The Impact of Social-Related Factors on the Severity of Prolonged Grief Symptoms: A Machine Learning Approach

Abstract

Background: The aim of this study is to use a machine learning (ML) approach to investigate the relationship between social dimensions of childhood trauma and attachment style with the severity of prolonged grief symptoms. Methods: This descriptive-correlational study was conducted on 516 bereaved individuals who were selected through convenience sampling. Participants completed the Prolonged Grief Scale (PG-R-13), Childhood Trauma (CTQ), and Attachment Styles (RAAS) Questionnaires with informed consent, and instead of using an independent questionnaire to measure social dimensions, items related to these dimensions were extracted from the questionnaires. Machine learning models, including linear regression, random forest, and gradient boosting, were used to analyze the data to predict the severity of prolonged grief symptoms. Results: The results showed that social dimensions related to childhood trauma were the strongest predictors of the severity of prolonged grief symptoms. The random forest model showed the highest predictive power (60.6% accuracy). It was also found that perceived social support was associated with a reduction in the severity of grief symptoms. Conclusions: This study demonstrated the role of social dimensions of psychological factors in prolonged grief symptoms. The use of ML provided a novel approach to uncovering hidden patterns in the grief-related process that could lead to personalized intervention.