Inside an Initiative to Diversify the Field of Computational Social Science

[PREVIEW CLIP OF NICOLE JENKINS]

I was really excited to be also around other researchers that look like me and share research interests in terms of racial inequality and those sorts of things. So it's not often that you're going to find a group of scholars of color who are doing impactful research, all interdisciplinary, from all over the world, right?

[J.B. WOGAN]

I’m J.B. Wogan from Mathematica and welcome back to On the Evidence, a show that examines what we know about today’s most urgent challenges and how we can make progress in addressing them. On today’s episode, we’re going to focus on a creative initiative designed with two purposes: First, to build the pipeline of researchers who use the methods and tools from data science and social science. And second, to increase the diversity of researchers in that interdisciplinary space so that they can directly confront inequity.

Earlier this year, Howard University and Mathematica sponsored a free, two-week training for graduate students, postdoctoral researchers, and beginning faculty in the fields of data science and social science. The training was part of a broader instructional program held at sites across the globe called the Summer Institutes in Computational Social Science, better known by the acronym S-I-C-S-S or “six.” The Howard-Mathematica SICSS was unique in that it was the first site to be hosted by a historically Black college or university and it was the first site to focus on anti-Black racism and inequity.

We’ll focus most of the episode on the context and content of the Howard-Mathematica site, but I want to start by giving a little background on the idea behind the summer institutes because it relates to a broader story that I’ve encountered during my time at Mathematica. And I’ll lean a bit on history that Mathematica’s president and CEO, Paul Decker, provided during a session at SICSS. In the 1970s, when companies like Mathematica were getting off the ground, data-driven social scientists were the disruptors in public policy. Here’s Paul.

[PAUL DECKER]

But back then usually the first step of a significant policy research study was, was to go out and collect primary data. And it was companies like Mathematica, a few companies like Mathematica, that did that. And so we were the missionaries really for rigorous research at that time in a world that didn't understand research, didn't understand data. But now over the, I'll say the past 10 years, that's been changing rapidly.

[J.B. WOGAN]

Data used to be rare and static. Now, Paul says, they’re everywhere and new technologies are arising that allow people to use those data more effectively. Paul joined Mathematica in the late 1980s and became CEO in 2007. For evidence of just how much technology has changed the policy research landscape even in the relatively short period that he’s been CEO, one need only listen closely to the language used by today’s researchers in describing their tools and data sources.

[PAUL DECKER]

Twitter, gamification, wearables, data scraping or web scraping, internet of things, github, Slack, R, R studio, facial analysis software, Affectiva, telehealth, mechanical turk, APIs, interrogation and observation data, automated text analysis, NLP, sentiment analysis, social media, media cloud, Venmo, and geo-tag, nobody would've known what the heck you were talking about if you used those words 15 years ago.

And so the impact of the new data and the new technology has had a profound effect on our market and on research and how we think about policy research. So it's this huge disruption in comparison to what we've experienced in the past. So I think of it as really a whole new era for policy research. I think of the first 25 years of my career as what I call the research and evaluation age, that age where, you know, where we were really the missionaries behind this novel concept of data and generating evidence and rigorous research. But now I think we've moved into a new age where today we’re swimming in data, those data are dynamic, new sources arise every day and the technology to exploit them, rises every day as well. And so I think that really fundamentally changes the environment in which we operate.

[J.B. WOGAN]

As digital technology advances, people who work at the intersection of evidence and decision making are using new tools and data sets to do things that just weren’t possible before, or at least to do them faster, at a greater scale, and more seamlessly than before. At Mathematica, we’re still hiring plenty of economists, sociologists, and other social scientists every year, but we’re also posting more job openings for data scientists and web developers. More and more of our projects require collaboration across the disciplines to create algorithms and build user-friendly online dashboards and data visualizations. The summer institutes are for people who want to be computational social scientists, people who sit at the intersection of these fields.

[MATT SALGANIK]

It's actually one of the first things that we talk about at the Summer Institute is what is computational social science? And the answer that I have is that it's anything that's cool.

[J.B. WOGAN]

That’s Matt Salganik, a professor of sociology at Princeton University.

