Diagnosing Depression, By Watching How You Surf The Web

According to new research, basically everyone who knows how to properly use a computer should seek counseling.

You’re downloading a movie from Bittorrent. Maybe it’s not exactly legal behavior, but other than Michael Bay getting by with five less dollars to adorn his money fireplace, no harm done, right?

Maybe…maybe not. A new study by the Missouri University of Science and Technology examined the internet patterns of 216 students, about 30% of whom met guidelines for depression. And what it found was that depressed students tended to have different browsing habits than their happier counterparts.

The use of P2P sharing—Bittorrent and similar file exchange programs—was a correlative indicator of depression, as was the practice of checking email more often. Another big indicator was exhibiting high amounts of what’s called "flow duration entropy"—otherwise known as multitasking—switching between disparate apps quickly. The study’s authors make no claims that these behaviors cause depression. Rather, P2P downloads and hyper multitasking are just potentially reliable symptoms of it, like the digital version of lethargy and weight gain. Their next step is to develop software to monitor web usage that could alert someone if they were exhibiting depressed behaviors online—maybe even alert a counselor in the process, too.

The mind races when you think what could happen if our computers weren’t just watching our clicks, but actually watching our behavior through face tracking, behavior analysis, and whatnot. Aside from the privacy concerns to such an idea (and there are many!), it does open the doors to a very interesting field in diagnostic research. My doctor sees me for about 10 minutes at a physical each year, as I stammer out the short list of aches I can remember. My computer follows my habits twelve hours a day in my natural environment. It could potentially diagnose all sorts of things my doctor will never see.

How long is it taking me to read articles? Maybe my eyes need to be checked. How fast am I writing emails lately, and are my typos getting worse? Maybe around 2pm every day, my hands get tired and I should be taking a break. Or maybe these innocuous details could spot precursors for dementia. Move to my phone, and the possibilities get even bigger. Have I been texting my friends less? Does the GPS say I’ve been staying at home more…or getting lost to places I’ve already been?

The first response I had to the study was simply, "well, I guess computer science majors are more depressed than the rest of us." But the more I think about the potential gains—spotting looming trouble through steps we’re all taking all day anyway, without any cost beyond a little software development—the more I’m inspired by the potential of such apps…and the more I’m a bit freaked out to be typing right now, worried some little robot in my computer is judging every word while just waiting for me to click on another pathetic Thought Catalog link.

[Hat tip: NYT; Images: Stock Elements and Dimendrol68 via Shutterstock]

Add New Comment


  • Stacerella

    If you find that freaky and leaves you a tad paranoid, download and install the add-in for Firefox called Collusion. It's eye opening. Even after two sites I visited. 

  • whatbagley

    This study reminds me of another study that basically said that if you liked to stay up late and browse the internet that you were practically a genius. WE'RE ALL GENIUSES PEOPLE.

  • Excalibur

    What the heck is "correlative indicator"? Seems to me that correlation is being "stretched" to replace causality. So there is definitely no causality but there is "correlative indicator" (??). So as I understand the researchers cannot measure depression directly and thus the validity of this correlation is questionable. So the validity is dependent on a specific theoretical construct. Such research studies that move in the direction of diagnosing sometimes behavior are dangerous as at best the diagnosis or any conclusion of a person being depressed is subjective.I am curious to know the funding of such research. I have to bet that the person leading this research is from a baby boomer generation, who does not really understand the Gen x and y. So when an individual does not fit their definition of normal habit then the person is "definitely" abnormal! Further, I think the goal here is how to diagnose someone to be depressed and then hook them on drugs and therapy. Then tomorrow they will be denied certain jobs and possibly tagged to be depressed. By then even if the lowest was not depressed initially, now the person would be. Why? Because all this would lower the person's self esteem and thus affect their actions and thus their relationships.

    The fact is a lot of people are bombarded with information, intentionally or unintentionally. Gen x and y, view and process information very differently than baby boomers. There are more sources to learn anything from and also more sources to express to. That type of behavior is associated with information processing, and not necessarily emotional behaviors. Depression is a result of an emotional disbalance; it is not a result of informational processing habits of an individual. Please don't confuse average readers with terms like "correlative indicators". It misleads people into thinking that its a causality.