Grants talk:Programs/Wikimedia Research Fund/Wikidata to Gender Rescue on Wikipedia (WiGeData)

Latest comment: 1 year ago by Bilorv in topic Feedback from — Bilorv (talk)

We ask you to respond to the following questions:

  1. In what ways do you think this research can support you or other members of the Wikimedia communities in the work that you do on the Wikimedia projects?
  2. What advice do you have for the authors to improve their research or the impact of their research? (We encourage you to share with the authors projects or initiatives that you think can benefit from the result of their research. This can help the authors connect their work with ongoing projects in the early stages of their research.)
  3. Please share any other feedback about this proposal that you think the Research Fund Committee should consider below.

Please use Add feedback button below to add your feedback.

Add your feedback

Feedback from Bridges2Information (talk) edit

1. In what ways do you think this research can support you or other members of the Wikimedia communities in the work that you do on the Wikimedia projects?

When discussing Wikipedia and the gender gap, almost all information is about English Wikipedia (although this is rarely mentioned). It is currently impossible to gather information about gender and Catalan bios and editors, because this information is not allowed on Catalan Wikipedia (from my understanding). Therefore, more information about gender and Wikipedia editors and bios is needed to better understand the gap in other languages and communities.

2. What advice do you have for the authors to improve their research or the impact of their research? (We encourage you to share with the authors projects or initiatives that you think can benefit from the result of their research. This can help the authors connect their work with ongoing projects in the early stages of their research.)

I do not have any.

3. Please share any other feedback about this proposal that you think the Research Fund Committee should consider below.

The researhers are well known and have delivered useful research in the past.

--Bridges2Information (talk) 23:07, 18 January 2023 (UTC)Reply

Thanks so much for your comments @Bridges2Information!! Nferranf (talk) 10:06, 20 January 2023 (UTC)Reply

Concerns for the proposal at Wiki Women in Red's talk pages edit

Hello all, Just to point reviewers to this discussion at WikiProject Women in Red, where a number of concerns were raised about this proposal last November. Some, but not all, of which were addressed. I would also point reviewers to this more recent discussion. Lajmmoore (talk) 07:52, 19 January 2023 (UTC)Reply

Thank you,@Lajmmoore, for all your insights and questions. We needed some time to work deeply with the proposal and know how to approach the concerns that you had highlighted and the ones we already had. I hope it is ok with you if I answer in the next section that you have opened. Nferranf (talk) 10:08, 20 January 2023 (UTC)Reply

Questions on leadership, outing, colonial understandings of gender, evaluation of WiGeDi edit

Hello all, First of all let me say I am very passionate about representation and the role Wikimedia can play. However, I posted this concerns as part of a wider discussion about this potential funding stream last November at Women in Red, and received no reply, so am re-posting them here. This falls I think under Question 3 from your list above:

1) I'm not sure the women-led projects you mention are the best to lead on this kind of work - surely this should be led a group like "Queering Wikipedia" for example?

2) I'm concerned about the potential for "outing" people who do not want their gender to be public knowledge. You don't need to provide evidence for gender on Wikidata, which might mean that some people who weren't open about their gender identity might be forced to be.

3) Many of the gender diversities that might be measured are outside western understandings of the binary - some Pasifika genders outright reject the idea of "trans-ness" as its not part of their culture, and I think there really needs to be some consultation made with people outside Europe, perhaps through ESEAP, and other regions, to check whether this kind of measurement would be suitable and even be welcomed?

4) Pressingly, not all gender diversities are recognised in WikiData categories for sex or gender (see Kalisito Biaukula who is vakasalewalewa as an exmaple), so it seems there's more work to be done on Wikidata about this.

5) I went to WP:WikiProject LGBT to see what discussion on this was like there, but there's no post. I think there, and WP:WikiProject Gender Studies would be good places to consult on this. (I see you did this yesterday)

6) Likewise I looked for a similar discussion on Wikidata, and couldn't see one. I really think you need to discuss this project there too.

7) Wikidata Gender Diversity (WiGeDi) that you mention finishes in May, so I'm wondering how you can build on the project when it's not complete?