[MATT SALGANIK]

But more seriously, I think of computational social science as research about people in societies that uses the tools and capabilities of the digital age.

[J.B. WOGAN]

In 2017, with funding from the Russell Sage Foundation, Matt co-founded the first summer institute with Chris Bail, a sociology professor at Duke University. Matt, by the way, is a member of Mathematica’s Board of Directors and the author of a book called Bit by Bit, which is for social scientists who want to tap into the potential of big data and for data scientists who want to apply lessons from social science to new technologies. Here’s Matt again.

[MATT SALGANIK]

And so I think, you know, we hear about research that takes advantage of data from Twitter and Facebook and mobile phones. So it's that, and more. And part of the reason why I don't try to define it very clearly is because it's changing very quickly. So when I taught my first course in computational social science in 2007, and it was called web-based social research because at that time I thought it only meant the web, but now I realize that in fact, the digital age is much broader than that. So it means mobile phones. It means wearable sensors. It means the internet of things, and what it will mean in five years and 10 years will also be different, right? And so I think one of our goals with the summer institute is to keep the definition broad, create a big tent where lots of new ideas can come in and where the field can continue to evolve.

[J.B. WOGAN]

In the first year, SICSS was only at Princeton and only had room for 30 in-person participants. Close to 300 people applied. So Matt and Chris made as much of the experience accessible to even those who couldn’t attend. They streamed all of the lectures and made other teaching materials available online. But they knew that a major benefit of the training was the opportunity to find a community of like-minded colleagues and mentors. So they turned to alumni of the first summer institute and invited them to launch more locations. Here’s Matt again.

[MATT SALGANIK]

And so in 2018, I believe we had seven partner locations. And then in 2019, I believe we had 11 or 12 and then 2020 came. And that was COVID. We were planning to have about 20. And in the end, we had about eight or nine, which I think is great for COVID. And then, this year, you know, because of COVID, everything was remote, and we still had 20 partner locations, around the world run by our alums and other members of the SICSS community.

[J.B. WOGAN]

One of the participants in 2019 was Naniette Coleman, a PhD candidate in the Sociology Department at the University of California, Berkeley.

[NANIETTE COLEMAN]

So, I should begin by saying I'm a bit of a non-traditional student.

[J.B. WOGAN]

That’s Naniette.

[NANIETTE COLEMAN]

For people who can't see me, I am of a bit older student. I decided to go back and get my PhD in my late thirties, so when I arrived at SICSS at Princeton, I already felt, and I can say this also about my arriving at Berkeley. Like I stuck out a little bit, so I'm older. I'd worked for a bit. I lived in Washington D.C. and worked for the federal government, also in international development for a bit. But beyond that, you know, as a woman of color, as a Black woman, I often find myself, and this is throughout my career, not just in academe, as the only in many spaces, the only woman, the only Black person. And now at this point in my life, I'm also one of the older people in the room, sometimes even older than my faculty. That's always fun. So when I arrived at Princeton, I will say I was always, I was already a little bit nervous, a little anxious entering into a new space where I was going to be learning this new toolkit, and I was excited to see other people of color in the room. I don't think I'm the only person who does a quick check, looks around, tend to see, tend to look and see who looks like you.

[J.B. WOGAN]

Naniette saw two people of color, one from Harvard and one from Northwestern, who would become friends during the institute. She also noted the faculty, Matt and Chris.

[NANIETTE COLEMAN]

After that experience, they not only became, you know, people who I admired greatly for their teaching prowess and research, but also friends, people who have sponsored and mentored me also. So I saw them in the front of the room, with all that said, they are both White men. I looked at our TA staff, again, incredible people who've both supported me, and through this challenging work, as a newbie, I also call myself a newbie to computational social science. But for the most part, they were also all people, that if I did that check in the room, I would not count as, as being on, you know, necessarily understanding my experience as a Black woman.

[J.B. WOGAN]

After her experience at SICSS, Naniette started thinking about how she could build on the SICSS template to provide an instructional program that would entice more scholars of color to become computational social scientists.

