Crowdsourced Peer-To-Peer Cognitive Therapy

Robert Morris, Stephen Schueller, and Rosalind Picard report an experiment which implemented assistance “cognitive reappraisal”, aimed at helping people work through and out of depression. This is modeled on therapeutic approaches which essentially help people reinterpret events in ways that offer options other than down spiraling negative thinking. The innovation is that they crowdsource the help, posting narratives and receiving instant feedback from strangers.

The mechanical Turks (“mechanical shrinks?”) responded to the posts to “(1) offer empathy, (2) identify cognitive distortions, or (3) help users reframe negative situations in ways that are more positive and adaptive”. The conversation is carefully designed, to the poster gets sympathetic feedback that first identifies “distortions” (e.g., over interpretations of another’s actions) and invites reinterpretation (e.g., alternative, less negative reasons for the reported events). Then they can “compare” their responses to suggestions from the crowd—and also get some alternative points of view on the situation.

Phew! Does this really work?

Evidently yes. The research shows that, at least for the situations tested, it seemed to improve the self reported behaviors, and the participants seemed to like the experience.

I can see some interesting implications and future research here.

I suspect that this technique will work best for people who are familiar and comfortable with such social media. I’d predict much less success for old geezers (over 40? Over 35? I dunno). This is not only a matter of fluency with the technology, it is also a matter of trust. You have to believe that the responses from the crowd are from real people, and that you can trust them.

I also suspect that getting instant responses is a key feature. This is not just a matter of keeping things in mind, it also indicates personal attention and therefore empathy. It’s surprising to get feelings of empathy from such anonymous and impersonal interactions, and there may be ways to make that even better. (Perhaps some kind of portrait of the “crowd”, to see that they are “people like me”, and are working selflessly because they are nice.)

By the way, I wouldn’t recommend automating this process (“Elizafication“). It might or might not work (and I’m sure there would be individual differences), but it might muddy the waters and make it harder to feel personally connected to any real people who are trying to help.  I know it will be tempting (it would be really cheap), but think carefully about the effects.

Finally, I would note the crucial implication of this work: if this kind of anonymous, instant feedback can help people change some of their negative thinking, then the same techniques could clearly push people into many kinds of destructive thinking. Many kinds of therapy have been used in ways that are not clearly in the interests of the participants, and this is no exception to that risk.

But more important, the basic technology is in wide use, unsystematically and with no thought to the well being of users. I refer to discussion groups such as Reddit—lots of instant feedback, offering reassessments and suggested alternative thinking. Even a few minutes exposure to such situations depresses me, so imagine what might happen to someone in a vulnerable condition? Not good.

I gather this idea is being developed further, so we will be hearing more.


  1. Morris RR, Schueller SM, Picard RW, Efficacy of a Web-Based, Crowdsourced Peer-To-Peer Cognitive Reappraisal Platform for Depression: Randomized Controlled Trial, J Med Internet Res 2015;17(3):e72
    DOI: 10.2196/jmir.4167


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s