Web Log: Study suggests tweets indicative of mental health

Posts with more negative language and swearing by users suggest ADHD

The study captured the tweets of 1,399 individuals with self-reported diagnoses of ADHD and compared them to a control set.

The study captured the tweets of 1,399 individuals with self-reported diagnoses of ADHD and compared them to a control set.

 

The language we use on Twitter may reveal whether or not we have attention deficit hyperactivity disorder (ADHD), according to a linguistic analysis of 1.3 million tweets carried out by researchers at the University of Pennsylvania. The study captured the tweets of 1,399 individuals with self-reported diagnoses of ADHD and compared them to a control set, revealing a pattern of less agreeable posts containing more negative language and more swearing by users with ADHD, who also posted more frequently than their counterparts.

While the study “found that social media language is predictive of ADHD” with an accuracy of 78.5 per cent, it also noted they had no way of knowing if the self-reported diagnoses were formally diagnosed by a mental health professional. They also had no way of knowing if these individuals had other underlying health conditions that could contribute to their behaviours.

This aside, there are the ethical implications of such research. Imagine a scenario where a potential employer or your insurance company dredges your tweets for signs of mental health issues or brain disorders without your knowledge. Not cool, big data. Not cool.

http://journals.sagepub.com/doi/pdf/10.1177/1087054717738083