A new preprint on arXiv examines whether large language model chatbots can take on the role of interviewers in qualitative research. The authors position AI-led interviews as a way to collect rich, open-ended data at a scale that traditional human interviews cannot easily reach.
The study is designed as a direct comparison between AI-led and human-led chat interviews. By pitting the two formats against each other, the researchers aim to assess whether chatbots can produce comparable interview experiences and data quality.
Because the abstract does not include results, the findings are not yet available from this source. The paper's significance lies in its attempt to test a practical use of LLMs for social science research, potentially offering a middle ground between structured surveys and in-depth interviews.