Generative AI in Qualitative Research and Related Transparency Problems: A Novel Heuristic for Disclosing Uses of AI

Abstract

Generative Artificial Intelligence (AI) tools, particularly large language models (LLMs), are rapidly transforming qualitative data analysis by offering unprecedented speed and scale. However, this integration introduces significant challenges to methodological transparency due to the algorithmic opacity of these “black-box” systems and their hidden decision points. Traditional qualitative reporting guidelines predate generative AI and lack specific guidance for disclosing AI usage. This paper addresses this critical gap by introducing a novel heuristic framework that poses 20 questions across five key themes: The Research Team, Participant Interaction, Study Design, Data Practices, and Data Analysis, providing disclosure actions for each. This framework blends and augments existing frameworks, aligning with principles to guide researchers in meticulously documenting their AI-mediated analytic choices. Methodologically, the framework advances the field by requiring explicit reporting on: the roles of AI tools and the AI literacy of the human research team; AI’s involvement in participant communication, informed consent processes, and safeguards for sensitive demographic data; the alignment of AI tools with theoretical frameworks, their influence on sampling strategies, and discussions during IRB review; data storage, AI’s role in data creation (e.g., synthetic data, transcription, interview protocols), and its assistance in determining data saturation and participant checking; and the precise contributions of AI to coding and thematic categorization, alongside detailed documentation of iterative human-AI interactions and prompts used. By fostering rigorous audit trails and comprehensive documentation, the framework aims to maximize transparency, ensure methodological rigor, and uphold ethical standards in AI-augmented qualitative inquiry, thereby enhancing the credibility of findings.