On August 23rd 2017, Google announced that it would direct users in the US who searched for terms related to clinical depression to the PHQ-9 questionnaire. The PHQ-9 is a clinically validated survey that tests for depression. The goal of this experiment is to encourage sufferers of depression to seek help and treatment.
Google’s mechanism is a fairly simple implementation of the idea that sufferers of mental health issues can be detected and targeted based on their online behavior. A 2017 report by the Australian claimed that Instagram and Facebook are tracking users’ mental health status, but didn’t allow advertisers to target users based on this.
Various researchers have proved this to be possible. Striking examples are De Choudhury et al. (2013), who used twitter data to predict postnatal depression prior to birth, or Katikalpudi et al. (2012) who use browsing meta-data to predict depression, without even looking at browsing content. In my own research, I was able to predict clinical depression with 22% chance-corrected accuracy (64% overall accuracy) based solely on Facebook Likes.
There are two sides to this coin. On the one hand, most would find it immoral to target people struggling with mental disorders for commercial gain. On the other, 300 million people worldwide suffer from depression alone. Only 24% of college students ever receive help for a depression, and targeting these people to encourage them to seek help can create tremendous social value.
My question to readers brings the topic back home: How would you feel about having your online behavior tracked, so that if you ever unknowingly start experiencing symptoms of depression, a system can warn you to go and seek help?
De Choudhury, M., Counts, S., & Horvitz, E. (2013). Predicting postpartum changes in emotion and behavior via social media. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems – CHI ’13 (p. 3267). http://doi.org/10.1145/2470654.2466447
Katikalapudi, R., Chellappan, S., Montgomery, F., Wunsch, D., & Lutzen, K. (2012). Associating internet usage with depressive behavior among college students. IEEE Technology and Society Magazine, 31(4), 73–80. http://doi.org/10.1109/MTS.2012.2225462
The Australian Reporters. (2017). Facebook-targets-insecure-young-people-to-sell-ads.
Dear Yannick,
Very interesting topic! Thank you for writing this post. As an answer to your question: I would not mind at all getting a warning from a system I use when they have indicators that strongly point towards a depression. I think widely adopting this kind of positive use of information tracking would help skeptics all around the world feel more comfortable with sharing their information as it actually becomes beneficial to the user. Our data is tracked for commercial purposes such as personalized advertising anyway: Why not add a positive, actually beneficial use to it for the user?
Another point of discussion here:
Would a depressed user actually take the effort to fill in the PHQ-9 form, or would they rather deny having any problems and therefore continue living with their disease, because depression (or any mental illness in general) is still taboo for many people? Wouldn’t unconscious/subtle treatment be better, especially for cases where no medical treatment is required (of course after the user has been asked to agree that his/her information will be used to help improve their health)? If yes, how would this be possible?
Hi Yannick,
Interesting topic! I believe tracking people and warn people of diseases like a depression isn’t such a bad thing. But on the other hand, how far can this go? When I place 3 pictures on my Facebook of myself with a bottle of beer. I don’t want Facebook to send me a message: ‘You’re drinking too much beer!’ I think that we as a society have to define the boundaries when a website may intervene on your health condition.
Interesting study. Although there is always a commercial incentive for Google, I think it’s a positive development that individuals suffering from depression can be helped rather sooner than later based on their search terms. I would imagine that Google will hold back and limit the ads/commercial things place, knowing that all commercial content they attach to medical advice will be scrutinised by the public.