Drilling into the Data Pipeline
Daniel Hopkins, associate professor of political science, uses big data to gain insight into the issues facing various segments of the American public.
When behavioral scientist Daniel Hopkins wants to study tiny segments of a population, he turns to big data.
“Big data gives social scientists new opportunities to target narrow populations of interest,” Hopkins says. “We gather and analyze large and complex datasets to learn more about how people form their opinions and attitudes.”
Hopkins, an associate professor of political science who also has an appointment in the Annenberg School of Communication, is a leader in the data revolution in social science. His work centers on American politics, with a special emphasis on racial and ethnic politics, local politics, political behavior, and research methods.
Hopkins collects data from public sources, including nationwide lists of every person in the U.S. registered to vote, to study subsets of a large group by names or birth dates. He can then focus to specific populations—for example, people in debt—or subsets of professions, like doctors and lawyers.
“The data revolution has been very helpful because there is a decline in [traditional] survey responses,” says Hopkins, who was named Clarence Stone Scholar by the Urban Politics section of the American Political Science Association in 2015. “We’re able to make use of large datasets, like voter files, to anchor surveys, and we can work in terabytes rather than megabytes of information. That’s a big advance.”
His most recent research has focused on attitudes toward immigrants and has widespread implications for the future of race relations in American society. A paper currently under review argues that where non-Hispanic whites lived as adolescents—not where they’ve lived as adults—predicts their level of prejudice toward African Americans.
Another research study by Hopkins, this one also under review, focuses on immigrants. The study concluded that “people who over-estimate the share of immigrants in the U.S. tend to be more opposed to immigration. Providing accurate information about actual immigrant population shares does little to dampen opposition to immigration.”
Locally, Hopkins works with the city of Philadelphia to help engage residents effectively in municipal programs. For example, he uses his insights on behavior and motivation to increase ridership in the Indego bike share system, especially among low-income residents, and he is helping roll out a new program to assist Philadelphia residents with paying their water bills.
Hopkins has also shed light on political and societal attitudes by conducting research of what he calls “text data.”
“We’ll look at how many distinctive words are used in Congress to find out what the prevailing rhetoric is and what word choices politicians are using to see how these influence attitudes,” Hopkins explains. He then cross-references this with what people are saying on social media—for example, political activists on Twitter. “We can see changes in attitude before and after a barrage of ads that are pro or con an issue by looking at the textual content of social media,” Hopkins says.
Using this type of textual research, he tracks trends like how Americans understand the Affordable Care Act and how they describe it in their own words.
“Bigger data isn’t always better,” Hopkins says, “but if you analyze groups of 100,000 instead of 10,000 you can really drill down on what distinct groups of people are thinking.”