Text Analysis LLM_ Return to work Interviews

This repository provides full methodological transparency materials for the LLM-assisted qualitative component of the study: Procedural Justice in AI-Mediated Return-to-Work Decisions: Privacy, Human Oversight, and Trust among Injured Workers. The package documents how a Large Language Model (LLM) was used as a structured analytic support tool within a human-led, theory-driven qualitative coding process. It includes coding scripts, prompts, validation procedures, example datasets, overlap metrics, and reproducibility materials. The LLM was used exclusively to assist structured extraction under predefined theoretical constructs. It did not generate new theory, replace human coding, or override human interpretive judgment. Purpose of the Repository The repository was created to ensure: Methodological transparency Replicability Parameter disclosure Prompt disclosure Validation transparency Stability testing documentation Compliance with emerging best practices in AI-assisted qualitative research It directly addresses editorial recommendations to open-source AI coding scripts and provide a clear step-by-step protocol.