open access
메뉴
ISSN : 1229-0688
In this study, we examined the prediction of nonsuicidal self-injury (NSSI) thoughts and behaviors, along with their frequency and recency in adolescents. Data from 470 students nationwide, initially collected for the development of a NSSI belief scale, were reanalyzed. We employed multivariate machine learning techniques, specifically classification and regression-based decoding. The results indicated that all eight scales significantly predicted NSSI thoughts and behaviors. Furthermore, specific items with the highest explanatory power were identified as significant predictors of NSSI thoughts and behaviors. In particular, the NSSI belief scale and urge to self-injury scale predicted the frequency of NSSI thoughts and behaviors, whereas the urge to self-injury scale and possible future self-injury scale predicted the recency of NSSI thoughts and behaviors. These findings suggest that the scales used in this study hold potential as tools for screening and predicting NSSI in adolescents, which could contribute to the prevention and early intervention strategies.
