Feeling #Downinthedumps? Twitter Already Knows

Tom Wang/Shutterstock

In a recent published article, researchers found that there are certain characteristics of social media posts that can indicate whether a person is depressed before the person actually becomes depressed (or at least aware that they are). The authors had subjects complete an online questionnaire to assess their level of depression and then asked for information about the onset of their depression. Researchers then asked for permission to view the participants' Twitter posts from the year leading up to the perceived onset of the depression.

The results indicated that depressed people, in general, posted the most on Twitter late at night, while non-depressed people posted most in the early evening. This difference may reflect the sleep difficulties common in depressed people and the increase in depressive symptoms that is often experienced at night.

The authors also found that, in general, depressed people expressed a lot of negative emotion in their posts, posted less frequently, described their depression symptoms, and used first person pronouns (I, me) to a larger degree than other people did in their posts. They also had smaller Twitter social networks (both fewer followers than average and following fewer people than others). The interpretation is that people who are depressed are posting about their own experiences to express their sadness to a small, close-knit group of people. It also indicates that people's depressive symptoms are reflected in their social-media posts.

Closely studying participants' posts during the time leading up to the reported onset of their depression, researchers found that the people began to post less frequently, reply to others less frequently, use third-person pronouns less often (he, she, they), and increase their expression of negative emotion and their use of first-person pronouns—as well as their use of expletives and words generally related to depression.

The findings suggests that as depression symptoms increased, subjects became increasingly self-focused and expressed their symptoms more in their social-media posts. They were also isolating more than before, at least on social media.

Using these types of factors, the researchers found that they could actually predict which people were getting depressed based entirely on the content of their Twitter posts in the time leading up to the perceived onset of their depression.

More broadly, this research suggests that social media posts may serve as a way to identify people who are at risk for depression through the use of software applications programmed to scan posts for risk factors and then potentially send personalized, confidential alerts to those individuals.

So maybe someday in the future, you could receive personalized mental-health warnings from a computer program based on your late-night tweets.

What do you think—would you find that helpful or intrusive?

Copyright Amy Przeworski.

This post and all portions of this post may NOT be duplicated or posted elsewhere (including on other websites) without permission of the author. However, a link to this post may be posted on your website without the need to request permission.

Follow me on Twitter @AmyPrzeworski https://twitter.com/AmyPrzeworski

Like my Facebook page: https://www.facebook.com/DrAmyP

Follow me on Pinterest: http://pinterest.com/amyp100/

Check out these related posts:

When You Feel Like Giving Up. A post about feeling helpless.

Embracing the Worst. A post about learning to be happy when everything seems wrong.

Becoming an Optimist: A post about discovering your optimistic side.