Unlocking the power of generative AI in qualitative data analysis: a deep dive into data from a physical activity service evaluation

Yu, Qicheng, Webb, Justin, Hunte, Raymon and Arachchilage, Chamila D. P. (2025) Unlocking the power of generative AI in qualitative data analysis: a deep dive into data from a physical activity service evaluation. In: 13th International Conference on Frontiers of Intelligent Computing: Theory and Applications (FICTA-2025) June 06 - 07, 2025, 6-7 June 2025, London Metropolitan University, London (UK) / Online. (In Press)

Abstract

Qualitative Data Analysis (QDA) is essential in social sciences and human-computer interaction, involving the interpretation of unstructured data like interviews and surveys. Traditional QDA methods are labour-intensive and prone to human bias. Generative AI (GenAI) offers new possibilities for automating QDA through natural language processing. This study, using qualitative focus group data from a physical activity programme evaluation, as a case study, compares GenAI models with human-led analysis. Findings show GenAI significantly improves efficiency, reducing analysis time from nine hours to one hour with human validation. However, GenAI struggles with nuanced contextual understanding and human judgment. The paper concludes that GenAI can complement QDA by streamlining workflows while maintaining quality through human oversight, offering valuable insights into integrating AI in qualitative research.

Documents
10375:52605
[thumbnail of Unlocking the Power of Generative AI in Qualitative Data Analysis.pdf]
Unlocking the Power of Generative AI in Qualitative Data Analysis.pdf - Accepted Version
Restricted to Repository staff only until 30 April 2026.

Download (645kB) | Request a copy
Details
Record
View Item View Item