Indiana University researchers have taken an interdisciplinary approach to studying social media impacts on mental health, discovering an intervention that could help curb the U.S. depression epidemic, according to a new study.
Eeshan Hasan is a doctoral student in the College of Arts and Sciences. Photo courtesy of Eeshan Hasan
It builds on research from Lorenzo Lorenzo-Luaces, an associate professor and associate chair in the Department of Psychological and Brain Sciences; and Johan Bollen, professor and chair of informatics in the Luddy School of Informatics, Computing and Engineering and the College's Cognitive Science Program. Both are co-authors on the report, along with Gunnar Epping, a doctoral student in the Department of Psychological and Brain Sciences and the Cognitive Science Program.
Bollen and Lorenzo-Luaces had previously found that depressed people are more likely to produce content that exhibits higher levels of negative, rigid and extreme language on social media, also known as distorted thinking. In the PNAS Nexus study, the team investigated whether cognitive behavioral therapy could train social media users to recognize posts that exhibit distorted thinking before interacting with it, in hopes of drastically reducing the amplification of such content.
Reducing engagement with distorted content online may ultimately prevent algorithms from suggesting such content.
"A lot is unknown with regard to how people interact with social media, the association with social media and depression, and whether something can be done to change it," Hasan said.
The researchers had participants take part in a brief, "one-shot" cognitive behavioral therapy intervention, which helps people identify and reshape distorted thinking into more balanced and helpful thoughts. Lorenzo-Luaces produced a brief overview of cognitive distortions, along with illustrative examples, that were presented to the participants.
The team had also designed a social media interface that mimicked the popular social platform X. They then generated a set of distorted and non-distorted posts, which included interactive buttons.
"We wanted something that actually looked like X," Hasan said. "When you clicked a heart, the heart would light up. It was designed with these ecological factors in mind."
Researchers designed a social media interface similar to X's to display distorted and non-distorted posts. Photo courtesy of Eeshan Hasan
"Our study found that individuals were extremely good at identifying distorted content with very little training," Hasan said. "After the intervention, there was a huge decline in how much people liked and interacted with it.
"An unexpected finding was that individuals with greater depression severity interacted - liked and retweeted - with distorted social media content more than the control group did. Fortunately, our intervention reduces the interaction with distorted content for depressed and non-depressed individuals."
This experimental distortion study is the first of its kind, attacking longstanding questions using a combination of cognitive science, clinical psychology and informatics. Social media is usually studied using large-scale datasets. But with this experimental set-up, users are interacting with the same content, which makes it possible to infer how individual factors influence behavior.
"Computational social scientists are increasingly integrating data, theory and experiment-driven approaches to address long-standing problems in psychology and the social sciences," Bollen said.
Jennifer Trueblood is the Ruth N. Halls Professor of Psychological and Brain Sciences and director of the Cognitive Science Program at IU Bloomington. Photo courtesy of the Department of Psychological and Brain Sciences
"This work shows how bringing together different perspectives can lead to creative, practical solutions for real-world challenges like social media and mental health," Trueblood said.
Despite this breakthrough, there are still hurdles in studying social media and mental health. One is access to data.
For example, when this project started in 2022, X's data access was much more open, and researchers could study the networks of its users. That has since changed, with X putting expensive paywalls in place.
Still, researchers are hopeful that others will adopt similar cognitive behavioral therapy interventions at the collective level - like at community organizations, workplaces or schools, or even on social platforms themselves - so more people can identify distorted thinking and limit their engagement with it online.
"Psycho-educational interventions like ours could be dispensed at scale on social media platforms to counteract the effects of distorted language," Hasan said. "This would improve quality of life and bring about real change to society."
