SHARINGACTION2019 Data and Gender: the importance of Data with gender perspective in Public Policies 2019/11/20/apunts/01

https://pad.femprocomuns.cat/sharingbcn2019genderdatapolicies

Data and Gender: the importance of Data with gender perspective in Public Policies

Sharing Cities Action Encounter Barcelona - 19-21 November 2019

http://www.sharingcitiesaction.net/encounter-2019/programme/

Collaborative notes at Teixidora.net, documentation to share and reuse.

https://www.teixidora.net/wiki/Sharing_Cities_Action_Encounter_2019

Conference notes

Eli - Hoy en dia, la info en lugar de venirnos con una nota de prensa llega en forma de datos. Para demonstrar que la brecha entre hombre y mujeres no es una leyenda urbana se necesitan datos.

No hay practicamente referentes femeninos en los libros de ciencia.

Los datos dicen que las mujeres siempre pierde, cualquiera sea la dimension.

Catherine - there is a lot of feminist scholarship and theories all around, that we can use for our societal debates: about social justice, ethical debates. All these debates that the feminist movements has brought about, which are extremely critical, can serve us in todays society for data science and technical discourses. Any kind of process of innovation we can't just do work about women as subject but we have to center women and marginalized individuals in our process. It is not just about doing work ABOUT women and marginalized communities but rather do work WITH women and marginalized communities.

Lidia - Why we need to adopt a gender perspective when we produce and publish data? There is a big issue of interpretation: there are a lot of scientific publications that does not publish the gender of the individual, or they provide false interpretation. A gender perspective improve the accuracy and avoid misinterpretation. In health research for example a lot of women deaths by cardiovascular disease could be avoided by doing research providing the sex of the participant sample. Another issue is about reproducibility of the research and data: when we dont publish the sex and gender of the sample we are inconsistent in terms of methodological reporting (how could we reproduce the experiment?), it is also a lack of transparency.

There is an issue on the generalizability of the results.

It offers new perspective and questions about the data. in natural language processing systems it is very evident: those who train algorythm are mainly men and this extremely biases the results.

There is also a difference between having the opportunity to access data (everybody potentially has the opportunity) but are these opportunities equally distributed? does anybody actually access data? It is not only about broadening participation in scientific research but also about the theoretical analytical framework used to collect and analyse the data.

Karma - we did not want to focus on the dramatic dimension of violence against women but rather about accountability: what is the size of the issue? One of the biggest problem when we talk about violence against women is that we dont know how many women are suffering. Because much of the suffering is psychological. And we dont have data about this dimension. Things are happening and violence against women is there, and data is voicing it. The problem: lack of data; but also lack of budget, also differences among data sources related to different criteria to get these data: how can you then analyze this data? And invest money to solve this problem? The professionals who are doing with women need more training, because they themselves say that they are not prepared to deal with these situations. We need to apply the istanbul convention.

Nuria - The regulation that the Generalitat approved last year points at not open data by law but open by default: we would like to open all the private datasets. Open more data, have data wiht higest quality possible (timely, compete and reliable). Quality in data is also about providing some additional info in the dataset, for example, geolocalized information, and about gender: if the dataset has info about gender or sex, it is mandatory to put it in the dataset. Data will help to improve a lot of public policies, as the gender equality policy. The third issue that the strategy is aiming at is that the data are to be useful: the goal is not to have the data useful but that the citizens and public administratora can use them for their own purspose, this is also achieved trhoguh data visualization, to ease data access and understanding and use. It is important to develop indicators which now are not there to measure gender pay gap or violecne against women.

If you dont have evidence it is just an opinion.

Valentina -why we need to open data about gender? Our smart cities are ruled by algorithm, which are trained mainly by men and which can only reproduced structural gender and other social inequalities. We have lost the opportunity to discuss with the city council about data openness accessibility and quality. And for examplke all the data that are there about residency, salary and other sociodemographics are not filtered about gender, this information is very relevant to develop public policies. We asked the politicians to see the city as a women would see it, to shift the discourse from data issue to feminist issues.