CU Boulder pollster emphasizes need for rigor in political surveys
Carey Stapleton, PhD candidate at CU Boulder, was survey lead for the American Politics Research Lab’s Colorado Political Climate Survey
The nature of political polling is changing in the United States, and Carey Stapleton, a PhD student in American politics and methodology at the University of Colorado Boulder, is equipping his undergraduate Survey Design and Analysis class, to meet these changes head-on.
Carey Stapleton
The result of this class was the recently released Colorado Political Climate Survey. Through creating a class centered around surveys and polls from CU’s American Politics Research Lab (APRL), Stapleton is teaching his students not only to think critically about politics and survey design, but also how to analyze the data of a large-scale poll.
Stapleton, who has “a substantial background in survey design,” was approached two years ago by Professors Scott Adler, Anand Sokhey and David Brown to create a class in survey design and political polling.
“We all sat around and developed the syllabus for the course, and we finally started teaching it last fall.” He says the class has grown by a factor of three in two years, and is continuing to pick up speed.
Stapleton says the Colorado Political Climate Survey took a couple years to perfect. “We tried to get both information that's relevant to Colorado specifically, but also kind of how Coloradans feel about broader national issues.”
It’s why the poll features questions ranging from intent to vote in the gubernatorial election to opinions on sports gambling, recreational marijuana laws, and if “Dreamers” should be allowed to remain in the country.
In 33 pages packed with data and analysis, the survey “is designed to gauge the public’s political and partisan leanings,” and correctly predicted the outcomes of the gubernatorial and congressional races, but revealed inconsistencies in polling with issues on the ballot like Propositions 73, 74 and 112.
In this midterm election, pollsters stood poised to redefine the relevance and format of political polls and surveys after the stunning upset in the 2016 presidential election defied the predictions of many major polls.
I tell my students all the time, ‘I don't care what you think in this class. I care that we do science in the appropriate way so that we can get to the reality, rather than what we want to be true.’”
On the nature of polls and the argument that they might yield biased information, Stapleton argues, “We're not here to tell people what to think about the results. We do this to get this information out there.” His responsibility is to be an impartial arbiter of the reality that the poll is reporting, he says.
Stapleton argues that transparency is key among polls and pollsters, citing the intersection of science and politics as an area that should be free from biases. He observes: “I tell my students all the time, ‘I don't care what you think in this class. I care that we do science in the appropriate way so that we can get to the reality, rather than what we want to be true.’”
Undergrads “are just now learning to think things through and come to conclusions,” which is why making sure students analyze the reason certain conclusions are reached is critical.
In his own research, Stapleton focuses on the role anger plays in American politics: “Emotions are critically important to our daily lives, and one thing that politicians can do is use their own emotional projections to influence how the public feels.”
The prevalence of emotional manipulation in today’s politics only reinforces the importance of unbiased, science-based surveys, he says. According to Stapleton, the goal of a survey like APRL’s is to create a more nuanced view of politics among voters and his students.
“I think that's our goal ultimately,” he muses. “Do people leave your class with a better grasp of reality and the truth? My hope is they do. And then do they take that and apply it in the real world? My hope is they do.”