通用版工作倦怠量表的编制及信效度研究

Job Burnout Scale-General Survey: Development,Reliability,and Validity

目的编制通用版工作倦怠量表,并检测其信效度。方法根据文献研究,以已有相关量表为基础,结合开放性问卷调查和半结构式访谈所获资料,建立通用版工作倦怠量表内容框架和条目池,采用"出声思维"法等修改、删除、合并条目,最终形成初始通用版工作倦怠量表。2013年1—2月,采用整群抽样法选取广州、厦门、泉州市企、事业单位员工450名为调查对象,采用一般资料问卷、初始通用版工作倦怠量表、中文版Maslach倦怠量表通用版(MBI-GS)对其进行调查。共发放问卷450份,回收有效问卷406份,有效回收率为90.2%。采用简单随机抽样法将调查对象分为A组(300名)和B组(106名),分别用于构建通用版工作倦怠量表及其信效度检验。采用简单随机抽样法从B组中选取50名企、事业单位员工,于7~10 d后采用初始通用版工作倦怠量表进行重测。通过项目分析、同质性检验和探索性因子分析(EFA)对条目进行筛选,进而形成通用版工作倦怠量表。结果通过项目分析、同质性检验、EFA,最终形成通用版工作倦怠量表,其包括情绪耗竭(包括7个条目)、躯体不适(包括8个条目)、身心疲惫感(包括5个条目)、认知行为效率下降(包括4个条目)4个维度、24个条目,累积贡献率为60.58%。通用版工作倦怠量表及其各维度的项目间平均相关系数(Miic)为0.41~0.52,Cronbach'sα系数为0.79~0.94,分半信度系数为0.71~0.92,重测信度系数为0.93~0.98。验证性因子分析(CFA)结果显示,χ2/df=1.76,RMSEA=0.08,RMR=0.05,NFI=0.77,TLI=0.86,IFI=0.88,CFI=0.88。中文版MBI-GS情感耗竭维度得分与通用版工作倦怠量表各维度得分及其总分的相关系数为0.42~0.60(P<0.001);中文版MBI-GS玩世不恭维度得分与通用版工作倦怠量表各维度得分及其总分的相关系数为0.46~0.66(P<0.001);中文版MBI-GS成就感低落维度得分与通用版工作倦怠量表情绪耗竭、躯体不适、身心疲惫感维度得分及其总分的相关系数为0.25~0.34(P<0.001);中文版MBI-GS总分与通用版工作倦怠量表各维度得分及其总分的相关系数为0.45~0.65(P<0.001)。结论本研究编制的通用版工作倦怠量表,包括情绪耗竭、躯体不适、身心疲惫感、认知行为效率下降4个维度及24个条目,具有较好的信度及效度。

Objective To develop a Job Burnout Scale-General Survey（ JBS-GS）,and test its reliability and validity. Methods We developed the content framework and item pool of the JBS-GS by reviewing the related literature,other scales,results of open-end interviews and semi-structured interviews. After modifying,deleting or combining the items by think aloud and other methods,the initial JBS-GS was formed. Four hundred and fifty employees of enterprises and institutions were selected by cluster sampling from the cities of Guangzhou, Xiamen and Quanzhou from January to February 2013. The Demographic Questionnaire,the initial JBS-GS and Maslach Burnout Inventory-General Survey [MBI-GS（ Chinese version） ]were used to investigate them. Of the 450 sets of questionnaires administered,406 were returned giving a response rate of 90. 2%. Then the 406 employees were randomly divided to group A（ n = 300） and group B（ n = 106） using simple random sampling to develop the JBS-GS and study its reliability and validity respectively. Seven to ten days later,50 from group B were selected with simple random sampling and surveyed by the initial JBS-GS again. The item analysis, homogeneity test,exploratory factor analysis（ EFA） were used to select the items and develop JBS-GS. Results After item analysis,homogeneity test,and EFA, the JBS-GS was developed, which consists of 4 dimensions, including emotional exhaustion（ 7 items）,somatic distress（ 8 items）, physical and mental fatigue（ 5 items）, and cognitive and behavioral impairment（ 4 items）,contributing 60. 58% to the total variance. The mean inter-item correlation coefficients（ Miic）,Cronbach＇s α coefficient,split-half reliability coefficient,test-retest reliability coefficient of JBS-GS and its 4 dimensions were 0. 41-0. 52,0. 79-0. 94,0. 71-0. 92,0. 93-0. 98,respectively. Confirmatory factor analysis（ CFA） result showed that χ2/df = 1. 76,RMSEA = 0. 08,RMR = 0. 05, NFI = 0. 77, TLI = 0. 86, IFI = 0. 88, CFI = 0. 88. The correlation coefficients between score of emotional exhaustion dimension of MBI-GS（ Chinese version） and scores of other 3 dimensions of JBS-GS and the total score of JBS-GS were 0. 42-0. 60（ P 0. 001）. The correlation coefficients between score of cynicism dimension of MBI-GS（ Chinese version）and scores of 4 dimensions of JBS-GS and total score of JBS-GS were 0. 46-0. 66（ P 0. 001）. The correlation coefficients between score of diminshed personal accomplishment dimension of MBI-GS（ Chinese version） and score of 3 dimensions of JBS-GS（ emotional exhaustion,somatic distress,physical and mental fatigue） and total score of JBS-GS were 0. 25-0. 34（ P 0. 001）. The correlation coefficients between total score of MBI-GS（ Chinese version） and scores of 4 dimensions of JBS-GS and total score of JBS-GS were 0. 45-0. 65（ P 0. 001）. Conclusion The JBS-GS developed by this study consists of 4 dimensions,including emotional exhaustion, somatic distress, physical and mental fatigue, and cognitive and behavioral impairment,and 24 items. It has good reliability and validity.