Artificial intelligence-based positive youth development intervention protocol in school settings

Positive Youth Development (PYD) conceptualizes youth as agentic, developmentally flexible individuals whose functioning reflects dynamic person-context transactions. Rapid developments in AI now offer new opportunities to support counselors and extend intervention-based programs. The purpose of this paper is to outline an eight-week AI-supported PYD intervention for high school students and frame a feasibility-oriented mixed-method outcome research design. The model integrates daily AI micro-nudges, short voluntary micro-coaching dialogues, and one weekly 10-15-min counselor mini-session. Trait-level outcomes (self-regulation and school belonging) are assessed pre-post; state-level micro-regulation is assessed daily via Ecological Momentary Assessment (EMA). Quantitative analysis uses Lakens's generalized within-subject effect-size formulation and baseline trend-corrected Tau-U; qualitative analysis uses deductive thematic analysis. The design is intended to demonstrate whether mechanism-consistent developmental micro-arrangements occur in naturalistic conditions rather than to test causal efficacy. This paper provides a research-ready protocol for future randomized controlled trials in AI-supported school-based intervention program.