Role Play: Conversational Roles as a Framework for Reflexive Practice in AI-Assisted Qualitative Research
Article
Thominet, L, Amorim, J, Acosta, K et al. (2024). Role Play: Conversational Roles as a Framework for Reflexive Practice in AI-Assisted Qualitative Research
. 10.1177/00472816241260044
Thominet, L, Amorim, J, Acosta, K et al. (2024). Role Play: Conversational Roles as a Framework for Reflexive Practice in AI-Assisted Qualitative Research
. 10.1177/00472816241260044
Previous literature has shown that generative artificial intelligence (GAI) software, including large language model (LLM) chatbots, might contribute to qualitative research studies. However, there is still a need to examine the relationships between researchers, GAI technologies, data, and findings. To address this need, our team conducted a thematic analysis of our reflexive journals from an LLM chatbot-assisted research project. We identified four roles that researchers adopted: managers closely monitored the LLM's work, teachers instructed the LLM on theories and methods, colleagues openly discussed the data with the LLM, and advocates worked with the LLM to improve user experiences. Planning for and playing with multiple roles also helped to enrich the research process. This study underscores the potential for using conversational roles as a framework to support reflexivity when working with GAI technologies on qualitative research.