Name
AI vs. Human Analysis in MedEd Qualitative Research: Insights from a
Physician Flourishing Program
Date & Time
Monday, June 8, 2026, 1:49 PM - 2:09 PM
Location Name
Oglethorpe G
Speakers
Authors
Julia Silverman, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell
Jennifer Groh, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell
Benjamin Weisman, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell
Presentation Topic(s)
Other
Description
PURPOSE: This study investigates the use of artificial intelligence (AI)
to expedite and enhance qualitative analysis research in medical education,
specifically through evaluating clinicians' experiences in a flourishing
program aimed at promoting humanism among healthcare professionals.
METHODS: Clinician focus group transcripts from a six-month program on
character, flourishing, and practical wisdom were analyzed by AI using Gemini
1.5 Pro through the Northwell AI Hub with a structured prompt to identify key
themes, excluding program logistics. Independently, two student researchers
manually coded the transcripts, reconciling discrepancies before finalizing
results. Comparisons were drawn between AI and human-coded themes.
RESULTS: AI identified four themes across 65 quotes: Personal Growth &
Self-Reflection (45.4%), Connectedness & Community (15.8%), Application
& Success (13.7%), and Character Strengths (26.4%). Human analysis
identified six themes across 71 quotes: Introspection (32.4%), Growth
(19.7%), Belonging (14%), Leadership (22.5%), Empathy (7%), and Resilience
(4.2%). While AI demonstrated efficiency in identifying more quotes quickly,
human analysis provided nuanced distinctions among themes.
CONCLUSIONS: AI serves as a beneficial supplement for initial theme
identification, offering efficiency in theme identification processes, yet
lacking contextual depth in certain themes like Empathy & Resilience,
which human analysis captured. Future studies should address AI prompt
variations and explore different AI models to refine AI's role in qualitative
medical education research.
to expedite and enhance qualitative analysis research in medical education,
specifically through evaluating clinicians' experiences in a flourishing
program aimed at promoting humanism among healthcare professionals.
METHODS: Clinician focus group transcripts from a six-month program on
character, flourishing, and practical wisdom were analyzed by AI using Gemini
1.5 Pro through the Northwell AI Hub with a structured prompt to identify key
themes, excluding program logistics. Independently, two student researchers
manually coded the transcripts, reconciling discrepancies before finalizing
results. Comparisons were drawn between AI and human-coded themes.
RESULTS: AI identified four themes across 65 quotes: Personal Growth &
Self-Reflection (45.4%), Connectedness & Community (15.8%), Application
& Success (13.7%), and Character Strengths (26.4%). Human analysis
identified six themes across 71 quotes: Introspection (32.4%), Growth
(19.7%), Belonging (14%), Leadership (22.5%), Empathy (7%), and Resilience
(4.2%). While AI demonstrated efficiency in identifying more quotes quickly,
human analysis provided nuanced distinctions among themes.
CONCLUSIONS: AI serves as a beneficial supplement for initial theme
identification, offering efficiency in theme identification processes, yet
lacking contextual depth in certain themes like Empathy & Resilience,
which human analysis captured. Future studies should address AI prompt
variations and explore different AI models to refine AI's role in qualitative
medical education research.
Presentation Tag(s)
Student Presentation, Best Student Oral Nominee