[NANIETTE COLEMAN]

We, people of color, tend to appreciate seeing a person in the front of the room, or a person, in the citation list and the article we're reading, or an author or a faculty member, or a data set, or considerations that include us. So when I imagined how I might do something like SICSS, and I can’t help but do that when I see something that’s done exceptionally. I ask myself, you know, if I was to do something like this, what might I add to it? I imagined the creation of a space where being a person of color was somewhat of a foregone conclusion. And there were considerations for our experience that went beyond just sort of the content of the data, but considered the unique challenges of being only, or other in academe and as well as what it might feel to have imposter syndrome, where, you know, I think this isn't unique to just the Black community or people of color, but feeling like you enter a space and maybe you're there by accident, or, you know, your application blew across the room by accident, landed on the top of the pile. You know, what would it be like if in that space, we prioritized the confidence, the building up, the encouragement of individuals and didn't ignore the fact that most people in that space probably had some anxiety about being there. What would that look like?

[J.B. WOGAN]

What that would look like, Naniette thought, was a summer institute hosted by an HBCU. When she mentioned the idea to Matt and Chris, they were supportive and connected her with someone else thinking along the same lines: Akira Bell, the chief information officer at Mathematica.

[AKIRA BELL]

A lot of my personal interests and my path before coming to Mathematica has been about, you know, getting more underrepresented minorities, represented in STEM careers. And that can't happen without significant investment in the pipeline before people show up for jobs.

[J.B. WOGAN]

That’s Akira.

Akira’s is an alumna of Princeton University and sits on Mathematica’s board with Matt, so she was already aware of the summer institute. Like Naniette, she loved the idea of increasing the pipeline of computational social scientists but thought there was an opportunity to leverage a host site, like an HBCU, to also boost the diversity of that pipeline. And she had one in mind: Howard University, which is near Mathematica’s office in Washington, D.C. Akira had met Howard’s president, Dr. Wayne A.I. Frederick, when they served together on a panel about the digital divide.

[AKIRA BELL]

And I knew that Howard in particular had an interest in equity from a research perspective, but also in data science, and I thought they would be a perfect fit to bring this to a larger HBCU audience.

[J.B. WOGAN]

Since computational social science isn’t so much a field as a loose amalgam of related fields, no one has crunched the numbers on diversity, or lack thereof, in computational social science. But data do exist about fields like economics, data science, and the STEM workforce, all of which suggest that Black, Latinx, and indigenous people are underrepresented in computational social science. Here’s Howard University’s president, Dr. Frederick, on why that representation matters.

[WAYNE A.I. FREDERICK]

Like no other industries or fields of exploration, the STEM disciplines have a tremendous influence on the trajectory of American society and the entire planet for that matter. And that's why it's so critical that these industries are diverse, so they can represent the interests of all people, not just the privileged. I think a society that works for everyone cannot afford underrepresentation. And unfortunately, Black people in particular are underrepresented in all aspects of STEM, in academia, in business, in research, and so much more.

[J.B. WOGAN]

Representation in research matters, Dr. Frederick says, because lived experience informs how people build and use data-driven tools, which then influences how those tools affect society.

[WAYNE A.I. FREDERICK]

The things like artificial intelligence, I think are critical. When you look at something like law enforcement, artificial intelligence can help do these jobs more efficiently without bringing harm to our citizens. But if we are not at the forefront of shaping what that artificial intelligence looks like, it can actually be used for further discrimination.

[J.B. WOGAN]

The digital technology available today is a double-edged sword, Akira says. It has the potential to enable researchers to collect more kinds of information, at greater scale, and in less time, which then helps decision makers in government, philanthropy, and in the business community respond to pressing issues with timely evidence. But it can go wrong, and has. There have been high-profile cases where algorithms made it less likely a Black person would receive bail, have a mortgage application approved or even receive necessary health care services. Here’s Akira again.

[AKIRA BELL]

This is a powerful time to be a data scientist. And I mean that quite literally. You know, we've got so many advances in processing and capability and algorithm generation, and machine learning. But to do that in a respectful way, in a way that that builds trust, that doesn't build distrust. We've got a lot of distrust for AI right now. We've got a lot of distrust for certain social services right now. We've got a public health system that is trusted by some and not trusted by others. So adding data science and things around machine learning and artificial intelligence can exacerbate some of that.

