What Your Instagram Posts Reveal about Your Mental State (and Why That's Important)
Instagram has long been known as a place for your best photos, for polished, pristine images that showcase you or your subjects in the best light, as opposed to the more in-the-moment, DIY feel of other networks. That focus on perfection became a point of contention last year when Insta-famous model Essena O'Neill spoke out about the lengths she would go to in order to present a perfect life via her Instagram account. O'Neill even went so far as to go back and edit the captions on her old posts to reflect the actual truth behind her "in the moment" shots.
O'Neill's stance highlighted a growing issue, particularly among younger users - that teens are often succumbing to peer pressure and working to present an unrealistic, and largely unhealthy, view of how they live. That then has a flow-on effect to other teen users who invariably compare their own lives to those being presented via the content of their friends. But what's posted on social, as noted by O'Neill, is very often not representative of people's real life experiences. There are far more posts about triumphs than there are about the struggles that go along with them.
The links between social media and psychology have been examined in many ways - one recent study sponsored by the National Institute for Mental Health identified a "strong and significant association between social media use and depression" among young adults, another by the University of Missouri found that Facebook use, in particular, could be linked to depression, particularly the "surveillance" element, where people use the site to check up on how their friends are doing and compare what they see to their own life. For all the positive benefits that social connectivity brings, the relative links between social media use and psychological concerns are undeniable, at least in some capacity, and when you also consider that suicide is the second leading cause of death for people aged between 10 and 24 in the U.S., the issue takes on even more urgency. Put simply, it's something that we all need to pay attention to and work towards a solution where possible.
In line with this, a new report by researchers from Harvard University and The University of Vermont takes a look at Instagram, specifically, and how the content people post on Instagram can be reflective of their mental state. But rather than examining their findings in an observational capacity, the research team are looking to use the data to detect users in need to assistance by creating an algorithm that can detect mental issues and connect people to relevant support services.
To do this, the researchers asked 166 Instagram users for permission to analyze their posts and also asked whether or not they had a diagnosis of clinical depression from a mental health professional. What they found was that people with depression over-indexed in several categories in regards to their Instagram post composition.
For example, people with depression prefer darker colors and more grays or blues than non-sufferers.
Depressed people also tend to tend to have more comments on their posts, but fewer likes, while they were also less likely to use image filters - though when they did, Inkwell, which makes everything black and white, was their primary option of choice.
The researchers also found that "depressed participants were more likely to post photos with faces, but had a lower average face count per photograph than healthy participants". This is likely indicative of these people having fewer social interactions.
The findings enabled the researchers to build an algorithm that was then used to analyze further Instagram images to determine whether it could detect depressed users based on their posts.
"When the machine gave a depression marker, it was right about 54% of the time, compared to unassisted primary physicians who correctly make a depression diagnosis about 42% of the time."
And while that level of success may seem low, it's still 25% better than human detection rates, which suggests that such a system, with further training, could one day be able to automatically detect mentally ill patients who may be in need of assistance based on their Instagram posts alone.
This is not the first time such a system has been proposed. A report released by Microsoft in 2013 used Twitter data in the same way, determining that tweet information could detect individuals with depression or signs of distress with 70% accuracy. Facebook data, too, can be used for the same purpose. In isolation, each set of results shows promise, but in concert, if there were a system established that could take into account indicators from all social platforms in this way, there could be a an extremely powerful tool in the making which could detect mentally ill individuals and connect them to relevant support services ahead of time.
The other option, of course, is to demonize social media and call for a ban, particularly among younger users. And there's some merit to that, but the reality is that social platforms are here to stay. Even if you were to remove one, another would take its place. And while social media can amplify the pressure on young users to present a perfect image of their life, there are also significant benefits to social connection. Isolated and disillusioned people can find communities of like-minded folk online, the ways and means in which social can concurrently be used for mental benefit also need to be taken into consideration and measured on balance with the potential negatives.
Through research like this, the hope is that we can ramp up the beneficial uses of such data to ensure more people get help when they need it, rather than being brought down further by online communication. It's a difficult issue, and as has been well-documented, one that Twitter, in particular, has long struggled to tackle. But through collaborative research and enhanced data systems, we're hopefully on track to being able to help more than hurt through our wider social interactions.
Follow Andrew Hutchinson on Twitter