Characterizing Audience Engagement and Assessing Its Impact on Social Media Disclosures of Mental Illnesses
Keywords:self-disclosure, twitter, social media, mental health, schizophrenia, audience, time series forecasting, stigma
Self-disclosures of mental illnesses have been identified to yield coping and therapeutic benefits. An important construct in the self-disclosure process is the audience with whom the individual interacts and shares their experiences. Mental illness self-disclosures are increasingly happening online. However, unlike online support communities where the audience comprises sympathetic peers with similar experiences, what the discloser gains from an ‘invisible’ audience on a general purpose, public social media platform is less understood. Focusing on a highly stigmatized mental illness, schizophrenia, this paper provides the first investigation characterizing the audience of disclosures of this condition on Twitter and how the audience’s engagement impacts future disclosures. Our results are based on a rich year-long temporal analysis of the data of nearly 400 disclosers and their nearly 400 thousand audiences. First, characterizing and modeling the audience engagement temporally, we find evidence of reciprocity in the disclosure process between the discloser and their audience. Then, situating our work in the Social Penetration Theory and operationalizing the disclosure process via a measure of intimacy, an auto-regressive time series model indicates that the patterns of audience engagement and content can forecast changes in the intimacy of disclosures. We discuss the implications for building socially engaging, supportive online spaces for stigmatized mental illness disclosures.