Authoritarian and Democratic Data Science in an Experimenting Society

How will the role of data science in democracy be transformed as software expands the public’s ability to conduct our own experiments at scale? In the 1940s-70s, debates over authoritarian uses of statistics led to new paradigms in social psychology, management theory, and policy evaluation. Today, large-scale social experiments and predictive modeling are reviving these debates. Technology platforms now conduct hundreds of undisclosed experiments per day on pricing and advertising, and the algorithms that shape our social lives remain opaque to to the public. Democratic methods for data science may offer an alternative to this corporate libertarian paternalism.

In this talk, hear about the history and future of democratic social experimentation, from Kurt Lewin and Karl Popper to Donald Campbell. You’ll also hear about CivilServant, software that supports communities to conduct their own experiments on algorithms and social behavior online.

J. Nathan Matias is a Ph.D. candidate at the MIT Media Lab Center for Civic Media, an affiliate at the Berkman-Klein Center at Harvard, and founder of CivilServant. He conducts independent, public interest research on flourishing, fair, and safe participation online. These include research on harassment reporting, volunteer moderation online (PDF), behavior change toward equality (PDF), social movements (PDF), and networks of gratitude.

Nathan has extensive experience in tech startups, nonprofits, and corporate research, including SwiftKey, Microsoft Research, and the Ministry of Stories. Nathan’s creative work and research have been covered extensively by international press, and he has published data journalism and intellectual history in the Atlantic, Guardian, PBS, and Boston Magazine.