Thanks very much, I will mention that it is a shame I am having to repeat my personal concerns here though, rather than receiving a reply several months ago on the talk page. Lajmmoore (talk) 08:04, 19 January 2023 (UTC)Reply

Hello @Lajmmoore,
Sorry for our late reply, as we said we needed some time to work in the proposal. But thank you for your questions, feedback, which made us think about our proposal and improve it. We answer to your points one by one:
1) Thank you for raising this concern, which we share with you. As gender studies researchers, we are aware of the ethical relevance of positionality and situated knowledge when we approach any research topic, particularly one that some of us do not experience in first person and involve the representation of gender-based oppressed groups. We considered this fact, and we valued that the proposal is significant because it aims to reduce the gender gap by making visible and improving representation of women and nonbinary people in Wikimedia. We consider we are not taking away space for the LGBT community, but acting from, with and for the women and LGBT community (in a certain way, for our own rights as there are both women and LGBT people in our group). Likewise, we consider this is an issue of feminist and transgender/ nonbinary social justice and will be as honest as possible with our potential biases derived from our positionality. In our broad Women & Wikipedia Research Project we are also collaborating with trans and nonbinary-identified people and making an effort to consider their voices and experiences (we are currently working on qualitative research on women and nonbinary identities experiences around representation in Wikipedia). We started working with women several years ago, and now our efforts will go to nonbinary people.
2) We are concerned about the potential for "outing" people who do not want their gender to be public knowledge. You don't need to provide evidence for gender on Wikidata, which might mean that some people who weren't open about their gender identity might be forced to be.
2) We understand your concern, but we consider that although gender ambiguity or invisibility may protect circumstantially and getting out is a personal decision, the systematic invisibility of gender on Wikipedia implies a bias that damages and harms women and nonbinary people in the short and long term by biasing their representation, preventing them from reaching the knowledge of more readers and therefore reducing their possibility of being a reference/ role-model and having a broader social impact, especially on all those readers and audiences who identify as women and nonbinary people.
Regarding wikipedists, this data will still be voluntary. As it is now, particularly in non-English languages with grammatical gender, the default is non-specified (male in languages with grammatical gender) and you need to change your gender in the profile with three options (non-specified, male or female). If people want to remain silent about their gender identity, it would be an option, but our aim is to better represent those people who do not have a possibility to be visible (their gender is not an available option to choose) or decide not to make their gender visible in order to avoid androcentric bias towards their work in Wikipedia (eg: because articles written by women are much more questioned).
3) Many of the gender diversities that might be measured are outside western understandings of the binary - some Pasifika genders outright reject the idea of "trans-ness" as it is not part of their culture, and I think there really needs to be some consultation made with people outside Europe, perhaps through ESEAP, and other regions, to check whether this kind of measurement would be suitable and even be welcomed?
3) You are right, gender identities are a very complex matter where culture, history and geography enormously condition its understanding, identification, and embodiment. We of course do not want to impose a gender identity or category over people from anywhere, and less from non-western cultures. However, we honestly think that the consideration of nonbinary as a Wikimedia category can work as an umbrella term where further discussions on nonbinary genders and cultural identities could blossom. We see this as a start point category that must be broadened, refined, and improved over time by the nonbinary community.
Regarding the consultation you propose, we agree this is a great idea, we will consider this for our empirical qualitative research and will try our best to collect different views on the issue to be as much culture-sensitive as possible. We would like to say, however, that we do not currently have the capacity to carry out an in-depth consultation study, as we see it would be another research project.
4) Pressingly, not all gender diversities are recognised in WikiData categories for sex or gender (see Kalisito Biaukula who is vakasalewalewa as an exmaple), so it seems there's more work to be done on Wikidata about this.