[J.B. WOGAN]

Akira says the next generation of computational social scientists, at Mathematica and elsewhere, need

[AKIRA BELL]

To think about ways in which we can use that technology for trust building to create more equity, to create a sense of integrity in what we're producing, a faith in the data that's being leveraged and how it's being leveraged, and to answer questions.

[J.B. WOGAN]

So Naniette and Akira connected with Dr. Frederick, who supported the idea and asked his senior advisor for strategic initiatives, Calvin Hadley, to help with planning the event. Original plans for an in-person gathering in 2020 were scrapped due to the pandemic, but in mid-June 2021, Mathematica and Howard University welcomed 26 participants for a special all-virtual Summer Institute for Computational Science focused on countering anti-Black racism and inequity. In announcing the event, Howard and Mathematica had encouraged applicants from underrepresented groups to apply, and it worked. The vast majority of individuals who applied were either alumni of an HBCU, PhD students at an HBCU, faculty at an HBCU, or data scientists at an HBCU. All of the participants were scholars of color.

One of those participants was Felix Owusu, a PhD candidate in public policy at Harvard whose dissertation research centers on racial disparities in the criminal legal system. Felix heard about the institute just a couple days before the application deadline and decided to drop everything to apply.

[FELIX OWUSU]

I was on the edge at first because I was like, this summer was set aside to really write my dissertation. Should I take two weeks to do this summer institute? And if it had been somewhere else and not at Howard and Mathematica, I very likely would have said, this is not the right time, but I didn't know if I was going be happening again. I thought it was a once in a lifetime opportunity.

[J.B. WOGAN]

As Naniette and Akira had anticipated, Howard’s involvement signaled to Felix that this convening would be different from other opportunities to network and gain new skills in computational social science.

[FELIX OWUSU]

In my mind, what Howard was able to bring was unique. And I think, especially in the sense that it is very much what is missing. I think in terms of Howard being able to bring in the right kinds of people and convene a conversation that's credible and specifically centers the experience of, you know, Black people, people experiencing anti-Black racism, and other kinds of bias. I think that's important and the consistent history of Howard's work in that space to address those kinds of issues, I think really, really means a lot.

[J.B. WOGAN]

For Felix, the fact that Mathematica was co-sponsoring the event with Howard added to the can’t-miss nature of the program.

[FELIX OWUSU]

So Mathematica is an institution that I've been familiar with for years. They're sort of a pillar in this applied research space, right? And so seeing that it was a tag team between Mathematica and Howard was really compelling for me. I think I could definitely trust Mathematica to bring in terms of any technical assistance or methodological research design assistance, and I think they did that with flying colors. That was a really appealing aspect of this. And I think as a person whose research tends to be pretty applied and is looking to continue to do work in that space, that was, I think something that was quite appealing.

[J.B. WOGAN]

Felix came to the institute hoping to learn more about automated text analysis, which he plans to use in his research about policing.

[FELIX OWUSU]

One thing that a lot of other jurisdictions have that is often overlooked is narrative information from police reports. You can see who was stopped, you might see where, or for what reason, that's being recorded, but you're not going to know necessarily what the situation entailed, what the event was like, or what that was going through the officer’s head. That's often recorded. And in Boston, for example, where I went to school and where I lived for awhile, that information is available online in their sort of publicly available police data. So I looked at that data, that I'd been familiar with and that people have looked at before, but dug into the narratives about the reasoning behind the stops and was able to figure out how are these officers justifying their stops differently for people based on the race of the person who is being stopped and other demographic information. And so I started to dig into it, and hope to spend more time going forward, working on this, looking at the justification and reasoning behind stops, and whether that's number one responsive to the demographics, the person being stopped. But also if that gives us insight into how policies or changes in reforms to the criminal legal system influence police activity and interactions.

[J.B. WOGAN]

Another participant in the institute was Nicole Jenkins, an assistant professor at Howard in the Department of Sociology and Criminology. Nicole’s ethnographic research focuses on understanding the experiences of Black women in institutions.

