A Text Mining-Based Analysis of Research Trends on Self-Efficacy in the Field of Kinesiology

The study aimed to analyze research trends related to self-efficacy in the field of kinesiology using text mining and Latent Dirichlet Allocation (LDA) topic modeling techniques. For this purpose, 500 abstracts of academic papers on self-efficacy published between 1993 and 2024 were collected through the RISS academic search portal. Subsequently, a series of analyses were conducted, including word frequency analysis, TF-IDF analysis, topic number estimation based on Coherence Score, LDA topic modeling, topic similarity analysis, and distribution analysis by periods. The results indicated that self-efficacy research in kinesiology has primarily focused on interventions involving physical activity and the interaction of psychological factors, with a recent expansion toward topics incorporating emotional and cognitive components. Specifically, keywords such as “physical activity,” “participation,” “self,” “psychology,” and “engagement” frequently appeared and the overlap among topics implies that the concept of self-efficacy is structured in a complex and integrated manner. Furthermore, the keywords analysis by periods exhibited that while early studies primarily emphasized physical activity, recent research has increasingly addressed topics related to mental health. This study provides a meaningful foundation for future research by empirically presenting the structural flow and thematic evolution of self-efficacy research in kinesiology through a quantitative, text-driven analytical approach.