Changes in Students’ Writing Self-Efficacy and AI Literacy Following Structured, Stage-Specific Guidance for AI-Supported Writing
The adoption of generative artificial intelligence (GenAI) raises questions about fostering AI literacy while preserving independent writing in criminal justice education. This exploratory embedded mixed-methods evaluation examined a structured, stage-specific AI-supported writing protocol in a Spring 2026 undergraduate course. Quantitative data included 30 pretest and 25 posttest surveys, with 21 matched responses for most paired analyses. Qualitative data comprised 25 student reflections. Unadjusted paired tests indicated significant gains in self-reported overall AI literacy, technical proficiency, and ethical competence, but none of 11 comparisons survived Bonferroni correction. Writing self-efficacy showed no significant change. Student reflections highlighted editing support, scaffolding, content clarification, verification concerns, selective use of AI assistance, nonuse, and reported ethical awareness. Findings offer preliminary, noncausal insights into the feasibility and perceived value of integrating critical, ethical GenAI guidance into disciplinary writing instruction.
Authors
- Sungil Han
- Jaeyong Choi
Institutions
- University of North Carolina at Charlotte (US)
- West Chester University (US)
Publication Details
- Journal
- Journal of Criminal Justice Education
- Published
- 2026-11-21
- DOI
- https://doi.org/10.1080/10511253.2026.2765773
- Primary Topic
- Artificial Intelligence in Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00