4) We agree. This matter is much more complex and is in continuous discussion within the nonbinary community. We strongly believe this difference and its cultural and de/post-colonial relevance must be considered, and we see this as a criticism happening within the Global LGBTIA+ community (leaded of course by non-western, BIPOC communities). Our contribution in this case has its limitations, we suggest that the mere existence of the “nonbinary” category on Wikipedia can work as a starting point to this nonconformity to western frames and categories to blossom, express and broaden categories. We do not feel legitimised to deploy a huge list of categories, but we think that this is a preliminary step that will favour a more nuanced gender identification and visibility in Wikipedia. Besides there are research projects such as WiGeDi (Grants:Programs/Wikimedia Research Fund/Wikidata Gender Diversity (WiGeDi)) from @Mushroom which are more focused on these concerns. We are working together as sister projects already, and we think both research projects will provide more inclusive gender representation.
6) As said, this is the main focus of WiGeDi from @Muschroom, and we will share methodology and results that will help both projects.
7) They finish in May23, but we are already doing research on this matters and being in touch with them since November22.
Finally, thanks again for sharing your thoughts and ideas, @Lajmmoore, and also thanks for your patience. We needed some time to work deeply on the proposal and then be able to answer all the comments like yours. We hope that you would understand.
@Lauferagui and @Nferranf (talk) 10:37, 20 January 2023 (UTC)Reply
Hello both, Thank you very much for the detailed response. I would like to point the funding reviewers to answers to points 2 and 3, and to check if I have understood what is inferred correctly?
  • For Point 2 you say "although gender ambiguity or invisibility may protect circumstantially and getting out is a personal decision, the systematic invisibility of gender on Wikipedia implies a bias that damages and harms women and nonbinary people in the short and long term by biasing their representation, preventing them from reaching the knowledge of more readers and therefore reducing their possibility of being a reference/ role-model and having a broader social impact, especially on all those readers and audiences who identify as women and nonbinary people" This comes across as if you think it is OK to out people who aren't open about their gender or sexuality, regardless of their wishes, which I think is fundamentally wrong. I would very much encourage to re-think this position and the potential for harm it might cause. For example, what is your project's ethical position if someone who you list as non-binary lives in a country where their gender is deemed illegal is harmed? It should be a human right for people to decide when and to whom they come out, and to not have a research project decide it for them (however well intentioned that might be)
  • For Point 3, there are two points there that I would resist your explanation of. The first is that "the consideration of nonbinary as a Wikimedia category can work as an umbrella term" fundamentally ignores the importance of culturally specific terminology for people. As westerners, for too long we have decided "umbrella terms" and i really think it is disrespectful to continue to do so without consultation. On the point of consultation you say "do not currently have the capacity to carry out an in-depth consultation study, as we see it would be another research project" - I would argue that it a bare minimum that you should undertake before this kind of work happens. Communities are not research projects, and if you are going to interfere with their representation online then the least you can do is ask if some representatives have feedback.
I would encourage funding decision-makers to press the consortium on these points, and use their connections within the movement to see if they can support the work on point 3. Regarding point 2, I do not think your position is ethical. Lajmmoore (talk) 20:55, 20 January 2023 (UTC)Reply
Hi again, @Lajmmoore, and thanks for sharing your concerns.
For Point 2: It’s worth noting that our proposal does not go along the lines of outing "people who aren't open about their gender or sexuality, regardless of their wishes". On the contrary, it wants to get available and accessible, for Wikipedia searching, properties and values that are already available in Wikidata. By no means, are new and not agreed properties or values.
For Point 3: Along the same lines, our project will not interfere with anyone's online representation; our project will increase values ​​already agreed upon by the Communities to improve the Wikipedia searching system.
Hope this clarifies the purpose of our proposal WiGeData.
@Miquel Centelles and @Nferranf (talk) 10:45, 23 January 2023 (UTC) Nferranf (talk) 09:47, 23 January 2023 (UTC)Reply

