Cox-Stratified Model in Relationship Analysis between Employee Mental Health and Resignation Decision
DOI:
https://doi.org/10.24036/ujsds/vol3-iss2/350Keywords:
Cox Stratified, Employee Turnover, Mental Health, Self-Control, Survival AnalysisAbstract
This study examines the relationship between employee mental health and turnover decisions using the Cox Stratified model. Utilizing secondary worker turnover data from Kaggle, the research investigates the impact of anxiety and self-control on job tenure. Results indicate that the Cox Proportional Hazard model significantly explains this relationship, with self-control emerging as a key factor negatively associated with turnover risk. Stratification of profession variables, which did not meet the proportional hazard assumption, revealed variations in survival rates across different professions. Professions requiring strong self-control, such as HR and sales, exhibited higher survival probabilities, whereas high-pressure professions like consulting andshowed lower survival rates. A reduced model confirmed the importance of self-control in employee retention. The findings suggest that interventions aimed at enhancing self-control could serve as an effective strategy for mitigating turnover, especially in high-stress occupations. Elevated job pressure can negatively impact employee mental well-being, potentially disrupting self-control and increasing anxiety levels. Future research could incorporate additional influential factors, such as job satisfaction, work environment, and social support, to further develop this research. Furthermore, the implementation of real-time data collection could enable continuous monitoring of mental conditions, behaviors, and relevant factors such as self-control and anxiety, providing dynamic insights over short time intervals.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Sari Agustin, Tessy Octavia Mukhti, Suci Rahmadani, Afifah Nabilah, Wafiq Alya Aufa

This work is licensed under a Creative Commons Attribution 4.0 International License.