Linguistic Markers Indicating Therapeutic Outcomes of Social Media Disclosures of Schizophrenia
This post summarizes a research paper about therapeutic outcomes of social media disclosures of schizophrenia that will be presented at the ACM Conference on Computer-Supported Cooperative Work and Social Computing 2018 (CSCW) on November 5th.
It was the 10th of October, World Mental Health Day. And, compared to the usual posts on travel, fashion and food, today Jules saw stories about peoples’ experiences with mental illnesses on social media. Browsing through the posts, she was shocked to find out that her friend Joe had been struggling with an anxiety disorder for the past few years.
Jules had always been skeptical about opening up and disclosing her own experiences. She was diagnosed with PTSD 10 years ago and since then she hasn’t been able to share her experience with anyone but her partner and therapist. “What was the point?”, she thought. She feared re-connecting to those memories, confronting the stigma, albeit her therapist advised her to show more emotional vulnerability. “What was the point?”, she thought. But, looking at Joe’s post, she was surprised at the number of people coming forward and sharing their own experiences, providing solidarity and support, breaking the stigma for themselves and others. She wanted to be a part of it, for Joe and for herself. So, with some apprehension, she went ahead and typed in “I’ve been diagnosed with PTSD…”.
The act of making ourselves known to others, also termed as self-disclosure, is an integral component of our social lives[1]. Particularly in experiences around mental illness that can be debilitating, lonely and overwhelming, self-disclosure is frequently adopted as a coping mechanism for emotional release, to break inhibitions, identify confidants, build trust and intimacy, and is known to have therapeutic benefits.
Historically, opening up about mental health experiences has been an activity that was situated between a therapist and client. Today, in contrast to such dyadic disclosures to a carefully selected listener, social media platforms have emerged as new arenas for broadcasting self-disclosures — referring to sharing personal, sensitive information in public contexts, often with large, invisible audiences [2]. With this shift from a careful listener to a large, public audience, how does the disclosure process and its expectant outcomes change?
How do we answer Jules’s question: “What was the point?”.
In our paper, we look at the specific case of schizophrenia disclosures on Twitter. We choose schizophrenia as it is one of the most stigmatized mental illnesses affecting more than 21 million people worldwide. And, sufferers are known to benefit therapeutically from intimate self-disclosures [3]. We ask the question: Can we quantify behavioral changes following disclosures of schizophrenia on social media? Focusing on one of the prominent goals of mental health disclosures, their expectant therapeutic outcomes, we ask: Do the therapeutic benefits of opening up manifest in behavioral changes on social media?
We collaborated with clinical psychiatrists in this work and used an expert validated dataset of 146 disclosures of schizophrenia shared on Twitter over a year-long period. We then propose a variety of psycholinguistic, structural, and topic-based linguistic measures to characterize and quantify the behavioral changes preceding and following the disclosures. Examples of the linguistic markers include psycholinguistic measures such as affective, cognitive, linguistic style, social and interpersonal attributes, linguistic structures such as readability, repeatability and complexity, domain specific content measures and topical measures such as topical coherence.
Therapeutic outcomes of social media disclosures
Along with significant linguistic differences before and after disclosures, we found indications of therapeutic outcomes following disclosures, including improved readability and coherence in language, future orientation, lower self preoccupation, and reduced discussion of symptoms and stigma perceptions.
We also found that following disclosures on Twitter, individuals express lowered stereotypy such as word repetitiveness, and demonstrate improved linguistic complexity, and topical coherence in the content shared on Twitter. To ensure these differences were not an effect of platform specific confounds, we quantified the same behaviors in a control group of Twitter users and found minimal differences.
Implications of Technology-mediated Self-disclosures
Expressive Writing on Twitter: Venting out negative emotions | Connecting to an audience | New model for writing therapy
To understand the implication of our findings, we draw on the expressive writing literature which suggests that venting out negative emotions, confronting inhibited thoughts are among the reasons why disclosure, beyond the therapist-client setting has therapeutic benefits [4]. This leads us to ponder whether social media platforms like Twitter are being adopted as an expressive writing platform, to vent out and give structure to stigmatized experiences. Our findings support that venting out via self-disclosures of schizophrenia are followed by increased focus on well-being and self-care, and reduced expression of symptoms and stigma perceptions.
“Write something every day”, she said
“even if it’s only a line,
it will protect you”
… how then should it defend us?
unless by strengthening
our fierce and obstinate centers. — Elaine Feinstein
Moving from identified, dyadic, private settings in traditional therapeutic disclosures, we observed individuals adapting Twitter’s public, broadcasting nature to share their experiences to large, invisible audience and derive therapeutic benefits. We conjecture that disclosers thereby draw unique benefits from this invisible audience: it is known that communicating with an invisible audience reduces inhibition leading to more candid disclosures around sensitive topics than is possible in front of an identified audience [5].
Finally, in the traditional expressive writing paradigm, people engage in long-form writing privately, pouring down their thoughts vividly in an elaborate manner. On the other hand, Twitter is spontaneous, real-time, and largely a public place for sharing content in 280 characters or fewer at a time. Appropriating Twitter for expressive writing thus implies that disclosers have identified mechanisms to adapt these in-the-moment writing forms to meaningfully communicate a variety of thoughts and emotions about their experience of a stigmatized illness like schizophrenia.
What next?
Our work reveals the potential of social media platforms as new therapeutic tools supporting broadcasting self-disclosures. How can social media platforms and the increasing number of technology-mediated therapy and counseling applications accommodate sensitive, stigmatized disclosures and support such behaviors in a meaningful manner. In the paper, we suggest two design directions towards building journaling tools for writing therapy and social support recommendations for individuals to gain therapeutic benefits collaboratively and reciprocally. We also discuss how these tools need to balance the disclosers’ therapeutic needs with boundary regulations, attending to their privacy concerns at the forefront.
For more details about our methods and findings please check out our full paper that has been accepted to CSCW 2018. For questions and comments about the work, please drop an email to Sindhu Ernala at sernala3 [at] gatech [dot] edu. Citation:
Ernala, S. K., Rizvi, A. F., Birnbaum, M. L., Kane, J. M., & De Choudhury, M. (2017). Linguistic markers indicating therapeutic outcomes of social media disclosures of schizophrenia. Proc ACM Hum-Comput Interact, 1(1), 43.
References:
- Joinson, A. N., & Paine, C. B. (2012). Self-disclosure, Privacy and the Internet.
- Marwick, A. E., & Boyd, D. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New media & society, 13(1), 114–133.
- Shimkunas, A. M. (1972). Demand for intimate self-disclosure and pathological verbalization in schizophrenia. Journal of Abnormal Psychology, 80(2), 197.
- Pennebaker, J. W., Mehl, M. R., & Niederhoffer, K. G. (2003). Psychological aspects of natural language use: Our words, our selves. Annual review of psychology, 54(1), 547–577.
- Joinson, A. N. (2001). Self‐disclosure in computer‐mediated communication: The role of self‐awareness and visual anonymity. European journal of social psychology, 31(2), 177–192.
