Hi Everyone.

We were involved in several projects trying to frame research questions to be less accidentally racist, colonial, etc. Each of the projects was designed to bring about more equitable outcomes, but their research questions, while well-intentioned, were accidentally putting the onus for changing in inequitable places. 

At We All Count, we like to think of the different variables involved in a project as puzzle pieces that can fit together well, poorly, or not at all. When we design research questions we try to decide which piece to rotate, shift or swap out to see improvement or ‘positive change’. Choosing which piece to study is easy when you ask ‘What’s the most equitable piece to adjust?’.

Read the whole story here.

We've just wrapped up our most recent Foundations of Data Equity Workshops online and the questions, comments and feedback from all the participants have definitely helped us improve our tools, develop some new learning opportunities and find ways to make the Data Equity Framework applicable to your work. Thank you! We're offering the workshop again in June.

Also in today's newsletter, I'm going to share one of the Q&As from the workshop, a link to a valuable new resource on gendered data collection from Morgan Klaus Scheurman, plus some additional new resources and art.


A Question+Answer from the Foundations of Data Equity workshop

In the data visualization world, there is a big push to get the background info into appendices and put the findings up front. We are finding that, when we do this, our clients like the reports and are more likely to use them. However, I'm now starting to question this approach when it comes to equity and wondering what your guidance is in terms of the funding and motivation statements. Should they appear on the back of the front cover or on the contents page? Is an appendix sufficient? I'm sold on the importance of these statements from an equity lens and am now trying to reconcile it with the push to have less narrative to make our reports more user- friendly.

Really awesome question. There isn’t only one answer to this. The answer is that It Depends. It depends on your intention and purpose. I totally get that putting all the contextual and background information into an appendix and simply presenting results is likely to get more people to actually look at the results. However, I would question whether this is the one and only primary goal. If I want my toddler to eat whatever I put in front of her, I might hand her mostly candy because she’s likely to eat it right away. However, if I want her to be healthy, I work to find ways to get her to eat her vegetables too. From my perspective, getting people to consume results without context is not a good holistic solution, even if it might be efficient. We’ve been working with this one quite a lot. We feel strongly that both the Funding Statement and the Motivation Statement need to go wherever the data goes.
However, it doesn’t mean that the full statements need to be at the front of every report. We have found that the blunter and more honest you’re willing to be, the less paragraphs on funding and motivation upfront with the results. For example, we have had success in putting the full-length Funding Statement and full-length Motivation Statement in the Appendix and then adding a box at the very front of the results as shown below that is linked to the statements in the appendix. Granted, this is a bit radical, but it does a very good job of getting the point across without using many paragraphs right upfront. You could modify it to meet your needs.
A huge thank you to Sara O'Keeffe from the CPPR for making this graphic recording of our most recent presentation at the Good Tech Fest. 
Foundations of Data Equity Workshop
Registration is open now.

It’s time to increase equity and ethics in your data projects and help your data products avoid the pitfalls of racism, sexism and all the other kinds of bias. It’s not going to work to grab a couple of tips here and there and expect to see any meaningful results. Our essential workshop gives you all those practical tools, checklists, and resources built into a comprehensive system for changing the way you work with data from beginning to end. 

The details on the format and syllabus are here.

If you are interested in attending and are struggling financially, reach out to me by email and we'll see what we can do to make it happen.

New resource on collecting and analyzing gender data.

In our ongoing quest to get some innovative ideas on how to collect and analyze social identity data, Dr. Alex Hanna pointed me in the direction of some amazing work being done by Morgan Klaus Scheurmen, which you can check out here.  I had the opportunity speak with Morgan this week and can share some very lightly edited parts of our discussion. Morgan has done some very impressive thinking, publishing and building in the area of intersectional notions of identity in marginalized communities' experiences with technology. His current research focuses on conceptions of gender in technology and its implications for safety and bias for transgender individuals. Having the opportunity to meet him was a highlight of my month.
Morgan is a PhD student at University of Colorado Boulder in the Information Science Department. His research is focussed on how gender specifically is represented in technical infrastructures, so that could be social media websites to algorithms. A lot of his work has focused on algorithms, but he’s also done a very recent study about the inclusivity of gender data collection forms.
Heather: So, what are you thinking now about a recommendation for data collection on gender?
Morgan: I think it's hard to come up with a single answer, so we've been talking about maybe making the primary recommendation that the designer of the form or the surveyor, or whatever is asking about gender, think carefully about what information they actually need. In terms of a doctor's office, you'd probably need a lot more very personal information than a social media website.

Heather: Right. That makes a lot of sense. Many things about equity are never one size fits all.
Morgan: We're trying to think through those options and maybe how you can explain that to the people who are filling out the forms so maybe they don't feel so creeped out or whatever. I think there's no one correct answer for every person, so there's always going to be that possibility of someone not feeling represented by the form or the survey unfortunately.

This is just the highlights of some of Morgan’s thinking on topics around gender and data and technology. He has published quite a bit of work on teaching algorithms to “see” gender, how gender automation impacts safety and much more. Enjoy it here and here.

We're looking for equity problems and successes.

If you have a story or idea you want to share, send me a note by replying to this email.

Project for Equity in Data Science
Copyright © 2020 Datassist, All rights reserved.

Did someone forward this email to you? Sign up here to get the next one directly to your inbox.

Want to change how you receive these emails?
You can
update your preferences or unsubscribe from this list.