far 2001:A60:F000:67:0:0:0:2 07:21, 18 March 2015 (UTC)
- I like this approach. I'd also be interested in particularly surveying women who 'should' edit--insofar as they have many of the matching skills and background compared to current editors... but who *still* do not edit. That would make the sample even more insightful for me. But a broad survey could confirm what prior informal interviews have yielded. Ocaasi (talk) 00:28, 20 March 2015 (UTC)
- I like the idea of conducting the survey the other way round to gain perspective from respondents not usually included in WMF surveys. I Strongly Support this idea!→Masssly (talk) 11:22, 26 March 2015 (UTC)
- Essential idea. +1 to Ocaasi's thoughts and to +Skud's
- I'm surprised we haven't done this kind of survey targeting non-male users before, especially given the gender imbalance has been on-going and broadly discussed. Just as we do user research before product launches, community research before program launches, this research seems very necessary. Shouston (WMF) (talk) 20:24, 27 March 2015 (UTC)
- great idea Ludopedia (talk) 19:49, 28 March 2015 (UTC)
- I think this is a very worthwhile idea. We won't get a rich understanding of editing behavior without examining those who don't edit. Libcub (talk) 22:52, 28 March 2015 (UTC)
For reasons of practicality -- primarily the difficulty in understanding and translating long-form answers -- I plan to conduct this survey in English only.
This will mean that the results will reflect only information about English-speaking non-contributors, and may not be relevant to non-English speaking groups. The results will therefore be of most use with regard to English Wikipedia and other English-language projects.
Because English is widely spoken throughout the world, and answers may vary widely based on cultural background, I intend to ask respondents about their country of residence and primary language spoken. This may help us break down answers in more useful ways.
The original IdeaLab idea suggested two surveys, one for women who have never contributed and one for women who used to contribute but ceased. To clarify: the current proposal is for a single survey of women who do not currently contribute.
The survey would include some questions about whether they have ever contributed, and if they have previously done so, may branch into some different questions depending on their contribution history. However, there is only one survey, and there is no intention to do in-depth research of ex-contributors' edit histories, or anything of that kind, at this time.
(There was a suggestion to only survey people who have never contributed. However, in casual conversations with women who don't contribute, I have found it's hard to make this determination up-front. Many people say they don't contribute, and then, when prompted, remember that they have made minor edits in the past. It seems better to acknowledge this range of non-contribution, especially because people who've tried and then stopped may have useful insights into what didn't appeal to them.)
Depth of investigation
The original IdeaLab idea talked about a survey and/or interviews. To clarify: the current grant proposal is for a survey which may include optional in-depth/long-form answers, however there is no intention to do one-on-one interviews at this time.
My goal is to reach a wide variety of non-contributors from many different backgrounds. To this end I will be putting together a wide-ranging list of outreach channels and seeking assistance to spread the word as widely as possible. For example, some potential outreach channels include:
- Mainstream and online media aimed at women (eg. The Mary Sue)
- Online forums/communities frequented primarily by women (eg. parenting forums, hobby forums, fandom)
- Specialist online forums/mailing lists for "Women in X" or professional fields with a high proportion of women (eg. teachers, librarians)
- Universities/colleges (especially departments with high proportions of women, or through academics who have previously been involved in Wikimedia-related outreach)
- Social media (eg. sharable "viral" content for Facebook etc)
- Individual outreach from current Wikipedians to their own friends and family who don't edit
I would hope to have a total outreach contact list of at least 100 channels (forums, mailing lists, bloggers, etc) from a wide variety of interests/backgrounds, and that by providing them with links to easily sharable "viral" material, they would help spread the word further.
With regard to individual outreach, anecdotally it seems that many Wikipedians have the experience of friends or family members who aren't interested in editing. I would expect that many of these Wikipedians could be enlisted to help recruit respondents from among their own circles.
Note: all time estimates are given as elapsed time.
First half of project
(approx 6 weeks elapsed time)
- Create a survey
- Review existing research (1 week)
- Design questions (1 week)
- Seek review/comments (from research experts and on-wiki for community input) (2 weeks)
- Revise (1 week)
- Develop preliminary list of potential forums/bloggers/etc to contact (0.5 weeks)
- Post on wiki to crowdsource further suggestions (2 weeks)
- Note: this can be done somewhat in parallel with the survey creation work, above
- Create outreach materials for survey (form email, viral image for social media, etc) - 1 week
- Announce survey
Second half of project
(approx 8 weeks elapsed time)
- Collect survey results (over 4 week period)
- Engage wiki community to assist in outreach for survey (ongoing over 4 weeks)
- Analyse results and write report (2 weeks)
- Seek review and feedback on report from research advisors, update accordingly (1 week)
- Write blog posts etc. to inform people of results (1 week)
- Participate in followup discussions on-wiki etc as needed
Lead researcher: $4,000
- Seek community review of survey design questions
- Seek community help in developing initial list of outreach contacts
- Engage community in outreach to find survey respondents (eg. through their non-wikimedia communities, family, friends, etc)
- Report will be of use to future gender gap projects in assessing their likely impact and effectiveness.
Measures of success
- 1000+ respondents (identified as women who don't edit wikipedia)
- 50+ respondents provide in-depth answers
- report published and made available under CC license
- results of study publicised throughout Wikimedia community and to the wider public