Approximately 50,000 autistic young adults exit high school and become eligible for the workforce each year. Federal legislation mandates pre-employment transition services (Pre-ETS) for these young adults while still connected to school. However, Pre-ETS employment outcomes continue to be low (~20% within 2 years after high school) with brief job tenure (~2-6 months) among autistic young adults. Most interventions focus on getting jobs, with limited research on interventions targeting mechanisms of job retention (e.g., workplace conversations, regulating emotions and social anxiety). This study evaluated the feasibility and initial effectiveness of adding WorkChat (community co-designed, web-based workplace conversation simulator) to Pre-ETS. We hypothesized that providers could feasibly deliver WorkChat; and Pre-ETS+WorkChat participants would adhere to the delivery, while improving their social cognition, workplace conversational skills, social anxiety, and job tenure compared to Pre-ETS-as-usual.
Methods
For this intent-to-treat, RCT, we recruited autistic young adults (ages 17–26) from five Pre-ETS programs serving rural, suburban, and urban communities. Participants were randomized 1:1 to Pre-ETS+WorkChat (n=19) versus Pre-ETS-as-usual (n=22). Transition specialists implemented WorkChat using a manualized, 15-session curriculum (core skills videos; simulated conversations with customers, coworkers, supervisors; and simulated “workdays” with automated feedback). Outcomes were assessed at pre- and post-test; including social cognition (social perception, theory of mind, emotion management), workplace conversational ability, and social anxiety. Job tenure outcomes (days employed, hours worked, and wages earned in longest-tenure job) were collected at 3-, 6-, and 9-month follow-up. Employment outcomes had non-normal distributions and were analyzed using Poisson loglinear regression (incidence rate ratios; adjusted for baseline differences in recent work experience). Mechanistic outcomes were analyzed using mixed-effects linear regression.
Results
Baseline characteristics were comparable except fewer WorkChat participants had recent work experience (37% vs 73%). Feasibility was supported: 100% completed core skill videos and 73.7% completed the scaffolded curriculum of practice conversations and virtual workday sessions despite pandemic-related staffing constraints. Compared with Pre-ETS as usual, WorkChat participants demonstrated greater pre–post improvements in emotion management, workplace conversational skills (role-play performance [scored by masked raters] with customer, coworker, supervisor), and reduced social anxiety (all p<.05; except social perception and theory-of-mind; p>.05). WorkChat was associated with greater job sustainability in the longest-tenured job: days employed (IRR=1.32, 95%CI 1.21–1.44), hours worked (IRR=1.57, 95%CI 1.51–1.63), and wages earned (IRR=1.96, 95%CI 1.94–1.98), adjusting for prior work history (all p<.001).
Conclusions and Implications
WorkChat was feasibly delivered with promising initial effectiveness on employment-relevant social and emotional mechanisms, and job tenure indicators. Findings suggest that co-designed, technology-based simulation can extend evidence-informed practice in community services and support more equitable access to sustainable employment for autistic young adults, a key social determinant of health and economic security. Though we advice caution in not over-interpreting results from a small sample with the majority being White and male. For social work practice, integrating structured conversational and emotion management training into services may strengthen early job experiences. Future research should conduct a fully powered trial to confirm effectiveness, validate mechanisms, assess “desirable employment” (fit with skills), and evaluate implementation and equity across historically underrepresented groups.
![[ Visit Client Website ]](images/banner.gif)