To explore and revise the factor structure, reliability and validity of the Chinese version of the Driver Stress Inventory (DSI) in the driver population in first-tier city of China. In this study, the questionnaire method was used to select the data collected by the "Research on the Driving Stress of Urban Salaries in Urban Traffic" carried out by the Institute of Psychology, Chinese Academy of Sciences in 2012, and a third-party survey agency was commissioned to select 300 people in Beijing, Shanghai, and Guangzhou respectively. A total of 900 drivers (with 889 valid data scales) were used to test their driving stress and other indicators by using the DSI and the self-compiled travel and stress relief scale. The test content collects driver self-reported stress data from five dimensions: Aggression, Dislike of driving, Hazard-Monitoring, Fatigue Proneness and Thrill-seeking. The DSI questionnaire of 445 cardinal numbers was revised by item analysis methods such as correlation test and T-test, and exploratory factor analysis method based on principal component analysis and optimal skew axis method (Promax).The results of item analysis showed that the total correlation coefficient of 10 items of the 48 items of the original DSI scale was lower than 0.3, and the total correlation coefficient of 6 items was not significant (r=-0.078-0.079, P>0.05), and the high and low groups were independent. There were significant differences in the results of the sample t test (t=-16.642-0.091, P<0.001), the 16 items were deleted, and the remaining 32 items; exploratory factor analysis showed that KMO=0.938>0.900, and the Bartlett's sphericity test result was significant (χ²=6 361.974, df=496, P<0.001), suitable for exploratory factor analysis, the results showed that 2 items constituted independent factors, did not meet the relevant standards of psychometrics and were deleted, and finally retained 30 items, and the internal consistency coefficient of the new scale was better than the original one(α=0.932>0.877); Based on the results of exploratory factor analysis, the model fitting indexes of 444 even-numbered samples such as RMSEA, SRMR, CFI and TLI were verified by confirmatory factor analysis, and the results showed that the index of each index was good (χ²=1 250.447, RMSEA=0.070, SRMR=0.068, CFI=0.839, TLI=0.823); criterion validity analysis found that each factor of the revised scale was significantly correlated with situational anxiety (r=0.190-0.556, P<0.01). In conclusion, the DSI (Chinese version) has good reliability and validity, and can be used as an assessment tool for driver stress in China.
为探讨并修订驾驶员压力量表（driver stress inventory，DSI）中文版在我国一线城市（北京、上海及广州）驾驶员人群中的因子结构及信效度。本研究采用问卷调查法，选取中国科学院心理研究所2012年开展的“城市工薪族市内交通出行焦虑情绪研究”收集的数据，该数据通过委托第三方调查机构在北京、上海及广州三地分别选取300名（共900人，有效数据量表889份）驾驶人员，采用驾驶员压力量表（DSI）和自编出行及压力缓解量表，通过线上作答的方式对其驾驶压力等指标进行测查，测查内容主要从驾驶攻击、不喜欢驾驶、危险监控，驾驶疲劳和刺激寻求五个维度收集驾驶员自陈压力数据。将总样本按奇偶分组，采用相关和t检验等项目分析方法及基于主成分分析法和最优斜交转轴法（Promax）的探索性因素分析方法对奇数编号445份样本进行DSI问卷修订；项目分析结果显示，原DSI量表48个条目中10个条目题总相关系数低于0.3，6个条目的题总相关关系不显著（r=-0.078~0.079，P>0.05），高低分组独立样本t检验结果均存在显著差异（t=-16.642~0.091，P<0.001），删除此16个条目，剩余32个条目；探索性因素分析表明KMO=0.938>0.900，Bartlett球形检验结果显著（χ²=6361.974，df=496，P<0.001），适合进行探索性因素分析，结果显示其中2个条目组成独立因子，不符合心理测量的相关标准并删除，最终保留30个条目，且新量表的内部一致性系数优于原量表（α=0.932>0.877）。基于探索性因素分析结果，进一步采用验证性因素分析方法对偶数编号444份样本的RMSEA、SRMR、CFI、TLI等模型拟合指标进行验证；结果显示，各项指标拟合指数良好（χ²=1 250.447，RMSEA=0.070，SRMR=0.068，CFI=0.839，TLI=0.823）；效标效度分析发现，修订后量表各因子与情境焦虑显著相关（r=0.190~0.556，P<0.01）。综上，DSI（中文版）的信效度良好，可作为我国驾驶人员压力的评估工具。.