0:00:00.2 Sally Laskey: Welcome to Resource On The Go, a podcast from the National Sexual Violence Resource Center on understanding, responding to, and preventing sexual abuse and assault. My name is Sally Laskey and I am the Evaluation Coordinator at the National Sexual Violence Resource Center. Today we are talking with Heather Krause, the founder of We All Count, and a data scientist and statistician with over a decade of experience building tools to support equity and ethics in data.

[music]

0:00:46.0 SL: Thank you so much for joining us today, Heather. Would you be able to give our listeners a brief introduction of who you are? 

0:00:56.3 Heather Krause: Hi, thank you for having me. I'm very excited to be here, and yes, my name is Heather. Heather Krause, and I am a statistician who is the founder of We All Count, which is a project for equity in data. And I'm really looking forward to talking with you today.

0:01:18.3 SL: Well, fantastic. Let's jump right in then. For those that might be new to this topic, could you describe what data equity is? 

0:01:30.7 HK: Sure. So data equity is about being attentive to the world views and power structures that have been centred in all data projects. There's this idea that if it's a number, it doesn't have a world view, or if it's quantitative data, it is inherently objective. And when you really dig into the mathematics and the process by which research is produced, those things are very much not true, and that there are lots of very simple ways to look and see how every data point has someone's world view embedded into it, and that the way that we... The whole scientific process of generating research involves many, many, many choices, and each one of those choices embeds a set of values or a lived experience or a world view.

0:02:39.7 HK: So in order to really use data effectively rather than pretending that it is objective or free of any world view, the way to get to data equity is to be intentional about whose world view and what set of values get embedded at each one of these choice points, and then be transparent about it so that anyone who's looking at our data or making decisions based on our research can understand what world views or sets of values have been embedded in those numbers when they're using them. So that's what data equity is.

0:03:18.3 SL: Excellent. So then what brought you to this work? And why do you think it's valuable for those of us working to prevent and to respond to sexual violence? 

0:03:31.0 HK: Sure, absolutely. It was actually an experience that I had at a conference, a large prestigious international conference on research about intimate partner violence, that was one of the big catalysts for me founding We All Count. And that was looking at a presentation that was being made about the relative rates of violence... Against intimate partner violence. I mean... Yeah, intimate partner violence and comparing different rates from different countries, and it was very clear to me by looking at the chart that it was accidentally comparing all types of data that weren't actually comparing apples to apples if you really looked at whose value system or whose world view had been embedded in that data.

0:04:28.6 HK: And so even understanding what policies work, or what locations or geographies or parts of the world have more or less violence, is impossible to do unless you're paying attention to the equity that's being embedded or not into the data. And so I think it's real... I mean, I could go on for the whole day, much longer than anyone would listen to a podcast, how important each one of those equity-based choices are, specifically in the area of using data to understand sexual violence.

0:05:05.1 SL: Well, it's interesting that you brought up how this experience at a conference led you to do very specific work in this area, because I was able to attend a webinar you presented entitled How To Not Use Data Like a Racist. In that webinar, you talked about studying up the power chain, and it's just... It stuck with me ever since that moment. And can you share what that is, and how it can help us better address racism and other structural issues that increase risk factors for sexual violence? 

0:05:49.0 HK: Yeah. Thank you for asking that question in that way, because I think that that is often under-considered when using data, because we have this idea, this myth that we've been taught that data is objective, we often think that as long as the person is well-intentioned and intelligent and well-resourced, then we can trust the data that they're giving us, and that is not the case, unfortunately, which is really a lack of tools and systems to really think deeply about the equity that is embedded in the data that's missing. And when you talk about studying up the power chain, and how important that is, when thinking about how individuals are not nearly... Individual victims of violence are not usually the source of the problem, but yet we find ourselves crafting research questions and collecting and analysing data implying that they are, without understanding that we are making that assumption embedded into our data.

0:07:05.5 HK: So even if we craft a research question that accidentally puts the locus of change and the expectation of change on to individuals who experience violence, rather than crafting a research question that puts the onus to change on systems like racism, like sexism, like colonialism, like classism, that are actually much more powerful levers of change if we are trying to eradicate sexual violence. And we have, absolutely have all the data and all the data tools and all the statistical resources that we need to ask research questions and build statistical models that put the onus to change on structures, not on victims, and we just have really bad habits of doing it the other way. There's no mathematical reason to do it that way. It's just a really bad habit.

