Â鶹ÒùÔº

Rigid and negative thought patterns linked to increasing political polarization online

The ideological divide between opposing political groups has been drastically increasing in various countries worldwide. This phenomenon, known as political polarization, can lead to greater social division, extremism and political violence.

Researchers at Indiana University and City College of New York recently carried out a study aimed at better understanding the psychology of political polarization and, more specifically, the thought patterns associated with extreme and opposing political views. Their findings, in Communications Psychology, unveiled a link between the rising political polarization online and distorted ways of thinking often associated with some mental health disorders, including anxiety and depression.

"The idea for this paper was born during the 2023 Conference on Complex Systems in Brazil," said Matteo Serafino, senior author of the paper. "In a keynote, Johan Bollen presented research showing a sharp rise in linguistic markers of cognitive distortions—rigid, negative thought patterns often linked to anxiety and depression—in public discourse over the last two decades. This immediately caught my attention.

"Our team had just published a paper on the growing intensity of political polarization, and I began wondering whether these trends might be connected. Later that evening, over dinner, I shared the idea with Andrew Edinger, a Ph.D. student in Johan's lab."

This conversation between Serafino, Bollen, Edinger and their colleague Hernan Makse ultimately initiated the research project that led to their recently published paper. The objective of this project was to explore the possibility that distorted, emotional thinking might be contributing to, or could even help to explain, the growing extremity that characterizes online political posts.

"We started by collecting over 85 million tweets from people talking about the U.S. presidential elections in 2016 and 2020," explained Serafino.

"Then, we looked for specific phrases in those tweets that match patterns of distorted thinking, like 'They're all evil' or 'Everything is ruined.' These come from cognitive behavioral therapy (CBT) and are known as cognitive distortions—ways of thinking that are overly negative or extreme."

To measure political polarization in online posts, the researchers analyzed political tweets, with a key focus on what online users retweeted and who had originally posted these tweets. This allowed them to estimate each user's political inclinations (i.e., left or right) and how strongly they supported one side or the other.

"Using these methods, we could compare, for each user, how much distorted thinking they showed in their tweets and how politically polarized they were," said Serafino. "We then used a statistical method called linear mixed-effects models to analyze these patterns over time."

Remarkably, Serafino and his colleagues found that between 2016 and 2020, the use of rigid or negative language by users on X increased by over 40%. Moreover, people who appeared to become more politically polarized over time appeared to use more cognitively distorted language in their tweets.

"Distorted thinking appeared to precede polarization in some cases," said Serafino. "This suggests that how we think—emotionally and cognitively—may be a key driver of polarization, not just what we believe. Of course, our observations do not imply causality."

In the future, the findings of this recent study could help to address the growing political division in various countries worldwide, not only by introducing new policies or fact-checking measures, but also by employing psychological tools. For instance, therapeutic exercises employed in cognitive behavioral therapy (CBT), a type of psychotherapy found to be promising for treating anxiety and depression, could also help people to recognize and change the distorted thought patterns they exhibit when engaging in political debates.

"This opens up new ideas for digital platforms, educators, or even civic campaigns to promote healthier, more constructive public discourse," said Serafino. "We're currently interested in exploring whether the same patterns we found on Twitter also appear on other social media platforms, like TikTok, Reddit, or Facebook. These platforms differ in how people communicate and comparing them could help us understand how distorted thinking spreads in different online environments."

As part of their future studies, Serafino and his colleagues also hope to better understand the extent to which cognitive distortions cause political polarization, or vice versa. So far, they have found a correlation between distorted thinking and political polarization online, yet they have not yet established a causality between the two.

"We did find hints that distorted thinking may come before polarization, but future research using more controlled or experimental methods is needed to confirm that," added Serafino.

"We're also very interested in how these patterns differ across generations. Some studies suggest younger people may be more susceptible to cognitive distortions online, but this hasn't been thoroughly tested. Understanding whether and why that's true could have big implications for education, mental health, and civic engagement."