This is the grant proposal referenced at the end of that article (currently under review as part of Inspire): https://meta.wikimedia.org/wiki/Grants:IdeaLab/More_Female_Architects_on_Wi… On Thu, Apr 9, 2015 at 6:31 PM, J Hayes &lt;slowking4(a)gmail.com&gt; wrote: -- Siko Bouterse Director of Community Resources Wikimedia Foundation, Inc. sbouterse(a)wikimedia.org *Imagine a world in which every single human being can freely share in the sum of all knowledge. * *Donate <https://donate.wikimedia.org> or click the "edit" button today, and help us make it a reality!*

Can anyone point to where this "troll" behavior happened? There don't seem to be a lot of specifics in this article, and I'm wondering if it's gender trolls (which are, alas, plentiful) or a culture clash between old editors and new ones over unfamiliar policies? On Sat, Apr 11, 2015 at 3:39 PM, Carol Moore dc &lt;carolmooredc(a)verizon.net&gt; wrote:

One can always just study the relevant articles. But often it's a double standard in application of policies. So if it's a guy architect with a couple low quality refs, people won't even bother to notice or respond. But if it's a woman architect with 7 or 8 solid ones, it becomes a cause celebre to delete the article. And none of that "give the women a chance to beef it up" nonsense either. It tends to be quite irrational and knee jerk. I've seen the same thing on articles about writers, professors, politicians, anyone with even a mild POV that goes against the alleged mainstream. Their articles sometimes are ruthlessly attacked and nitpicked. But if you just put a tag for better references (or any references at all!) on articles about individuals with an allegedly more mainstream view who editors merely claim are important in their field, you may get a lot of grief. That's what systemic bias is all about it. On 4/11/2015 5:55 PM, Rob wrote:

Inclusionism and deletionism is a longstanding battleground where the community is awfully inconsistent. I decline a fair few incorrect speedy deletion tags, some of them so egregious it is very hard to assume good faith and not treat the tagger as a vandal. I don't know whether there is a pattern of articles on women being more likely to be targeted by deletionists, or whether this is a matter of perspective, you know about the articles that you care about that are deleted and you see articles that you don't care about that have survived. What you are less likely to know about are the articles that you don't care about and that have been deleted. If I'm right then there is a common misperception that ones own particular area is sometimes judged to a higher standard. But this would be an interesting area for a couple of studies. Firstly looking at gender ratios of deleted and undeleted bios to see if there is an overall gender skew. Secondly look at the deleters and deletion taggers to see which ones have gender skews in their deletionism. Of course sometimes there will be clear reasons why there is a gender skew, I'd expect the editor who keeps an eye on the category "mixed martial artists" will mostly be tagging blokes for deletion. I'd also expect that the editor who monitors the model category will disproportionately be tagging women. But if we have deletionists who are disproportionately targeting women for no discernible reason then it would be good to identify them. Regards Jonathan

On 04/12/2015 03:38 PM, Emilio J. Rodríguez-Posada wrote: I have python code that's pretty good at guessing the gender of biographical subjects, but originally it scraped HTML given a list of names. If someone had some code for retrieving the wikitext and determining that it is a biography (neither of which would be hard) it would be very easy to determine. Here's some pseudocode: ``` #!/usr/bin/python2.7 def is_bio(article): '''TRUE if article is not '{{hsdis}}' and has '{{infobox person}}''' for title in titles: males = females = unknowns = 0 if title_exists: article = get_wiki(title) if is_bio(article): gender = guess_gender(article) print('%s: %s' %(title, gender)) if gender = male: males += 1 else gender = female females += 1 else: unknowns += 1 print('males = %s; females = %s; unknowns = %s' %(males, females , unknowns)) ```

Actually I think it would be useful to measure all existing female bios vs all existing male bios for the proportion of those which have been previously deleted and recreated. I have a theory that it is much more difficult to create bios of females in whatever category due to the systemic academic bias aginst including women's biographies in the list of "reliable sources" mostly used in Wikipedia. I would be especially interested in comparison of male-female ration of bios in established dictionaries of biography and how these compare to Wikipedia, and of those, how many such bios were previously deleted on Wikipedia and recreated. On Mon, Apr 13, 2015 at 7:02 PM, WereSpielChequers < werespielchequers(a)gmail.com&gt; wrote:

Interesting, thanks for the links! We also now have mix-n-match and Charles Matthews has matched the complete Oxford Dictionary of National Biography, With autolist I could probably look at those male-female ratios per occupation. Might be interesting. I don't know how to get at the "deleted & recreated" data though On Mon, Apr 13, 2015 at 8:01 PM, Joseph Reagle &lt;joseph.2011(a)reagle.org&gt; wrote:

maybe a note to leigh themadattar User:Thelmadatter? who was mentioned in a blog https://blog.wikimedia.org/2015/04/13/wiki-learning-edit-a-thon-mexico/ good reason to go to wikimania On Fri, Apr 17, 2015 at 9:09 AM, Neotarf &lt;neotarf(a)gmail.com&gt; wrote:

Emilio - thanks for the reminder of that excellent page! On Mon, Apr 13, 2015 at 1:18 PM, Jane Darnell &lt;jane023(a)gmail.com&gt; wrote: Agreed. I think one of the most effective ways to counter this sort of systemic bias is to find dictionaries of biography & encyclopedic histories of women and digitize them / make them available to editors. Those sources often do exist, though they are less commonly known or available online. We have a decent reason for them to relicense those works, especially if we can more actively help with the digitization as a result. It would be a small but precise blow against systemic bias to say "the following areas have historical & reporting bias; so we make extra effort to find and recognize additional sources, and vary criteria in inverse proportion to that bias". Sam

I think everyone's idea is good. You can tackle these problems from a "big data" perspective as Joseph is proposing which has the disadvantage of being possibly imprecise in its assessment of gender but can potentially scan the whole of Wikipedia providing a massive dataset. And you can look at "small data" like existing lists as others are suggesting; with this approach, the assessment of gender will be more accurate but the sample size will be quite small (inviting the question whether a larger sample would reveal something different). There is a lot of benefit of doing both. Big data (a highly quantitative approach) is good at testing for unusual patterns in the data but not so good in explaining them; doing a more qualitative approach with small data analysis is likely to yield explanations. If small data yields a theory on the more precise characterisation of a pattern, then the theory can be tested on the big data to see if appears to hold more widely. The two approaches complement and reinforce one another. Neither approach is better or worse than the other, simply different tools used in research. Studying small samples requires skills that most researchers have, which is why it tends to be more popular. Big data approaches need firstly a large dataset (and Wikipedia is a very large dataset!) and secondly a strong IT skill set to work with that data, which limits the number of researchers who can undertake that kind of investigation. If Joseph has the skills to tackle a big data investigation, I hope he will tackle it. I look forward to hearing the results from any investigation. Kerry _____ From: gendergap-bounces(a)lists.wikimedia.org [mailto:gendergap-bounces@lists.wikimedia.org] On Behalf Of J Hayes Sent: Tuesday, 14 April 2015 5:56 AM To: Addressing gender equity and exploring ways to increase the participationof women within Wikimedia projects. Subject: Re: [Gendergap] Article: Wikipedia trolls now vs. women architects i also collected some anecdotal data about deletions and speedy deletions of MacArthur Fellows. you could study differential deletions of them or Fellows of the Royal Society. On Mon, Apr 13, 2015 at 2:44 PM, Samuel Klein &lt;meta.sj(a)gmail.com&gt; wrote: Emilio - thanks for the reminder of that excellent page! On Mon, Apr 13, 2015 at 1:18 PM, Jane Darnell &lt;jane023(a)gmail.com&gt; wrote: I have a theory that it is much more difficult to create bios of females in whatever category due to the systemic academic bias aginst including women's biographies in the list of "reliable sources" mostly used in Wikipedia. I would be especially interested in comparison of male-female ration of bios in established dictionaries of biography and how these compare to Wikipedia, and of those, how many such bios were previously deleted on Wikipedia and recreated. Agreed. I think one of the most effective ways to counter this sort of systemic bias is to find dictionaries of biography & encyclopedic histories of women and digitize them / make them available to editors. Those sources often do exist, though they are less commonly known or available online. We have a decent reason for them to relicense those works, especially if we can more actively help with the digitization as a result. It would be a small but precise blow against systemic bias to say "the following areas have historical & reporting bias; so we make extra effort to find and recognize additional sources, and vary criteria in inverse proportion to that bias". Sam _______________________________________________ Gendergap mailing list Gendergap(a)lists.wikimedia.org To manage your subscription preferences, including unsubscribing, please visit: https://lists.wikimedia.org/mailman/listinfo/gendergap