0:08:09.7 SL: Right. Well, I'm glad you bring that up, and talk about the fact that we have the tools, because white folks, including myself, and white led agencies and organisations hold the disproportionate amount of power and resources and influence. I'm sitting in my home office with a mic and a laptop, and in a workplace that is supporting doing work from home at the moment, and we have a lot of influence in the work to prevent sexual violence. What are some strategies you would recommend to those working to end sexual violence who might be sitting in agencies such as the one I'm sitting in and others, to ensure that the experiences of black, indigenous and people of colour are centered and not tokenized? 

0:09:17.9 HK: Yes, well, that is a very big question. [chuckle]

0:09:21.2 SL: It is, it is.

0:09:23.9 HK: So I can speak to that from a data perspective, and again, there are many, but a couple that I would recommend starting with is really to re-examine our use in the research sector of social identity variables. So it's very kind of ironic that in the data science sector, when we gather up ourselves and we say we wanna do data science for good, the first thing we do is put out some kind of a survey that says, "Raise your hand if you're black." If you really think about that, it is very, very problematic. And so much of the time, we collect social identity data with an intention to help use this data to pierce through racism and violence, but then we accidentally use social identity data as a proxy for a whole bunch of other lived experiences, and in that way, we accidentally double down on racism.

0:10:30.9 HK: So if we ask somebody for their racial-social identity, and then they tell us, and then because of that, we make a whole bunch of assumptions about what their lived experiences are, because it's convenient for us. Data science is a lot about making things efficient and convenient. So rather than asking about somebody's social identity, to use that as a proxy for their maybe experiences of marginalisation or oppression or of racism, we recommend you ask directly about their experiences of marginalisation, oppression and racism, rather than asking them what colour they are and then making a bunch of assumptions. So that's kind of a really important practical thing I would suggest.

0:11:22.8 HK: I am not suggesting not collecting social identity data, because I do think it's very important, the work of Abigail Echo-Hawk, for example, from Seattle around missing and murdered indigenous women has been extremely useful in pointing out how important is that we do know, that we do count people who are black or people who are indigenous, and we understand the depths of the problem, but not to use that as a proxy for what their experiences are.

0:12:05.3 SL: I understand that distinction. I think that's really helpful in that... So our work in sexual violence prevention is often to make visible what has been invisible previously, but then there also is this complex story that we need to be telling, especially as we are working in community with people to change these underlying factors that support...

0:12:49.9 HK: Yes.

0:12:51.7 SL: Sexual abuse and assault and harassment.

0:12:54.1 HK: Yes, and to, when you're using data, not equate race with racism. So when you talk about making the invisible visible, that's so important. So we certainly want to raise awareness about the fact that these are problems, but the problem isn't that there are indigenous women. The problem is that there is racism against indigenous women, and sometimes we can accidentally, in our quest to make racism visible as a cause of this problem, we use race as a proxy for that, and that is the problem with the data.

0:13:38.9 SL: I'm really glad you brought up Abigail's work, and specifically the violence experienced by native people, and native women specifically. Could you share your experiences with data sovereignty issues and how we as a movement to end sexual assault, harassment and abuse can support indigenous communities specifically? 

0:14:09.6 HK: Great, great question. And a couple of years ago, I would have said yes, but now I am a little bit wiser and I can say no, no, I can't because I don't have any direct lived experience of being an indigenous person, but I am very fortunate to be in collaboration with many indigenous people, specifically around data sovereignty and data equity, and what I'm learning is that I'm so deeply steeped in the academic system of a white person that I'm still very much in the process of undoing all of that learning, and even the things that I thought were kind of helping to kind of blend Western post-industrial quantitative research with indigenous methodologies were still very much centred in whiteness.

0:15:10.8 HK: And so yeah, I'm really learning that I have so much further to go before I can really speak to this at all. One of the things that I have just had a conversation with about yesterday, which really moved me was how cavalier my attitude is, and many people similar to me. How cavalier our attitude is towards data, and even in... I was talking to somebody who was doing work with aboriginal children in Australia, and they would not let any data be collected from these aboriginal children unless there was a permanent full-time data spiritual advisor and protector that was gonna go with that data. And I just thought, that is amazing. I am so far away from this, but boy, do I wanna get there and learn things like that. That just moved me to my core, and I thought, absolutely.

