基于经典测量理论和项目反应理论的《职业紧张简易调查问卷》适用性研究

Study on the applicability of the Brief Job Stress Questionnaire based on classical test theory and item response theory

目的评估《职业紧张简易调查问卷》(Brief Job Stress Questionnaire,BJSQ)在制造业和服务业员工中的问卷质量和条目特性。方法于2021年12月至2022年12月,采用目的抽样和方便抽样相结合的方法,选取北京市制造业和服务业各4家企业员工2077人为研究对象。采用中文版BJSQ从工作紧张因素、紧张反应、社会支持因素和满意度4个维度、57个条目开展职业紧张状况调查,采用经典测量理论(classical test theory,CTT)分析问卷的效度和信度,采用项目反应理论(item response theory,IRT)中的Semejima等级反应模型分析各条目的区分度a、难度系数b和条目信息量。结果2077名研究对象年龄中位数[M(Q_(1),Q_(3))]为33(28,37)岁,工龄M(Q_(1),Q_(3))为4(2,8)年,制造业和服务业分别为723(34.8%)、1354(65.2%)人。经探索性因子分析提取11个公因子,累计方差贡献率为62.823%,各条目公因子方差为0.451~0.865,因子载荷值范围为0.413~0.825,其中有5个条目存在交叉载荷;验证性因子分析结果显示,模型拟合度指标中近似误差均方根为0.055,比较拟合指数为0.950,Tucker-Lewis指数为0.948,标准化均方根残差为0.066;内容效度结果显示,中文版BJSQ总得分与4个维度得分均呈正相关关系(r_(s)=0.487~0.936,P<0.05)。信度分析结果显示,总问卷的Cronbach'sα系数为0.945,工作紧张因素、紧张反应和社会支持因素3个维度的Cronbach'sα系数分别为0.775、0.957、0.830;总问卷的Spearman-Brown系数为0.866,3个维度的Spearman-Brown系数分别为0.572、0.882和0.772。IRT分析结果显示,57个条目中仅1个条目的区分度a值<0.30;各条目的难度系数b值为-12.02~11.09,其中8个条目难度系数过高,3个条目难度系数过低,3个条目不符合要求;各条目的平均信息量为0.022~2.566,有47个条目的平均信息量>0.088(5/57)。结论中文版BJSQ在我国制造业和服务业职业人群中具有良好的信效度,大部分条目性能较好,但仍有部分条目需要进一步改进优化或删除。

ObjectiveTo evaluate the quality and item characteristics of the Brief Job Stress Questionnaire(BJSQ)among employees in manufacturing and service industries.MethodsFrom December 2021 to December 2022,a total of 2077 employees from 4 manufacturing and service enterprises in Beijing were selected by the method of combining purpose sampling and convenience sampling.The Chinese version of BJSQ was used to carry out a survey on occupational stress from 57 items in 4 dimensions including job stress factors,stress response,social support factors and satisfaction.Classical test theory(CTT)was used to analyze the validity and reliability of the questionnaire.The Semejima hierarchical response model in item response theory(IRT)was used to analyze the differentiation a,difficulty coefficient b and information content of each item.ResultsAmong the 2077 subjects,the age M(Q_(1),Q_(3))was 33(28,37)years old,and the working age M(Q_(1),Q_(3))was 4(2,8)years.There were 723(34.8%)people in manufacturing industry and 1354(65.2%)people in service industry.Eleven common factors were extracted by exploratory factor analysis,with a cumulative variance contribution rate of 62.823%,the variance of the common factors of each entry ranging from 0.451 to 0.865,and the range of factor loading values from 0.413 to 0.825,with 5 items having cross-loadings.The results of the validation factor analysis showed that the model fit indexes of root means square error of approximation was 0.055,comparative fit index was 0.950,Tucker Lewis index was 0.948,and standardized root mean square was 0.066.Content validity results showed that the total scores of the Chinese version of the BJSQ were positively correlated with the scores of the 4 dimensions(r_(s)=0.487-0.936,P<0.05).The results of the reliability analysis showed that the Cronbach's alpha coefficient for the total questionnaire was 0.945,and the Cronbach's alpha coefficients of job stress factors,stress response and social support factors were 0.775,0.957,and 0.830,respectively.The Spearman-Brown coefficient for the total questionnaire was 0.866,and for the 3 dimensions,it was 0.572,0.882 and 0.772.The results of IRT analysis showed that only 1 of the 57 items had a differentiation a-value<0.30,the difficulty coefficient b-value of each item ranged from-12.02 to 11.09,of which 8 items had a difficulty coefficient that was too high,3 items had a difficulty coefficient that was too low,and 3 items did not meet the requirements,and the average amount of information in each item ranged from 0.022 to 2.566,and there were 47 items with average amount of information>0.088(5/57).ConclusionThe Chinese version of BJSQ has good reliability and validity in the typical occupational groups of manufacturing and service industries in China,and most of the items have good performance,but some items still need to be further improved and optimized or deleted.