Social science is in tumult. The federal government pays for a shrinking share of research. A replication crisis undermines the validity of many peer-reviewed studies. Public skepticism abounds.
Another unsettling force: the rise of private companies like Facebook, which increasingly control essential data about human behavior.
Can scholars tap that data to do good? Can they protect privacy at the same time?
These are the kinds of questions Alondra Nelson is trying to answer. Nelson, a professor of sociology at Columbia University, took over in 2017 as president of the Social Science Research Council (SSRC), an independent nonprofit group that promotes innovation in scholarship.
As an academic, Nelson is known for her 2016 book on genetic genealogy, The Social Life of DNA: Race, Reparations, and Reconciliation After the Genome (Beacon), which chronicles how African-Americans turned genetic data into a tool for reckoning with the history of slavery. She is also an institution-builder who served as Columbia’s first dean of social science.
Her experience informs the council’s new report about the future of social science, “To Secure Knowledge.” In it, Nelson and her colleagues describe an urgent need for a “new institutional infrastructure supporting social-science research.” Fresh partnerships, they argue, are vital. The council coordinates one such effort, to make Facebook data available to academics studying the impact of social media on democracy.
The Chronicle spoke with Nelson about Facebook, the shifting landscape of research ethics, and a new project that has her interviewing Obama-administration officials.
•
You spent years studying the social implications of genetic ancestry testing. How do you think about the role of corporations in research?
My book The Social Life of DNA is really about the first 10 years of this new start-up industry. And one of the lessons from direct-to-consumer genetic ancestry testing for data research, and social research more generally, is that part of the disruption often takes place in regard to research norms. I’ve watched with great interest how scholars who go to work in industry wrestle with the new norms necessary for research outside the academy, where we like to imagine they are more sacred.
As direct-to-consumer genetic ancestry testing is circulated in the world, it’s framed as highly scientific and accurate. But the industry doesn’t have the norms we would expect from gold-standard science. The databases are opaque. The algorithms and weights given to genetic markers and the like are not shared among companies or with the consumer. So it’s hard to have faith in the findings, beyond what the companies themselves say.
You’re part of the SSRC-Facebook Social Data Initiative , which has been described as “the first time that the social-media giant has agreed to release large amounts of proprietary data to outside scholars.” What potential do you see?
To the extent that private companies have collaborated with social-science or quantitative researchers to date, it’s been with a lot of constraints and restrictions — the type that can make it more difficult for research to be accountable, transparent, and reliable. The model created by Gary King and Nate Persily, which is our blueprint for this work in some ways, is an attempt to move to a space in which the companies are not handpicking researchers, not funding research only on the questions they would like to have proposed.
Scholars who go to work in industry wrestle with the new norms necessary for research outside the academy, where we like to imagine they are more sacred.
One alternative has been individual researchers who make private arrangements with various platform-technology companies. Sometimes these arrangements require that the researchers show the companies the data before publishing it, or give the company some say over where it is published or whether it can be shared so it can be verified, replicated, and the like. That’s not optimal social research. Another model is that people gather various data sets, often on their own. These aren’t data sets that are used collectively by a research community.
What I think we will have here is cohorts of researchers plumbing the same data sets. The power of academic insight is really an aggregation of researchers working together on related problems, citing one another, and being able to reference the same data.
This program was announced in the wake of the Cambridge Analytica scandal, alongside other changes that restricted researchers’ and commercial users’ access to Facebook data. Some scholars faulted both changes for effectively making Facebook an even more powerful information gatekeeper. What do you make of that criticism?
The Social Media and Democracy Research Grant program has created a process involving scores of confidential peer reviewers from around the world, with expertise in information science, quantitative social science, political science, and the like. It’s arguably the most academically rigorous vehicle for access to this data that’s been developed to date.
But is Facebook becoming a more powerful arbiter of information?
The research questions are proposed not by Facebook, or by the SSRC, but by those who respond to the call for proposals.
The privacy scholar Sarah E. Igo has said the toughest question won’t necessarily be scientific standards, but protecting human subjects. She urged the SSRC not simply to rely on the system of institutional review boards but “to reconsider, and possibly revise, the rules of social inquiry.”
Built into our process was an understanding that, while we require the highest standards of what universities and institutional review boards require, those were not built for a big-data society. And so we’re working with a National Science Foundation data-ethics team called Pervade to generate some new research insights on data ethics and privacy. That’s allowing researchers to study our process. We’re also working with another group of researchers to develop, as part of the peer-review process, an ethics-review process.
Part of the reason we’ve engaged researchers is that some of this is an empirical question. What are the best practices? What issues are different for this moment that we need to be thinking about?
Your next book is on science policy in Barack Obama’s administration. Why were you drawn to that, and what do you expect to say?
I’m interested in the attempts to create ethical frames around science-and-technology projects — big projects, moon shots.
There was a big commitment in the Obama administration to moving forward with science. And it didn’t shy away from white papers and events suggesting that artificial intelligence or big-data analysis came with tremendous downsides that should be explored. I’m interested in a presidential administration endeavoring to frame that conversation, even if it never intended not to go ahead with any of these projects.
You’ve done research on the Black Panthers, the social life of DNA, and now the Obama administration. What connects these interests? You wrote in your DNA book that you were “born from the union of a cryptographer and a mechanical technician.” Maybe that has something to do with it?
Left to my parents’ devices, I would have gone to medical school or something. But I was always much more interested in the social and political issues around science. What unites all these projects are the intersections of science and technology and inequality.
One line to follow is the Black Panther Party’s involvement in the first-ever grass-roots screening and testing program for genetic disease in the United States. From there I move on to genetic ancestry testing and the early adoption by African-Americans. It’s not an accident that the first chapter of the new book is about the rollout of big-data genetics research as a federal science project.
Some of the bigger questions are: What is the particular burden for communities of color, for African-Americans in particular, who have been historically victimized by scientific research, who have been the canaries in the coal mine of bad research ethics? And how do we at the same time encourage people in these communities to be engaged in the STEM fields, to have robust health-seeking behaviors? As a researcher, I’m interested in the ways these communities almost miraculously find ways to thread that needle.
This interview has been edited for length and clarity.
Marc Parry writes about scholars and the work they do. Follow him on Twitter @marcparry or email him at marc.parry@chronicle.com.