Enhancing reflective practice with ChatGPT: A new approach to assignment design

<b>What was the educational challenge?</b> Reflective practice (RP) fosters professional growth but is often hindered by unclear purpose and minimal guidance. Designing comprehensive RP assignments is challenging because it takes time and not all faculty possess the skills to generate effective assignments. This innovation addressed two challenges: (1) creating clear, scaffolded reflective assignments using the transparent assessment framework (TAF) and (2) reducing faculty workload in assignment design. <b>What was the solution?</b> ChatGPT 4o was used to design RP assignments in health professions education (HPE). The TAF guided the design of the assessment. <b>How was the solution implemented?</b> A four-step process was followed to facilitate the design of the assignment. ChatGPT 4o was prompted to design the assignment and refined for the course. <b>What lessons were learned that are relevant to a wider global audience?</b> A pilot study in three graduate-level HPE courses showed ChatGPT-assisted assignments improved clarity, structure, and student performance while decreasing faculty preparation time. <b>What are the next steps?</b> We plan to expand this research to obtain student feedback on the effectiveness of the redesigned assignment and to explore the effectiveness of AI-generated reflective assignments with medical students who may require more structured guidance and contextualized prompts.