A new study by a graduate student with the Harvard University psychology department reports that social media posts can provide clues as to whether or not a user is likely to suffer from depression.
In a CBS News article on the study, lead researcher Andrew Reece said that “computer software designed to scan photos for these hidden signals accurately diagnosed people with depression seven out of 10 times.”
The hidden signals and clues tended to exist in the users’ choices of coloring and subjects. Depressed users’ posts contained more grays and darker blues, and often had the Inkwell filter applied, which turns color photos black and white. Non-depressed users tended to use a brighter filter called Valencia. The study also noted that photos posted by depressed individuals contained fewer faces.
Is a 70% accurate diagnosis rate notable? Absolutely. In fact, according to Reece and other study authors, the computer program’s detection rate is more accurate than that of general practitioners who average a correct depression diagnosis only about 42% of the time.
Why It Matters
More and more studies are drawing a link between social media use and stress and negative feelings. And with all generations (not just the assumed millennials) spending an increasingly large percentage of their time using social networks, any tools that may help identify at-risk users are welcome.
The article and study authors note that questions remain, specifically nagging ethical and privacy-based ones. If better diagnosis of depression and potentially suicide prevention are at the heart of the matter, hopefully these questions can be answered sooner rather than later.