Validating Qualitative Research with Multi-LLM Thematic Analysis and Dual Reliability Metrics

This research explores a novel approach to validating qualitative research by leveraging multiple LLMs for thematic analysis. The combination of Cohen's Kappa and semantic similarity offers a potentially robust method for assessing the reliability of LLM-generated insights.

• •Applies multiple LLMs for thematic analysis, potentially improving robustness and generalizability.
• •Employs dual reliability metrics (Cohen's Kappa and semantic similarity) for enhanced validation.
• •Focuses on strengthening the methodological rigor of qualitative research using AI.