Name
Beyond Content Review: AI-Guided Reflection to Strengthen Metacognition and Self-Regulated Learning in Premedical Students
Date & Time
Monday, June 8, 2026, 1:30 PM - 1:45 PM
Location Name
Oglethorpe G
Speakers
Authors
Antonie Rice, Baylor College of Medicine
Shelley Kumar, Baylor College of Medicine
Presentation Topic(s)
Technology and Innovation
Description
PURPOSEA significant pedagogical gap exists regarding the effectiveness
of tailored curricula that explicitly teach pre-medical and
post-baccalaureate students metacognition and self-regulated learning (SRL).
This gap is compounded by limited exploration of how Artificial Intelligence
(AI) can function as a scalable, confidential reflection coach, a key
component of professional identity formation. We evaluated the impact of a
novel “Success Strategies” course, incorporating an AI-enabled reflection tool,
on students’ SRL skills and confidence.
METHODSWe conducted a mixed-methods convergent study with 24
post-baccalaureate students enrolled in an eight-week course using Team-Based
Learning (TBL) and weekly AI-guided reflection (custom GPT trained on Gibbs’
reflective cycle). Quantitative data included pre-post assessments with
validated instruments (General Perceived Self-Efficacy Scale [GPSE],
Reflection in Learning Scale [RILS], Diversity of Strategies for Motivation
in Learning [DSML]) and the System Usability Scale (SUS). Qualitative data
comprised thematic analysis of open-ended reflections exploring changes in
learning strategies and experiences using AI.
RESULTSQuantitatively, reflection skills improved (RILS: pre M=62.6 vs post
M=71.1, p=0.007), as did self-efficacy (GPSE: pre M=46.5 vs post M=57.7,
p<0.001). DSML domains showed gains in Course Utility (p=0.03) and Test
Anxiety management (p=0.002), while Self-Regulation did not change
significantly (p=0.09). SUS ratings were high (median 4–5/5), indicating
strong perceived usability of the AI coach. Qualitative analysis generated
three themes: metacognitive transformation, locus of control shift, and AI
limitations.
CONCLUSIONSTBL integrated with structured reflective practice and an
AI-enabled coach fostered professional identity development through increased
self-efficacy, reflection, and process-oriented thinking. However, unchanged
Self-Regulation scores, alongside reports of generic, context-blind AI
feedback, highlight an implementation gap: AI that lacks contextual
specificity may limit sustained behavioral change despite high satisfaction.
of tailored curricula that explicitly teach pre-medical and
post-baccalaureate students metacognition and self-regulated learning (SRL).
This gap is compounded by limited exploration of how Artificial Intelligence
(AI) can function as a scalable, confidential reflection coach, a key
component of professional identity formation. We evaluated the impact of a
novel “Success Strategies” course, incorporating an AI-enabled reflection tool,
on students’ SRL skills and confidence.
METHODSWe conducted a mixed-methods convergent study with 24
post-baccalaureate students enrolled in an eight-week course using Team-Based
Learning (TBL) and weekly AI-guided reflection (custom GPT trained on Gibbs’
reflective cycle). Quantitative data included pre-post assessments with
validated instruments (General Perceived Self-Efficacy Scale [GPSE],
Reflection in Learning Scale [RILS], Diversity of Strategies for Motivation
in Learning [DSML]) and the System Usability Scale (SUS). Qualitative data
comprised thematic analysis of open-ended reflections exploring changes in
learning strategies and experiences using AI.
RESULTSQuantitatively, reflection skills improved (RILS: pre M=62.6 vs post
M=71.1, p=0.007), as did self-efficacy (GPSE: pre M=46.5 vs post M=57.7,
p<0.001). DSML domains showed gains in Course Utility (p=0.03) and Test
Anxiety management (p=0.002), while Self-Regulation did not change
significantly (p=0.09). SUS ratings were high (median 4–5/5), indicating
strong perceived usability of the AI coach. Qualitative analysis generated
three themes: metacognitive transformation, locus of control shift, and AI
limitations.
CONCLUSIONSTBL integrated with structured reflective practice and an
AI-enabled coach fostered professional identity development through increased
self-efficacy, reflection, and process-oriented thinking. However, unchanged
Self-Regulation scores, alongside reports of generic, context-blind AI
feedback, highlight an implementation gap: AI that lacks contextual
specificity may limit sustained behavioral change despite high satisfaction.
Presentation Tag(s)
Best Faculty Oral Nominee