If Cosmo and Buzzfeed have taught us anything, it’s that we love personality quizzes. Sure, many of them aren’t valid measures of personality, but it can still be fun to find out what kind of Disney princess you are or what your food truck preference says about the way you handle rejection in life. 

Vintage Quiz from “The Girl Friend and the Boy Friend” Magazine May 1953 – via Envisioning the American Dream

But the logic behind these fun quizzes can has a big impact in social science, because they are all based on looking for patterns in how people answer questions. We can reverse-engineer the process; instead of going in with a set of personality types and designing a survey, researchers can use a method called Latent Class Analysis to look at completed surveys and see which patterns of answers emerge from the data. By comparing those patterns to existing theories, they can come up with new categories that explain how people think, especially people who fall in between the strong or obvious categories.

The Pew Research Center has done this with different styles of religious experiences, and you can take a quiz to see which type best fits you. 

Bart Bonikowski and Paul DiMaggio use this approach to identify different kinds of nationalism in the U.S. There are ardent nationalists and people who are disengaged from nationalism, but the middle is more interesting. Between these two groups, there are also people with relatively moderate national pride who still think only certain people are “truly American,” and there are folks who have higher national pride, but a more inclusive vision of who belongs.

I also used this method in a recent paper with Jack Delehanty and Penny Edgell looking at different kinds of religious expression in the public sphere. In a new paper coming soon, our team also finds patterns in how people think about who shares their vision for American society.

Religion, nationalism, and even racism? These are heavier topics than the typical personality quiz covers, but the cool part about this method is that it is less intrusive than directly asking people what they think about these topics. When we ask simpler questions—but more of them—and then look for patterns in the answers, we can learn a lot more about what they actually think.

Evan Stewart is an assistant professor of sociology at University of Massachusetts Boston starting in the fall, 2019. You can follow him on Twitter.