Sheet 11 of 17 · Synthesis
Student success
Adoption rates describe what students do with AI. This sheet asks how AI affects their development, stage by stage, and shows where the evidence runs out.
Fig. 11.1: The student success continuum
Six stages from arrival to life after college. For each: what the evidence in this dashboard says, how strong it is, and what is missing.
The student success continuum
- Preparation. What knowledge, habits and expectations do students bring from high school? 84% of US high-school students used GenAI for schoolwork by May 2025; only 30% of teens had discussed safe AI use with a teacher. Evidence: Moderate. Gap: Use is measured; competency at entry is not.
- Transition. How do students adapt to college expectations, integrity rules and disciplinary uses of AI? 23% of Dartmouth’s incoming class came from schools that banned AI in every class. Two-year students were twice as likely as four-year students to be unclear on AI rules (20% vs 10%). Evidence: Limited. Gap: No study follows students across the transition.
- Engagement. Does AI change participation, help-seeking, faculty contact and belonging? About 22% of teens’ first moves when stuck on schoolwork are AI-mediated. 39% of current Dartmouth students feel pressure to use AI because peers do. Evidence: Limited. Gap: Whether AI extends human support or replaces it.
- Learning. Does AI build knowledge, critical thinking and independent competence? A scaffolded AI tutor more than doubled learning gains; unrestricted chatbot access lowered unaided exam scores by 17%. Evidence: Stronger, but narrow. Gap: Short, single-course studies; little on everyday use.
- Persistence. Do AI access, support and confidence relate to staying enrolled? No study in this evidence base links AI use to persistence or retention. Evidence: None found. Gap: The largest gap on the continuum.
- Completion and flourishing. What happens to competence, agency, well-being and readiness for life after college? 58% of US seniors say they need stronger AI skills for work; 28% say their program meaningfully integrated AI. Evidence: Limited. Gap: Self-reported readiness; no tracking into employment.