Stress, a common phenomenon among students, especially college students, is worsened by part-time working and studying. This paper discusses the correlation between age, working hours, sleep, and Patient Health Questionnaire (PHQ-9) and Perceived Stress Scale (PSS) scores as assessed quantitatively. Among the participants, 280 completed an online survey, the results were statistically processed using Ordinary Least Squares (OLS) regression and machine learning algorithms (Random Forest and Gradient Boosting). Among all the variables, anxiety came out as the most significant predictor of stress (r = 0.5344, p ≤ 0.05), besides work hours and sleep. Specifically, the Random Forest (RF) model yielded the lowest Mean Squared Error (MSE) of all the models at 13.4562. Some recommendations include counselling, flexible working, and teaching about the need for rest to lessen tension in their lives.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Artificial Intelligence and Predictive Analytics for Building Multi-dimensional Psychological Evaluation Models to Address Employment Stressors in College Students

  • Chengtao Qin

摘要

Stress, a common phenomenon among students, especially college students, is worsened by part-time working and studying. This paper discusses the correlation between age, working hours, sleep, and Patient Health Questionnaire (PHQ-9) and Perceived Stress Scale (PSS) scores as assessed quantitatively. Among the participants, 280 completed an online survey, the results were statistically processed using Ordinary Least Squares (OLS) regression and machine learning algorithms (Random Forest and Gradient Boosting). Among all the variables, anxiety came out as the most significant predictor of stress (r = 0.5344, p ≤ 0.05), besides work hours and sleep. Specifically, the Random Forest (RF) model yielded the lowest Mean Squared Error (MSE) of all the models at 13.4562. Some recommendations include counselling, flexible working, and teaching about the need for rest to lessen tension in their lives.