AI4Qual: A Comprehensive Field Guide to LLM-Supported Qualitative Research (Tutorial)

Abstract

This half-day (≈ 180 minutes), hands-on tutorial translates recent findings on large language model (LLM) support for qualitative research into a concise, end-to-end workflow grounded in our prior studies. Part I (LLM-supported semi-structured interviewing) distills design principles for AI-generated follow-up questions—covering role assignment, engagement patterns, user perceptions, and practical tactics for integrating LLM prompts during interviews while honoring consent and study protocols. Part II (LLM-supported coding and analysis) operationalizes results from prior work on using ChatGPT/LLMs for qualitative coding: moving from open coding to categories/themes, interpreting human–LLM alignment and inter-rater reliability findings at a conceptual level, and building a lightweight RAG-backed evidence path that links codes/themes to supporting excerpts. We conclude with a short discussion on applying these methods across domains and in instructional contexts. Attendees leave with slides and exemplar materials derived from the papers, plus a minimal RAG tookit that they can adapt to their own datasets.

Ähnliche Arbeiten