Northwestern researchers develop new model capturing effect of political environment on political opinion

Students+gather+to+watch+the+2016+presidential+debate.+The+new+model%2C+based+on+mathematical+equations%2C+will+help+show+how+populations+change+their+opinions+over+time+as+they+are+exposed+to+political+content+such+as+news+media%2C+campaign+ads+and+personal+exchanges.%0A

Daily file photo by Jeffrey Wang

Students gather to watch the 2016 presidential debate. The new model, based on mathematical equations, will help show how populations change their opinions over time as they are exposed to political content such as news media, campaign ads and personal exchanges.

Vivian Xia, Assistant Campus Editor

Two Northwestern researchers recently developed the first quantitative model that captures how politicized environments affect U.S. political opinion.

The model, based on mathematical equations, will help show how populations change their opinions over time as they are exposed to political content such as news media, campaign ads and personal exchanges. McCormick Prof. Daniel Abrams, the senior author of the study, said the group mapped political opinions from liberal to conservative on a scale from -1 to 1.

“You mostly need only one dimension to describe a person’s political beliefs in the U.S.,” Abrams said. “It’s not that you’ll perfectly capture everything they believe ― (the model will) capture a large amount of the variation. Most people who are left on one issue will also be left on other issues, so you don’t really need to have separate dimensions for each issue.”

Ph.D. candidate David Sabin-Miller, who led the study, said there are environments conducive to certain political viewpoints. If someone enters such a space and those viewpoints are close to their current position, they will move toward them. But, he added, if those viewpoints are too dissonant from their current position, they will actually go the opposite direction.

“The core of the model is the idea that people form their political ideology based on things they experience rather than any sort of inherent personality stuff,” Sabin-Miller said.

In a sphere as contentious as politics, he added, the tendency to move away from ideologies of the opposite party is a core aspect of tribalism and has become a really important dynamic in political advertising.

In conceiving the project, Sabin-Miller said he had an idea that he wanted to model how the population is polarizing and use math to “tackle this problem.”

Sabin-Miller said he, like many others, is very concerned about the current polarization of politics and has had some family experiences of having to be very careful about how to talk to people so as to not “make people defensive.”

“A lot of people probably have similar family experiences of feeling like politics is… getting in the way of just being people together, shared humanity kind of stuff,” Sabin-Miller said. “I really hope that we can figure out how some of these algorithms online are allowing us to live in totally different worlds by our own preference, and that’s having some unintended consequences on a massive scale.”

Santa Fe Institute postdoctoral fellow Vicky Yang, who used to work with Abrams, said one of the biggest contributions she sees from this model is the tracking of changes in belief without the use of a network model for mapping populations.

“There are limitations of studying opinion dynamics in the network form, because you have to really know a lot about how people are connected in order for you to analyze what happens,” Yang said. “And often we don’t know how people are wired together exactly, and this paper is offering a way for us to still draw conclusions without knowing these things.”

Email: [email protected]
Twitter: @vivianxia7

Related Stories:

Kellogg researchers develop predictive model for terrorist violence

NU researchers develop wearable device that helps monitor COVID-19 symptoms

Comments