Comments by Johnbod edit

I must say I'm still very unclear as to what the project would actually be doing. I was puzzled by "It is currently impossible to gather information about gender and Catalan bios and editors, because this information is not allowed on Catalan Wikipedia (from my understanding)." All the information about mass statistics about biographies on English Wikipedia has long been done by Wikidata queries picking Wikidata objects with a match to an en:wp biography. I don't see why this approach doesn't work on the Catalan or any other language wp. Am I missing something? Johnbod (talk) 04:48, 27 January 2023 (UTC)Reply

Thanks for your observations @Johnbod.
Yes, that it is true. In the Catalan Wikipedia or the Italian Wikipedia it is not possible to access directly to the Wikipedian categories "women" and "non-binary" which so many other Wikipedias allow their readers to do. Also in gendered languages, such as the Italian and Catalan, not having these categories available is even worse as the label used for the category uses the male gender, so you can reach a male label for "nurses" which mainly includes women or the male label "deaths by uterus cancer" which mainly affects to cisgender women.
There were two deliberations and an election in the Catalan Wikipedia, and it was decided not to allow the use of the "women" and "non-binary people" categories. But it was allowed to use a search engine that uses Wikidata to retrieve only the gender identity with its Wikidata values. This search engine has poorly been applied in terms of usability and consistency. So, we see this as an opportunity to explore better how Wikidata can help to make more visible and accessible content, in our case study in particular, different gender identities available in Wikidata which can be applied in Wikipedia content.
We will assess the classification of categories on Wikipedia and ontologies in Wikidata, from heuristic, user tests and automatic tools from a gender perspective. Then we will work on the design of a prototype of navigation and search that will help content to be more accessible and really trying to answer the user needs. Nferranf (talk) 08:57, 27 January 2023 (UTC)Reply
Ok, so you can use Wikidata, just as en:wp does. So I'm still not clear why "It is currently impossible to gather information about gender and Catalan bios and editors" - categories are probably a poor way of doing this. Perhaps you could explain the plan in more detail. Johnbod (talk) 02:43, 29 January 2023 (UTC)Reply
Currently, Wikipedia is using categories to organize its content. These categories are a controlled vocabulary or a folksonomy where any editor can create categories or subcategories if needed. That is in the English, Spanish, whatever Wikipedia. But some Wikipedias, after deliberations and elections, can decide to erase categories and not allow using them in the future, that is the case of study here, the Catalan and Italian which do not allow using "women" and "non-binary" tags. Then, we propose to organize Wikipedia content (navigation and search) using the ontologies of Wikidata and the experiment will be performed with a gender pespective. But in the future, this pilot that we will do in the Catalan Wikipedia can be broadened to other Wikipedias and other sections of the ontologies, so it can organize other types of content.
At the moment, Wikidata is used on Wikipedia for certain services, for example, the Infobox or the category "alive people" is created from Wikidata too. Besides, Wikidata is quite open in terms of gender, and the values of the entity include a greater diversity of gender identities, compared to Wikipedia. In fact, there are several research papers that have confirmed that the gender gap in Wikidata is not as big as Wikipedia because it is closer to reality. So Wikidata has a smaller gender gap as it is mirroring better the public sphere. Nferranf (talk) 15:50, 30 January 2023 (UTC)Reply

Feedback from — Bilorv (talk) edit

1. In what ways do you think this research can support you or other members of the Wikimedia communities in the work that you do on the Wikimedia projects?

I'm confused as to how this proposal is supposed to support me in the work that I do. As a member of WikiProject Women in Red on the English Wikipedia, I'm interested in reducing the gender gap through the creation of content: for me that's encyclopedia articles, but I understand Wikidata volunteers see value in certain types of more abstract structured data. This proposal says it aims to ... [reduce] the gender gap in Wikipedia by visualizing gender diversities collected in Wikidata. Which is it? Does it aim to reduce the gender gap or "visualize it"?

2. What advice do you have for the authors to improve their research or the impact of their research? (We encourage you to share with the authors projects or initiatives that you think can benefit from the result of their research. This can help the authors connect their work with ongoing projects in the early stages of their research.)

I've seen stat after stat about the gender gap, and this has actually been quite widely reported in the media, unlike many other issues in Wikipedia. I would respectfully argue that what we do not need is new ways to categorise gender bias, but more resources to tackle it. A budget of around US$50,000 is proposed. This money could make a huge difference to a particular gender gap if it was spent on expanding the digital resources (journals, magazines, archives) available to volunteers, on physical source material, or on Wikimedians in Residence. Or, the WMF could spend it on making essential bugfixes and technical maintenance work that all volunteers (including WiR members) need to edit.

3. Please share any other feedback about this proposal that you think the Research Fund Committee should consider below.

Bilorv (talk) 21:55, 18 February 2023 (UTC)Reply

Return to "Programs/Wikimedia Research Fund/Wikidata to Gender Rescue on Wikipedia (WiGeData)" page.