What makes text-based online chat work? Communication features of effective chat helpers’ responses to chat users

Background: Text-based online chat (TBC) offers accessible mental health support for youths and young persons who may avoid traditional face-to-face services. While research demonstrates TBC effectiveness, limited studies examine which specific chat helpers’ communication features contribute to positive outcomes. This study identified textual and linguistic characteristics of chat helpers’ responses associated with chat users’ perceptions of service usefulness in an online synchronous text-based chat service in Singapore.Methods:We analyzed 1,062 chat sessions from youths and young persons aged 12-25 years collected between 2016-2020. We examined 26 communication features used by chat helpers including emoticon use, affective language, session duration, response lag time, four types of social support (emotional, informational, esteem, network), pronoun usage patterns, politeness markers, question frequency, linguistic complexity measures, and superlatives. Analyses included univariate logistic regression and Latent Profile Analysis to identify communication styles.Results:Eleven communication features significantly predicted positive outcomes. Emoticon use showed the strongest effect (7.3 percentage point increase), followed by affective language (6.4 points) and session duration (4.1 points). Three social support types were effective: network support (4.0 points), esteem support (3.8 points), and emotional support (2.7 points). Linguistic features including superlatives and words per sentence also predicted success. First-person pronouns used by chat helpers showed negative effects. Three distinct communication profiles emerged, with "Emotionally Expressive" and "Inquiry-Driven" styles achieving 85% success rates compared to 75% for "Typical" style.Conclusion: Effective TBC requires specific text-based skills including emotional expression through emoticons and affective language, and providing social support. Training programs should develop these competencies while allowing for different communication styles such as emotionally expressive or inquiry-driven approaches.