[NICOLE JENKINS]

I think what was most appealing for me was the opportunity to sort of work in an interdisciplinary research setting.

[J.B. WOGAN]

That’s Nicole.

[NICOLE JENKINS]

Often you find that quantitative scholars are collaborating with other quantitative scholars and qualitative scholars are collaborating with other qualitative scholars and then the same sort of applies to research hubs, right? So sociologists with sociologists, and of course scholars are looking to collaborate interdisciplinary more and more. But any time there's an opportunity to do work like that, I think researchers become really excited.

[J.B. WOGAN]

Beyond the interdisciplinary nature of the institute, Nicole liked that it was focused on addressing anti-Black racism and that the event was designed to be welcoming to Black researchers.

[NICOLE JENKINS]

I was really excited to be around other researchers that look like me and share research interests in terms of racial inequality and those sorts of things.

So it's not often that you're going to find a group of scholars of color who are doing impactful research, all interdisciplinary, from all over the world, right? This is very difficult to do. Of course, there are conferences and these sorts of things, but this was unique also in that it was held at an HBCU.

[J.B. WOGAN]

Nicole came to the institute interested in learning about tools like web scraping on social media that could give her new insights about Black women’s experiences with hair-based discrimination. The current focus of her research is on the CROWN Act, legislation that bans race-based hair discrimination in work and educational settings. As of October 2021, a version of the CROWN act has become law in some states and localities, but not at the federal level.

[NICOLE JENKINS]

I think that most people are under the assumption that you cannot discriminate against a person based on their hair texture. They would just assume like you can't do that, but you actually can, you know, it's federal law that employers are allowed to not hire a person if they have locks.

[J.B. WOGAN]

The topical focus of countering anti-Black racism also appealed to Jeremy Prim. He’s a PhD candidate in the Department of Sociology at the University of California, Davis.

[JEREMY PRIM]

So I think for my own research kind of agenda, it was a perfect fit. My research focuses predominantly on the way the carceral state is embedded in different educational spaces, specifically looking at policing, exclusionary discipline, and how that disproportionately affects Black and brown students. So when I saw the topic, I was like, that's directly related to the work that I do because when you talk about anti-Blackness you have talk about the type of exclusion and surveillance that is really ingrained in that theoretical framing.

[J.B. WOGAN]

Jeremy is also an alumnus of an HBCU, Morehouse College, and seeing that an HBCU was hosting the program made him more interested in applying.

[JEREMY PRIM]

I've been at a PWI, predominantly White institution, for the past five years. I said, you know, what would it be like to actually kind of be back at a HBCU, whether virtually or in person, just to kind of get that feeling again, that same feeling that I had four years ago? And I definitely got that same type of fulfillment and enrichment, just with even more support, with the team from Mathematica and just all the different TAs we were able to get that support from.

[J.B. WOGAN]

Jeremy mentioned the support he was hoping to find at a program hosted by an HBCU and that’s something that came up in multiple interviews. Naniette told me she wanted to create a welcoming environment where participants didn’t feel anxiety about whether they deserved to be there. The message that participants did belong, that their work mattered, resonated with the participants. Here’s Nicole Jenkins again.

[NICOLE JENKINS]

Imposter syndrome is a real thing. You know, we're all, all of us are sort of sitting there like, am I supposed to be here? Are they going to figure me out?

[J.B. WOGAN]

The program’s line-up of guest speakers confronted that fear head on.

[NICOLE JENKINS]

This space was absolutely affirming. So above all that I learned, which was amazing, I was able to spend my time each day with this fabulous group of like brilliant scholars and every day there were these words of affirmation, sort of reminding one another, how incredible our research is and how powerful the work is that we're doing and how, you know, we belong in these different spaces.

[J.B. WOGAN]

Jeremy Prim remembers something that Akira Bell, the chief information officer at Mathematica, said to the group, which stuck with him throughout the two weeks.

[JEREMY PRIM]

She pretty much told us, she said that like, the reason why each of you were selected is that you all have a lens that you are looking at a social problem with and there is value in that lens. So just take your time, take your time, take a deep breath. Like really just try to get as much as you can from this experience, because your lens is just going to get even sharper.

