“In Minutes Instead of Weeks”: Discursive Constructions of Generative AI and Qualitative Data Analysis

Abstract

The use of qualitative data analysis software (QDAS) platforms have always posed a dilemma for researchers, and the integration of generative artificial intelligence (AI) tools are complexifying this relationship even further. The way QDAS companies are positioning this new development will impact how researchers understand what qualitative analysis is and what it could be. Using discourse analysis methods, we explored how ATLAS.ti, NVivo, and MAXQDA websites constructed the relationship between AI-assist and qualitative research methods. We noted four “discursive dilemmas” across the websites: (a) automated insight-generation versus systematic meaning-making; (b) chatting with documents versus analyzing data; (c) high speed versus high engagement; and (d) novelty versus agency. While some level of hyperbolic discourse can be expected from corporations whose goal is to sell products, we argue that the discourses used on these websites may be incompatible with the epistemological foundations of qualitative research.