Generative Artificial Intelligence for Enhancing Qualitative Research (GENQ)

• In Event: AERA Roundtable Session Thursday 4:15 pm - Gold 4
  In Roundtable Session: Method in Transition: Ethics, Temporality, and AI in Evolving Qualitative Research Landscapes

Abstract

We explore how Claude, a generative AI (GenAI) tool, can be effectively and efficiently leveraged in qualitative research. Using a three-arm experimental design, we study the optimal ways human-AI collaboration can support thematic analysis of a complex longitudinal dataset with 120 interviews on structural racism in sepsis care across three U.S. health systems. Thematic analysis produced by each arm is assessed by an interdisciplinary review panel. We evaluate the trade-offs between investment of time and cost, and analytic generativity. Preliminary findings indicate that Claude considerably improves efficiency while identifying cross-cutting patterns across complex qualitative data. This study provides methodological guidance regarding how to best integrate GenAI into qualitative research workflows without compromising interpretive depth, methodological rigor and contextual sensitivity.

Authors

• Murat Turk, Academy of Research, Design, and Training (ARD@T)

• Sinem Toraman Turk, Yale University

• Erika L. Linnander, Yale University

• Leslie Curry, Yale University