Scientists have explained the phenomenon of mood deterioration due to social media algorithms

Social networks have long used algorithms that select content based on user interests. However, a new study has shown that personalized recommendations can not only reflect a person's emotional state, but also enhance it. Scientists have found that people with more severe symptoms of depression are more likely to see negative content about relationships, and their brains react to it differently than to ordinary popular videos. How algorithms form a personal feed, why it's not just about the time spent on social networks, and whether it's possible to break this vicious circle — in the Izvestia article.

The brain was first studied while watching a personal feed

The study was conducted by psychologists from the University of Texas at Dallas. Unlike most previous studies, which were based on surveys of participants, scientists observed what happens to the human brain directly while watching content on a social network. For this purpose, electroencephalography (EEG) was used, which allows recording brain activity in almost real time. The results of the study are published in the journal Computers in Human Behavior.

The peculiarity of the experiment was that the participants were not shown videos pre-selected by the researchers. Instead, they reviewed their own recommendations, which the algorithms had already generated based on their habits and interests. The researchers then compared the brain's response to the personalized feed and to ordinary popular videos that were not related to a particular person's preferences.

Study co-author Stacy Warren, in turn, added that the EEG allows you to capture information processing by the brain in just milliseconds. Thanks to this, scientists have obtained objective data about what happens to a person at the moment of viewing the recommended content, and not after it, when impressions can already be distorted by memory.

Scientists have found a link between smartphone use and human behavior

The algorithms reflect the user's emotional state

The study showed that participants with more severe symptoms of depression were more likely to receive recommendations with negative emotional content. The most common topic of such videos was relationships between people — romantic conflicts, quarrels of friends, breakups and other emotionally difficult situations.

At the same time, the brain activity of these participants differed from the reaction to ordinary viral videos. The researchers found changes characteristic of depressive states, in particular the features of the so-called frontal alpha asymmetry, an indicator that is often used in the study of emotional information processing. The authors believe that algorithms do not create depression by themselves, but they can enhance an existing emotional state of a person.

The researchers also found that users with more severe symptoms of depression saw fewer positive videos related to support, friendship, or pleasant social interactions. Instead, their feed more often contains materials with negative emotional connotations.

From "What I eat in a day" to fast food: how social media is changing the diet of teenagers

Why wasn't screen time the main factor?

In recent years, many studies have linked the deterioration of mental health primarily to the amount of time a person spends on social media. However, new work shows that this indicator alone may not be informative enough.

The authors specifically took into account screen time as an additional factor in order to exclude its influence on the results. It turned out that what is much more important is not the length of use of social networks, but the nature of the content that the user receives thanks to algorithmic recommendations. One person can watch neutral or positive videos for several hours, while the other will face a lot of emotionally difficult posts in less time.

That's why the researchers suggest shifting attention from simply limiting screen time to analyzing how recommendation systems work. In their opinion, algorithms are able to gradually form an information environment that supports the user's existing emotional state.

Beauty by algorithm: how social networks are changing teenagers' beauty habits

Is it possible to break the "vicious circle" of recommendations

According to the authors of the study, the main task now is not to completely abandon social networks, but to train users to interact more consciously with algorithms. The researchers believe that it is important to understand why the feed shows exactly this kind of content and how your own actions affect further recommendations.

Scientists pay special attention to teenagers. As Professor Pamela Tang noted, a few years ago, the main recommendation was to reduce the time that children spend on social media. However, today this approach no longer seems sufficient.

Researchers have already started a new long-term study involving adolescents aged 13-16. They plan to track how algorithmic recommendations affect the emotional state of young users over a long period of time and whether strategies can be developed to reduce this effect.

Thus, the work shows that the impact of social media is determined not only by the time spent in front of the screen. An equally important factor is what kind of content algorithms decide to show to a particular person and how much this personalized feed corresponds to their current emotional state.