0:16:21.4 SL: I am feeling that, yeah, in my core.

0:16:24.1 HK: So... Yeah, I don't think we can support indigenous communities. I think we can get humble and hopefully learn.

0:16:32.5 SL: Absolutely, and there's so much about... A lot of our work is focusing on helping to support folks that are looking to evaluate their homegrown programmes, and so I really appreciate you talking about humility. This kind of came up as I was having a conversation with a co-worker about the great deal of tension that exists between social change work and when there are some existing funding structures for that change work. And the question that came up as we were talking is how transformative can we really be working within systems and structures that have caused harm, and in many cases are continuing to cause harm? How do we manage this tension, and what tools might help us to navigate funding mandates while supporting and lifting up specific community stories and needs? 

0:17:45.0 HK: Yeah, frankly, I don't know if it's possible. Yeah.

0:17:51.6 SL: It's hard. I guess you talked about at the beginning this need of just transparency.

0:18:00.0 HK: Yes, yes.

0:18:00.3 SL: And part of it seems to even just acknowledge the tension, and that being important.

0:18:06.7 HK: Yeah, yeah. One of the tools that we have found to be really effective is something that we've developed over the past couple of years that we call the funding web, which gets right at the heart of this. The funding web is a very simple thing. We've used it with small children, and we've also used it with leaders at the United Nations, and it is a very tangible diagram of how money, data, and decision-making are changing hands across major stakeholders in any data or research project, and we have found getting something like that down on paper, there's not words. It's just circles and lines, and it really illuminates very quickly what power dynamics are being supported by various funding structures, whether they're governmental or philanthropic or academic, corporate.

0:19:06.9 HK: And what we've also found that this funding web helps have these conversations in a more strategic and sometimes more effective way. If you're sitting down with a funder and you're pointing to an arrow that shows that the victims of violence are contributing all of the data to this project, but getting no money and no decision-making, that is a really non-controversial, tangible, grounded way to say, "Hey, I think we could bring some balance of power here if we drew some arrows that were a little different." So I'm not sure how far you can go within current structures, but if we are operating within current structures, I have found that this funding web tool is very simple and allows for these conversations to happen in a productive, less radical oriented way.

0:20:09.9 SL: Thank you. I know our listeners love a practical tool to help with this really important work. As we close up our time together today, is there any other information you'd like to share with the group? Anything you'd like to recommend where folks can go find additional information and support around data equity work? 

0:20:38.7 HK: Sure, well, you can find a lot of the tools and the various stories that I was talking about and images at our kind of main hub, which is weallcount.com, but to just kind of circle back to something we talked about earlier, I would just read and watch everything that Abigail Echo-Hawk does. I think she nails these points very, very well, and she has lots of YouTube videos. She was actually in Vogue magazine last month, which is amazingly cool, to do with some work she was doing with COVID. And speaking of COVID, I would really recommend the Data For Black Lives organisation.

0:21:21.6 HK: That's founded by Yeshimabeit Milner, and they have been doing some great work around COVID data, and ensuring and advocating and developing tools to make sure that COVID data is not used as a weapon against black communities, but rather kind of harnessed to support black communities. And the Indigenous Data Sovereignty Network, I think it grew out of the University of Arizona, but it's national now, actually international now, is always my go-to source if I need data or if I need an expert on data equity. Oh, and just this week or last week on Netflix, the algorithmic Justice League's movie, Coded Bias, just came on Netflix. So if you have a Netflix account, watch Coded Bias. Great movie.

0:22:16.9 SL: I love it. Thank you so much. I'm so grateful for your time. Thank you for the wonderful resources, for all the work that is happening, and how you're helping to lift up and share fantastic folks that are doing this work. I'm hoping actually that some of them might be future guests on the podcast.

0:22:41.3 HK: Yes! Highly recommended.

0:22:44.9 SL: Well...

0:22:46.2 HK: Well, thank you for having me. It was a real delight and a privilege, and I really respect the work that you're doing, and I think it's great, and I look forward to reading all about it going forwards.

[music]

0:23:12.2 SL: Thank you for listening to this episode of Resource On The Go. For more resources and information about preventing sexual assault, visit our website at www.nsvrc.org. For links to resources discussed and other episode resources, go to www.nsvrc.org/podcasts. You can also get in touch with us by emailing resources@nsvrc.org.