Adaptive self-reflection as a social media self-effect: Insights from computational text analyses of self-disclosures of unreported sexual victimization in a hashtag campaign

Abstract

Hashtag campaigns calling out sexual violence and rape myths offer a unique context for disclosing sexual victimization on social media. This study investigates the applicability of adaptive self-reflection as a potential self-effect from such public disclosures of unreported sexual victimization experiences by analyzing 92,583 tweets that invoked #WhyIDidntReport. A supervised machine learning classifier determined that 61.8% of the tweets were self-disclosures of sexual victimization. Linguistic Inquiry and Word Count (LIWC) analysis showed statistically significant differences in four psycholinguistic dimensions (greater use of past focus, cognitive processes, insight, and causation words) connected with reflective processing in tweets with self-disclosed sexual victimization compared to those without. Additionally, topic modeling and thematic analysis identified nine salient topics within the self-disclosing tweets, comprising three self-distanced representations (i.e., relatively abstract and insightful construals) of the unwanted experiences: (a) acknowledging one’s previously unacknowledged victimization, (b) reaffirming one’s rationale for not reporting, and (c) decrying invalidating response to one’s disclosure. Moving beyond reception effects and social support in extant research about social media as a coping tool, this study provides new empirical insights into the potential of social media to promote expressive meaning-making of upsetting and traumatic experiences in ways that support recovery and resilience.