大学生抑郁影响因素的多元线性回归和路径分析

Multiple linear regression and path analysis on influencing factors of depression of college students

目的以多元线性回归和路径分析方法深入探讨大学生抑郁影响因素及作用机制。方法利用多级整群随机抽样方法,对1 490名大学生进行调查和量表评测,采用多元线性回归和路径分析进行生活事件、应对方式和社会支持等因素与抑郁间关系及作用模式的分析。结果9.2%的大学生有自杀意念,55.1%的大学生存在抑郁情绪。标准化回归系数显示积极应对、主观支持、受惩罚和客观支持为抑郁的保护因素;学习压力、其他事件、消极应对和健康适应问题为抑郁的危险因素。路径分析发现学习压力可直接导致抑郁,路径系数为0.195,人际关系仅通过间接途径作用于抑郁,路径系数为-0.005,积极应对和社会支持对抑郁有直接的缓解作用,路径系数分别为-0.244和-0.159。结论抑郁的发生与生物、心理和社会等多因素有关,多元线性回归分析和路径分析在探讨抑郁的发生机制过程中各有侧重,相互补充。

Objective To explore the relationship and mechanism of depression of college students with life events, coping style and social support by multiple linear regression and path analysis. Methods By multistage sampling, 1 490 college students were recruited. BDI, ASLEC, SCSQ and SSRS were used as the survey instrument. Multiple linear regression and path analysis were used to probe the main factors of depression and its role model. Results 9.2% students had suicidal ideation, 55.1% students had depression problems. Standardized regression coefficients showed that active coping, subjective support, punishment factor and objective support were protective factors for the occurrence of depression; study pressure, other events, negative coping and health adaptation problems were risk factors. Path analysis showed that study pressure could induce depression through a direct way whose path coefficients was 0. 195. Interpersonal relation had no direct relation to depression; it could only through indirect channels （mediated by social support, coping style） influence the occurrence of depression. Active coping style and social support could also inhibit the occurrence of depression independently, with the path coefficients being 0. 244 and - 0. 159, respectively. Conclusions The occurrence of depression is a combined effect of biological, psychological and social factors. In exploring the mechanism of depression, both multiple regression analysis and path analysis have their own focus; the results can be mutually complementary to each other.