How to Psychoanalyze a Twitter Feed

Psychologists explore the possibility of using microblogs for psychoanalysis.

Posted Nov 18, 2019

Rawpixel / Pxhere
Source: Rawpixel / Pxhere

It is no stretch to say that our activity on social media provides a window into our psyche. Most of us are eager to share our hopes, fears, frustrations, and accomplishments on social media—and we do so at a rapid-fire pace. 

But it is possible that the information we willfully put out into cyberspace could change the delivery of therapy and counseling services in the future?

A new article published in the journal Frontiers in Psychology explores the idea that the psychoanalytic method could be reasonably applied to a person's digital activity on microblogging sites such as Twitter. The authors surmise that this semi-Orwellian future might be closer than we think.

"A key element of [psychoanalytic] therapy sessions is that the therapist does not actively guide the patient in a certain direction, but the patient talks freely about any topic only internally guided by her/his own unconscious mind," state the researchers. "Overall, this provides us with a practical connection between psychology and the data provided by microblogs."

The researchers argue that the sheer volume of content produced on social media platforms is reason enough to believe in the possibility of a "Twitter-fueled" method of psychoanalysis. According to the researchers, an especially active Twitter user might contribute the equivalent of a hundred book pages per year in 140-character tweets. Given that psychoanalysis has already been applied to literature in the form of psychoanalytic literary criticism, it is no great leap to assume that a similar analysis could be conducted on one's social media feed.

In fact, researchers are already using information from people's social media posts to predict emotional states. The graph below, for instance, depicts a running total of the emotions expressed in Bill Gates' tweets. (Bill Gates joined Twitter in 2009 and has contributed over 3,000 tweets since joining.) The most interesting outcome of this type of analysis, in the researchers' view, is the way the data might be used to examine personality change over time. In the case of Bill Gates, it is clear that the positive emotions reflected in his tweets have decreased over time while the negative emotions in his tweets have remained more stable (although they show some decline as well).

 Emmert-Streib et al. (2019)
Sentiment analysis of Bill Gates' Twitter activity.
Source: Emmert-Streib et al. (2019)

The researchers believe that, for all the benefits that might come from the digitization of clinical treatment methods, there are risks as well. For one, the amount of psychological information that can be gathered from someone's digital footprint raises serious privacy concerns. For instance, a study conducted by researchers at Columbia University found that constructing psychological profiles based on people's social media use increased advertising click-through rates by up to 40% and purchase decisions by up to 50%. 

The authors write, "We think that information about the human personality of individuals is among the most sensitive information one can obtain about someone. For this reason, our results suggest that users of microblogging platforms should be made aware of such threats so they can decide if they are willing to provide such information to the public."


Emmert-Streib, F., Yli-Harja, O., & Dehmer, M. (2019). Utilizing social media data for psychoanalysis to study human personality. Frontiers in Psychology, 10, 2596.