Study finds politics get in the way

"Our findings have implications for the spread of false news, for political popularization and for social divisions."

By GIL HOFFMAN
Computer (photo credit: REUTERS)
Computer
(photo credit: REUTERS)
Learning about someone's political beliefs interferes with a person's ability to assess expertise, as people judge like-minded peers as being more expert in fields completely unrelated to politics, according to a new University College London study released this week.
In the paper, published in Cognition, the researchers found that people turned to peers with similar political views for help on a task that had nothing to do with politics, instead of seeking help from someone who was doing better at the task but didn't share their political leanings.
"Our findings have implications for the spread of false news, for political popularization and for social divisions," said the study's senior author, Professor Tali Sharot. "If we are aware of a person's political leanings, for example on social media, we will be more likely to accept their take on a myriad of issues without scrutiny."
The researchers from UCL and Harvard University tasked 97 study participants with categorizing geometric shapes based on the shape's features. The participants were told whether their answers were correct and were also shown the answers of four other "co-players" who had completed the same task.
The co-players were actually computer algorithms designed to perform the task either extremely well or relatively poorly.
While completing the shape task, participants were also asked whether they agreed or disagreed with a number of political statements, such as "Immigration gives a boost to the national economy." The participants again saw how their "co-players" answered these questions. The researchers predetermined that some of the "co-players" would mostly agree with the participants on political issues while others would mostly disagree.
In a second stage of the experiment, the participants were shown a new set of shapes and were asked to categorize them correctly, with a financial reward promised that depended on their performance. For each shape they had an opportunity to view the answer of one of the co-players before making their final decision.
To increase their chances of answering correctly, the participants should have chosen to hear from those who had already demonstrated expertise on the task, regardless of political views. Yet instead, the participants sought and then followed the advice of the politically like-minded, even when a politically dissimilar co-player was better at the task.
"When we examined participants' impressions of the co-players, we found they overestimated how good the politically like-minded were at the shape categorization task," Sharot said. "This misperception drove the participants to seek advice from the politically like-minded."