大语言模型在质性研究中的应用探索

Large Language Models (LLMs) demonstrate significant potential in qualitative research, particularly in the critical processes of data coding and interpretive construction. In data coding, LLMs excel at preserving contextual information and conceptual relationships while focusing on events or themes, balancing both holistic and analytical needs of event structure analysis. During interpretive construction, LLMs can assist researchers in comprehensive review, exploration of underlying motivations, integration of socio-cultural contexts, and insightful interpretation and construction. This article presents an application framework and operational methods for LLMs in qualitative research through examples, providing researchers with a quick-start guide. The article also proposes future research directions, including expanding LLM applications across different research paradigms and examining the applicability and effectiveness of LLMs in qualitative research. Overall, while LLMs show notable advantages in rapidly integrating large amounts of information and offering multi-perspective analyses, researchers maintain a crucial role in the qualitative research process.