Profiling the Digital Mosaic of Uncensored Suicidal Thought and Behavior: A Theory-Driven Network Analysis of Online Written Expression

Abstract

The highly stigmatizing and taboo nature of suicidal thought and behavior (STB) makes it especially prone to censorship across modern communication platforms, ultimately impacting breadth of phenomenological detection. Applying ideation-to-action models of suicide to uniquely naturalistic data sources may serve to deepen understanding of digital STB expression. STB written communication across N=839 posts on the pro-choice suicide forum, Sanctioned Suicide, was qualitatively coded through development of a 20-item inventory based on the Integrated Motivational-Volitional (IMV) model. Reliable items were modeled using network analysis techniques to highlight patterns in post content. Coping strain-related stressors and feelings of defeat/entrapment most frequently characterized posts. Recapitulating the IMV framework, network modeling implicated defeat/entrapment as the most central node among items. Further analysis of the estimated network structure highlighted three primary themes of STB expression: contexts of powerlessness, trauma-mediated alienation, and crowdsourced preparation. This research offered an analytical foray into the expression of unmitigated STB. Findings aligned with documented associations in the literature, establishing a baseline summary of STB content within a marginalized, at-risk community. Future work should focus on modeling detailed semantic and social engagement data within this extraordinary microcosm to further explore, describe, and develop digital profiles of STB risk.

The Hidden Depths of Suicidal Discourse: Network Analysis and Natural Language Processing Unmask Uncensored Expression

“…Most concerningly, upon diving into the specific tokens of topics (Figure 3), a prominent representation of methods-related content and sharing activity was uncovered. This finding has also been observed in a previous analysis of the forum (Lekkas et al, 2023). The presence of a close token association between "information" and "thread" is a consequence of the ubiquity of suicidal preparation discourse.…”

“…First, several tokens belonging to seemingly innocuous topics such as "drink"/"drinking", "taste", "bag", "exit", "dose", and "help" were among the most central in the network. This underlines the prominence of a major content theme across Sanctioned Suicide, namely the appeal for, and sharing of, information related to suicide-related methods (Lekkas et al, 2023). This is secondarily supported by the close association of "information" and "thread" in the network.…”

Breaking the Silence: Leveraging Social Interaction Data to Identify High-Risk Suicide Users Online Using Network Analysis and Machine Learning

“…Upholding a pro-choice, censorship-free philosophy, Sanctioned Suicide has deservedly come under fire due to public outcry regarding suicides that were believed to have been facilitated through the content hosted within its virtual walls (Barbaro, 2021). The forum's content covers a variety of topics and themes, with the most concerning in regards to the solicitation and sharing of suicide methods-related advice (Lekkas et al, 2023). Although this obviates any potentially positive impact this community has on its highly vulnerable members, it is still important to note that individuals come to Sanctioned Suicide for different reasons.…”