[J.B. WOGAN]

Since SICSS launched in 2017, participants have published articles in peer-reviewed journals that started with a team project at SICSS. WIRED magazine has covered research that originated from SICSS. Many of the SICSS alumni have become professors and are exposing more students to new ways of doing research with digital tools. I asked Matt Salganik what he thought the biggest impacts were from SICSS.

[MATT SALGANIK]

I actually think one of the biggest impacts is the community that we're creating. So one of the things I'm particularly happy about is the way that the people from SICSS continue to be colleagues and help each other and learn from each other over time.

[J.B. WOGAN]

That was definitely true of Felix’s experience. He says his cohort at Howard was supportive, energetic, and engaged, despite all of the activities occurring over the computer. And that engagement continues.

[FELIX OWUSU]

We have our own group chats that are persistent. We have a Slack channel and people are continuing to have, you know, co-working writing sessions to give feedback on people's ideas.

One of the things that Naniette really engendered in the group was humble brags or sort of giving people their due recognition when people have positive things happen in their life. It's been a challenging, you know, past couple of years at this point. And so really celebrating people's victories is something that we've carried forward as well. So I've been really happy to find a community of people that I feel like today I could turn to, if I had a question or I just wanted somebody to like be on Zoom with, to work in a less isolating way. And that's been a great takeaway.

[J.B. WOGAN]

The Howard-Mathematica SICSS might also start a trend of other sites that incorporates diversity and equity into their training. Here’s Naniette Coleman again.

[NANIETTE COLEMAN]

I expect that some of the individuals that attended might be inspired to start a location at a Hispanic serving institution or one of our Indigenous colleges, either in North America or abroad. You know, there there've been some unique conversations about places that this can expand beyond the sort of the tradition that Matt and Chris have created. So I'm excited to see how this will proliferate and what our student participants will do with what they learned.

[J.B. WOGAN]

In reflecting on what the Howard-Mathematica SICSS accomplished, Akira hopes it provided resources to groups of students who are underrepresented in computational social science.

[AKIRA BELL]

And that they think of really impactful ways of applying those resources to the country and the world's most important problems about equity and access, to, you know, resources, benefits, and kind of justice elements. What we saw from them, because their research interests were very broad, and I got a chance to sit in on their, um, their project proposals of what they're going to work on post-SICSS. And it's an amazing set of everywhere from thinking about the future of AI and telehealth and how that, you know, facial recognition for African-Americans work, to thinking about how do we perfect NLP algorithms for African-American vernacular English and, you know, even scraping sentiments, scraping for social justice issues. There was a very wide array of types of problems to think about, and that's exactly what we wanted: some new voices, some new lenses, there were some problems in there that I had never even thought about might happen as a result of introducing, you know, kind of data science and intelligence systems to certain problems. And they're thinking about it because they're coming at it from a different lens, so that's exactly what we hope to achieve.

[J.B. WOGAN]

Thinking back to the early planning conversations about holding a SICSS at Howard, Naniette remembers sensing that a SICSS for aspiring Black computational social scientists had great potential.

[NANIETTE COLEMAN]

It just seemed like a space where, with thoughtful investment, the dividends that would just multiply, just capitalize and multiply across not just HBCUs, but across the Black diaspora, would just be immense. So it was a fortuitous moment that not only turned into something pretty beautiful, but I think affected all of us who were involved in it in a beautiful way as well.

[J.B. WOGAN]

Thanks to my guests, Matt Salganik, Naniette Coleman, Akira Bell, Felix Owusu, Nicole Jenkins, and Jeremy Prim. You also heard clips from Paul Decker and Dr. Wayne A.I. Frederick. Those clips came from Paul’s and Dr. Frederick’s remarks at plenary sessions at the Howard-Mathematica SICSS, with their permission. As always, thank you for listening to another episode of On the Evidence, the Mathematica podcast. There are a few ways you can keep up with the show. You can subscribe wherever you listen to podcasts or you can follow us on Twitter. I’m at JBWogan. Mathematica is at MathematicaNow.