《职业倦怠通用量表》的多人群信度、效度分析

王瑾; 刘晓曼; 孙彦彦; 楼建林; 袁方; 孙道远; 凌瑞杰; 余丹; 李霜

doi:10.11836/JEOM22365

摘要:

背景《职业倦怠量表–通用版》作为常用职业倦怠评估工具之一，缺乏在我国不同人群中使用的可靠性与有效性验证。

目的通过对《职业倦怠量表–通用版》进行汉化、预试后形成《职业倦怠通用量表》，分析该量表在我国多职业人群中使用的信度与效度。

方法基于“中国疾控中心职业卫生所重点人群职业病危害专项调查项目”，采用多阶段分层整群抽样方法，对教师、消防员、制造业员工、医务人员和交通警察五类典型行业重点职业人群进行问卷施测。采用验证性因素分析检验《职业倦怠通用量表》的单因素、双因素和三因素结构模型，通过卡方自由度比值（χ²/ν）、渐进残差均方和平方根（RMSEA）、适配度指数（GFI）、比较适配指数（CFI）和简约调整后的非规准适配度指数（PNFI）评价模型的拟合效度，以职业紧张、抑郁症状和睡眠障碍为校标，通过Spearman相关分析计算量表的校标关联效度，采用Cronbach's α系数和组合信度系数（CR）评价量表的内部一致性信度。

结果本研究共调查3485人，有效问卷3375份，有效应答率96.84%。验证性因素分析结果显示，在不同职业人群中，调整后的《职业倦怠通用量表》三因素结构模型（即耗竭、愤世嫉俗和职业效能）χ²/ν<4，RMSEA范围为0.032~0.069，GFI>0.90，CFI>0.90，PNFI>0.70，以上拟合指标均优于单因素或双因素结构模型。Spearman相关分析结果显示，研究对象全人群的职业紧张、抑郁症状、睡眠障碍得分分别与《职业倦怠通用量表》总分以及耗竭和愤世嫉俗维度得分呈正相关，与职业效能维度得分呈负相关，相关系数 \left| r \right| 范围为0.139~0.662（P<0.01），而消防员和交通警察的职业效能维度与抑郁症状、睡眠障碍得分相关性无统计学意义（P>0.05）。不同职业人群中总量表的Cronbach's α系数范围为0.819~0.899，CR值范围为0.941~0.978。

结论《职业倦怠通用量表》具有良好的信度、效度和不同职业人群应用效果。

Abstract:

Background As one of the common tools for job burnout assessment, Maslach Burnout Inventory-General Survey (MBI-GS), its reliability and validity across different populations in China have not been examined yet.

Objective To evaluate the reliability and validity of General Burnout Scale (GBS) by multiple occupational groups through the translation and preliminary test of MBI-GS.

Methods Based on the Special Project of Occupational Hazards in Key Populations led by the Institute of Occupational Health and Poison Control, China CDC, key occupational groups in five typical industries were selected by multi-stage stratified cluster sampling, including teachers, firefighters, manufacturing workers, medical staff, and traffic police. Confirmatory factor analysis (CFA) was used to evaluate the construct validity of GBS by single-factor, two-factor, and three-factor structure models. The model fitness was assessed by ratio of the chi-square statistic to the respective degrees of freedom (χ²/ν), root mean square error of approximation (RMSEA), goodness-of-fit index (GFI), comparative fit index (CFI), and parsimony-adjusted non-normed fit index (PNFI). Spearman correlation analysis was used to calculate the criterion validity of GBS with occupational stress, depressive symptoms, and sleep disorders. Cronbach's α coefficient and composite reliability (CR) coefficient were used to evaluate the internal consistency reliability of GBS.

Results A total of 3485 subjects were surveyed in this study, 3375 valid questionnaires were recovered with a valid response rate of 96.84%. The results of CFA showed that in the adjusted three-factor structure model of GBS (exhaustion, cynicism, and professional efficacy), the χ²/ν < 4, the RMSEA ranged from 0.032 to 0.069, the GFI > 0.90, the CFI > 0.90, and the PNFI > 0.70, which illustrated a good fitness than that of the single- or the two-factor structure models in different occupational groups. The Spearman correlation analysis results showed that occupational stress, depressive symptoms, and sleep disorders were positively correlated with exhaustion and cynicism dimensions, and negatively correlated with professional efficacy dimension of the GBS, with the \left| r \right| ranging from 0.139 to 0.662 (P<0.05) in overall subjects. For traffic police and firefighters, professional efficacy was not correlated with depressive symptoms or sleep disorders (P>0.05). The Cronbach's α coefficients ranged from 0.819 to 0.899, and the CR values ranged from 0.941 to 0.978 in different occupational groups.

Conclusion The GBS shows high reliability and validity, as well as good application effects in different occupational groups.

《职业倦怠通用量表》的多人群信度、效度分析

Reliability and validity of General Burnout Scale in multiple populations