Carnegie Mellon University

Dan Silverman Political Science Research Methods

June 18, 2020

Disrupted research provides Dan Silverman's students with important lessons in adaptation

By Bill Brink

Set up a table with brownies, the plan went. Who doesn’t like brownies? That should be enough of a carrot to halt students between classes and nudge them toward a voter registration table.

But the COVID-19 pandemic nixed the brownies. And the focus groups. And the in-person interviews. Students working on their group research projects in Institute for Politics and Strategy Postdoctoral Fellow Dan Silverman’s Political Science Research Methods class faced the same challenges as the rest of the world: a disruption of nearly everything.

The groups all studied the subject of political participation on college campuses. Why are some students active and others not? What accounts for the difference in engagement? Five groups of students took five different approaches: Surveys, quantitative analysis, interviews, case studies, and experiments. They all faced pandemic-related challenges.

“You would have people drop out of the study, saying, “I can’t deal with this right now,” said Renée Nikolov, a member of the interviews group.

“I think having to stay at home, and this pandemic, might have contributed to our limited sample size,” said Tanvi Siddharthan, from the group initially assigned to case studies. “In the list of things you have to do, answering a survey really isn’t one of them. When we collected data, it was us reaching out to our personal networks.”

According to Tufts University data, 40.3 percent of college students voted in the 2018 midterm elections, up from 19.3 percent in 2014. In the 2016 presidential election, 44.8 percent of college students aged eighteen to twenty-one voted. The pandemic has forced the campaigns for the 2020 presidential election into the virtual realm, away from the rallies and events that usually characterize such campaigns, and college voter turnout will again prove important if participation is to rebound from 2016.

The students in Professor Silverman’s class tried to find underlying factors, testing the effects of students’ socioeconomic status and their parents’ political views on their propensity to join or shun the political process.

Rigorous collegiate education can leave little time for the consideration of much else, so the survey group decided to examine the possible effect of stress on political participation. Using a scale of one to five, they collected information on stress, political participation, school, and demographics, but the surveys proved challenging. They had to remove half of the surveys from the data set because respondents either left sections blank or submitted junk responses. They did not find the impact of stress level to be statistically significant.

“When we did the very preliminary analysis, you’d expect the base level of stress, the higher level of stress, if you were a five, you were less likely to vote than a four,” said Sujay Utkarsh, one of the group members. “It turned out to be the opposite scenario.”

The quantitative analysis group used data from the American National Election Studies to determine the effects of financial security and access to technology on political participation.

“Given that so much politically has gone online in the past decade, specifically, we thought there might be a really fruitful relationship there to explore,” group member Ruth Pace said.

The study found that smartphone use correlated with voting behavior.

The interview group hypothesized that social science and humanities majors would engage in politics at a higher rate than STEM majors. Because of the pandemic, the interviews were conducted virtually, via Zoom.

“Interviews allow for a more personal [connection] and prompt more honest answers because you can read the room, in a way, and you can sense body language,” group member Leila Berger said.

The majority of students interviewed believed that both education in general and their specific major influenced their decision to participate in politics, but a smaller percentage of students felt that their major made an impact.

“Across the humanities, even modern languages … you can see how professors are pretty intentional with making [politics] a point of conversation, and weaving that into the discussion,” group member Audrey Pederson said. “I think if we’re able to encourage that type of dialogue across other fields of study, it would be really beneficial.”

The case studies group planned to identify factors that lead students to attend political protests using focus groups, but the pandemic relegated them to Google forms. They tested two variables: parental political ideology and income status. Hoping for fifty responses, they received ninety-eight. The majority of both the survey respondents and their parents identified as liberal, but the study found no link between ideology/income and participation in political protests, their chosen indicator.

“Students want to use their voices and engage in these political processes, but they just find the transaction cost too high, or they find barriers to these protests rather than an outright dislike of political protest,” Siddharthan said.

Once the brownies were off the table, the experiment group used a survey to study the effects of articles extolling the positive impact of political participation on students’ views of the process, but could not establish statistical significance. A randomization assigned only 25 percent of the sample size to the control group, further clouding the results.

“We didn’t set it to be 50/50,” group member Aden Halpern said. “There’s a setting where you can randomize it to be an even number of both conditions, and perhaps in future experiments we can do that.”

The virtual presentation attracted a few members of the Carnegie Mellon community. Center for Student Diversity and Inclusion Executive Director M. Shernell Smith, Assistant Dean of Student Affairs for Assessment and Strategic Initiatives Joanna Dickert, Associate Vice President for Community Standards and Diversity Initiatives Holly Hippensteel, Associate Dean of Student Affairs Elizabeth Vaughan, the director of Student Leadership, Involvement, and Civic Engagement (SLICE), and SLICE Coordinator Katterin White all attended the presentations.

All the groups developed essential research skills. But the pandemic also taught every team another lesson: that research is about adaptation to circumstances unforeseen. On that score, Silverman’s students gained valuable experience that will help them moving forward at CMU and beyond.