Abstract
<p>Background: Text-based chat (TBC) services are among the most accessible mental healthsupport for youth, yet more is known about whether they help than about how they help, thatis, what workers do in sessions. Without an account of these worker skills, TBC services cannotbe reliably trained for or evaluated. This study described the techniques chat workers use,drawing on the Behavior Change Technique Taxonomy (BCTTv1), and linked them to user-reported outcomes.Methods: We analyzed 1,439 real chat sessions (44,916 worker turns) from 73 workers atQuickChat, an anonymous, free, synchronous TBC service for youths aged 12 to 25 in Singapore.Each worker turn was coded for the presence of 32 behavior change techniques (BCTs). Weexamined how often, densely, and broadly techniques were used; how far use varied acrossworkers; and associations with three post-session outcomes: understanding, coping, andperspective.Results: All 32 techniques appeared, but practice was concentrated in a few: social support(unspecified and emotional), problem solving, re-attribution, and framing/reframing accountedfor over half of all instances. Use varied mainly by session, not by worker, indicating skillsshared across the workforce, not individual styles. Three were positively associated with alloutcomes, namely social support (unspecified and emotional) and information about social andenvironmental consequences, though effects were small.Conclusions: Chat workers deliver change through a small set of core skills, centered on socialsupport and shared across the workforce. These findings offer a first step toward a programtheory of TBC to guide training and evaluation.</p>