(PS8-90) Text-based teen-to-teen helpline conversations impacts user expressed affectivity and suicidality

Background: Suicide continues to be a leading cause of death for youth and young adults. Teen-to-teen mental health helplines provide an accessible, low-cost intervention to address this public health priority. The present study examines text-based conversations between adolescents on helpline platforms to understand how language shapes affect and suicidality in teens. 

Methods: We examined 9,836 unique text-based conversations between January, 2022 and December, 2024 at a digital, community-based teen-to-teen mental health helpline. Conversational markers of affectivity and suicidality were analyzed using Linguistic Inquiry and Word Count (LIWC; Pennebaker et al., 2022) and natural language processing (NLP; Hirschberg & Manning, 2015) approaches. Full sentiment analysis using NLP approaches will be included in the presentation. 

Results: Our preliminary results indicated that many teen-to-teen helpline conversations occurred not in a single session, but across several days (on average, 3 days and 7.35 hours). On average, conversations consisted of about 16 posts (M = 16.28, SD = 31.07) with agents (M=9, SD=16.11) posting more than users (M = 7, SD = 15.51). Initial analysis of helpline conversations demonstrated depression (23%) and suicide and self-harm (15%) were topics frequently discussed by users when accessing this service, thus, highlighting the importance of this research in promoting adolescent mental health. 

Our initial analysis using LIWC examined the use of I-words, positive and negative tone, and authenticity throughout text-based conversations between teens. We found preliminary patterns whereby the use of I-statements increased and positive tone increased over time. On the other hand, negative tone decreased. Mixed evidence was found for authenticity (i.e., speaking in an honest way and without filtering words) requiring further analysis to understand granular effects. Analyses of full sample will examine these speech patterns across two years of teen-to-teen conversational data on this helpline platform. 

Conclusion: Given the underutilization of mental health services in adolescents and the rapid increase in digital interventions, understanding how teens engage with and experience text-based helplines are critical. The current study aims to characterize conversation patterns in relation to suicide risk using a novel approach of studying conversation in a text-based teen-to-teen helpline. Our results will explore how adolescents use peer-to-peer conversations to express emotions and reduce suicidality and aim to inform areas of strength and quality targets within this understudied service sector.