Researchers have developed a novel multi-phase pipeline for qualitative analysis that integrates Large Language Models (LLMs) with human expertise. This method was used to create a hierarchical codebook from a large dataset of K-12 educator interactions with an AI platform. The process involved LLMs generating initial labels and annotations, while human researchers maintained conceptual control, refining definitions and interpretive frameworks. The resulting codebook, comprising 72 items across 19 categories and six domains, was validated through systematic human coding, achieving reliability with multi-label annotation measures. AI
IMPACT This methodology offers a scalable approach for qualitative researchers to analyze large interaction corpora, potentially accelerating insights into AI use in educational settings.
RANK_REASON The cluster contains an academic paper detailing a new methodology for qualitative research using LLMs. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
