What Happens At The Nexus of Qualitative Research and Artificial Intelligence

Among the many questions raised by the growing access to artificial intelligence (AI) technology is whether such tools can - or should - be used in research settings. Among qualitative research, there is both enthusiasm and concern about the ways that AI is used in data analysis. Proponents argue for AI in qualitative research in pursuit of more efficient research that uses larger data sets and responds to concerns about the rigor of analysis of the unstructured data in qualitative projects. Critics caution the use of a technology known for bias and hallucinations and that raises major ethical questions about data privacy and the broad social impacts of the AI industry.

"Qualitative" is a broad category; each project will include a unique collection of data analysis strategies (such as inductive or deductive coding, developing code frequencies, or comparing qualitative data across categories) depending on the methodology, data types, and disciplinary expectations. Likewise, AI carries many different meanings across communities, referring to both forms of technology and the cultural experience of AI integration and resistance. The debate at this intersection, therefore, requires detail on the range of technology and analysis strategies for which researchers use it.

In this presentation, I report on a thematic analysis of more than 150 research articles using both qualitative methods and AI technology. Through this analysis, I identify (1) which technologies are used in qualitative research assisted with AI (2) motivations for using AI and (3) data analysis strategies for which AI is used. Understanding the variety of practices that can take place at the nexus of AI and qualitative research will support data professionals who do their own qualitative work or engage with researchers drawing on qualitative methods who are considering AI.