In 2008, during his last year of graduate school at Duke University, government professor Brendan Nyhan won a fellowship to launch an “innovative teaching project,” according to Nyhan. Through the fellowship, Nyhan created what has since become one of Dartmouth’s most unique government courses — GOVT 83.21, “Experiments in Politics.”
The class is a “a lab-style seminar in which [students] design, field and analyze an experimental study of misperceptions,” according to its syllabus, with the “ultimate goal” of publishing “a scholarly article in a peer-reviewed journal.” The class conducts research centered on political misperceptions — false and unsupported beliefs held by many people — since Nyhan’s research lies in this field, allowing him to more aptly advise students.
The class boasts a 100% publication success rate, with four articles published and the rest “in the pipeline for publication,” according to Nyhan. The “most influential” study produced by the class, according to Nyhan, was an evaluation of Facebook’s efforts to identify misinformation after the 2016 election, which has been cited “more than 200 times,” he explained.
“People have come to see the class and the research that’s come out of it … and have started teaching versions of the class of their own,” Nyhan said. “Professor [Mia] Costa and Professor [Charles] Crabtree teach their own versions of it [at Dartmouth] … One of my friends from graduate school has taught at [the University of California, Los Angeles] with a huge class. [The students at UCLA] don’t all carry out the research, but a subset of them do.”
The small class size — nine students are enrolled this term, and there is typically a limit of 16 students — paired with Nyhan’s focus on hands-on, student-led research attracts a wide array of Dartmouth students, though the course’s innovative nature often necessitates a heavier workload.
Nyhan first taught “Experiments in Politics” at Duke in 2008, he said. After arriving at Dartmouth in 2011, Nyhan said other professors in the government department were “open to [his] ideas,” which allowed him to begin teaching a version of the course at the College that same year. “Experiments in Politics” has been taught a total of “six or seven” times since 2011, according to Nyhan.
According to the course syllabus, students spend the first three weeks solidifying their understanding of experimental design. Each student then brainstorms four to five potential experiments before writing a proposal about how one could be carried out. The students present their proposals, and the class votes for the two most interesting and feasible options, according to Aidan Ferrin ’24, a student in the class.
In the fourth and fifth weeks of the class, students split into two groups, which each design surveys and collect and analyze data for one of the two pilot studies before presenting their findings, according to the syllabus. The class then votes on one of the pilots to become the publishable experiment. Nyhan said he abstains from voting to ensure that the research remains student-driven.
Nyhan said most undergraduate research “is conceived by the professor” and involves students “helping them carry out the vision.” By contrast, the structure of “Experiments in Politics” allows him to combine his experience in political misperceptions with students’ “creativity.”
“I think the combination of those two has kind of shaken loose some really interesting ideas that I wouldn’t have come up with,” he said.
The class spends the rest of the term on experiment development, survey broadcasting and analysis, after which each student writes a 4,000-word article about the study, according to the syllabus. According to Ferrin, Nyhan then takes the best pieces from each student’s paper to create the final article, which he sends to publishers.
Ferrin said the emphasis within the class on creating a publishable study creates a “real sense of ownership among students.”
While Ferrin has previously published papers, both through an independent study in the public policy department and with the Geisel School of Medicine, he said he felt this experience was different from his other publications due to its student-led nature.
“With this, I’m doing everything — start to finish, conceptualization to publishing,” Ferrin said. “We are all invested in this in a way I think is tricky to do if you’re researching with a professor. If it’s a professor, it’s their project.”
Nyhan hopes that this approach to the course allows his students to be “the kinds of people who can understand how research works” — regardless of the field they pursue.
“You could be working in a hospital, in an education system, in local government and you could be evaluating the effectiveness of whatever it is you’re doing using the same kinds of approaches,” Nyhan added.
According to Marina Wang ’25, a student in the class, Nyhan’s broad approach to research has already surfaced in some of her other projects.
“I’m currently in the process of working with the [Dartmouth Outing Club] to design a DOC census survey,” Wang said. “I’ve been able to actually draw elements from this class over to what I’m doing in the DOC, which is really exciting.”
Until recently, however, these overlaps would not have been possible. The field of political science has only begun to emphasize quantitative research in the past few years, according to Nyhan, who added that he began teaching the class “when experiments were starting to be more widely used in political science.”
Nyhan initially offered iterations of the class before the proliferation of online survey platforms, leading “Experiments in Politics” students to collect data in person. Nyhan said that when he taught the class in 2008, he and his students “took laptops to a hospital … and offered people cash to take [their] study online.”
As online survey platforms have become more popular, the studies undertaken by “Experiments in Politics” students have grown more sophisticated. Nyhan and his students no longer have to take wads of cash and checked-out laptops to meet respondents in person, he explained. The class is now cross-listed in the quantitative social science and government departments, reflecting the increasingly quantitative approach of its research.
“The beauty of this class is that you are bringing together so many different students from all across campus with such different interests and training,” Wang said. “Some of us are stronger writers, some of us have more coding experience and some of us have various angles on social issues that we want to … implement into how we do our experiment.”
Wang noted that the class is “one of the more intense” that she has taken at Dartmouth due to its fast pace, which is “compounded by the fact that we are on the quarter system.”
Ferrin added that the class has “completely taken over [his] life,” though the workload has allowed him to learn high levels of information quickly.
“It’s all so fast that you just do,” Ferrin said. “And you pop out the other side … and we all know how to make a decent experiment now.”
GOVT 83.21 not only sheds light on political misperceptions but also promotes a student-led approach to research and the self-actualization that emerges when true passion and investment are involved. Perhaps the best way to learn, as Ferrin said, is to “just do.”
