The language you use on social media can reveal depression

Language use in social media can be a useful tool for social scientists, because it reflects living conditions in areas the posts originate from.

“There is a correlation between social inequalities and language patterns in social media,” Lucas M. Bietti says.

He is an associate professor at NTNU’s Department of Psychology.

Researchers find the most depression and the least adaptability in poorer areas that are home to a larger proportion of minorities.

Lower adaptability is a term used to describe how people react to big changes, such as losing their jobs.

30 million posts

Researchers studied 30 million X/Twitter posts from the United States. They compared the language used in these tweets to the living conditions in the counties from which the tweets were posted.

“How we use language reveals something about ourselves and the conditions in which we live,” Bietti says.

He collaborated on this study with postdoctoral fellow Eric Mayor from the University of Basel. The results have now been published in the journal Heliyon.

Ethnic minorities are more prone to depression

“The language patterns reveal that, on average, people from ethnic minorities show clearer signs of depression and lower adaptability,” Bietti says.

The researchers also found correlations between how rich or poor an area is and the proportion of ethnic minorities that show signs of depression and poor adaptability.

“The poorer an area is, and the larger the proportion of ethnic minorities, the clearer the language signs we find of depression and low adaptability in this area,” he says.

In general, many ethnic minorities in the United States have lower levels of education and lower income than the majority.

Reflects reality

Bietti and Mayor have previously investigated the language used on X/Twitter in relation to the time of the week the messages are sent. That study showed that much more positive language is used in the afternoon than in the morning, which is the busiest time of day for most people.

Our emotions generally get stronger outside of office hours, and we like ourselves the most as the weekend approaches.

Seen collectively, social media posts reflect reality.

Useful tool

“Analysing language patterns in social media posts is a useful method that can help us understand trends in mental health,” says Bietti.

The method enables us to compare people’s mood in a specific area at different times and over time.

“The method is particularly useful when public health data from an area are lacking or not good enough,” he says.

Among other things, the method can help us estimate the impact that significant events have on mental health.

“The indicators can show us where we should implement various health campaigns and social initiatives, especially in areas where living conditions are poorer than elsewhere,” Bietti concludes.

Reference:

Mayor, E. & Bietti, L.M. Language use on Twitter reflects social structure and social disparities, Heliyon, vol. 10, 2023. DOI: 10.1016/j.heliyon.2023.e23528

———

Read the Norwegian version of this article at forskning.no

This content is paid for and presented by NTNU

This content is created by NTNU's communication staff, who use this platform to communicate science and share results from research with the public. NTNU is one of more than 80 owners of ScienceNorway.no. Read more here.

More content from NTNU:

• 

  Artificial nose can sniff out damaged fruit and spoiled meat

• 

  Want to be the best? It’s all in your mindset

• 

  Millions of Earth's species remain unknown to us. Can we still identify them?

• 

  AI expert: We should not have AI that exploits people's weaknesses

• 

  Afghanistan veterans are more prone to anger

• 

  Could a digital driving instructor offer the same quality as a human, but at a lower cost?