Leveraging Interview-Informed LLMs to Model Survey Responses: Comparative Insights from AI-Generated and Human Data
This work addresses the problem of integrating qualitative and quantitative methodologies for social science researchers, though it is incremental in refining existing LLM applications.
The study tackled the challenge of aligning qualitative and quantitative data in mixed methods research by using interview-informed large language models (LLMs) to predict human survey responses, finding that LLMs captured overall patterns but with lower variability than humans, and that interview data improved diversity for some models.
Mixed methods research integrates quantitative and qualitative data but faces challenges in aligning their distinct structures, particularly in examining measurement characteristics and individual response patterns. Advances in large language models (LLMs) offer promising solutions by generating synthetic survey responses informed by qualitative data. This study investigates whether LLMs, guided by personal interviews, can reliably predict human survey responses, using the Behavioral Regulations in Exercise Questionnaire (BREQ) and interviews from after-school program staff as a case study. Results indicate that LLMs capture overall response patterns but exhibit lower variability than humans. Incorporating interview data improves response diversity for some models (e.g., Claude, GPT), while well-crafted prompts and low-temperature settings enhance alignment between LLM and human responses. Demographic information had less impact than interview content on alignment accuracy. These findings underscore the potential of interview-informed LLMs to bridge qualitative and quantitative methodologies while revealing limitations in response variability, emotional interpretation, and psychometric fidelity. Future research should refine prompt design, explore bias mitigation, and optimize model settings to enhance the validity of LLM-generated survey data in social science research.