公立医疗机构医生工作满意度结构的验证性因素分析

Confirmatory Factor Analysis of Job Satisfaction Structure of Physicians from Public Medical Institutions

目的通过探讨公立医疗机构医生工作满意度的潜在因素结构,评价测量工具的结构效度。方法自编医生工作满意度问卷(PJSQ-PMI),根据预调查探索性因素分析(EFA)结果提出假设模型,利用正式调查数据(N=1451)进行多样本、多阶层验证性因素分析(CFA)。结果由58个原始指标组成的五因素模型Model A、A'的拟合效果不佳,而经项目组合后的简化修饰模型Model B'(含13个指标)与实际数据的拟合状态理想(GFI、NFI、CFI、IFI>0.9,RMSEA=0.075,RMR=0.024),在不同地区的样本中都获得支持;进一步构建的二阶模型Model C同样达到拟合要求(RMSEA=0.086,RMR=0.025,γ≥0.64),与Model B'的拟合卡方值之差△CMIN=10.163,△DF=5,P>0.05,根据简效法则接受Model C为较佳模型。结论经EFA与CFA相结合的多样本交叉检验显示,公立医疗机构医生工作满意度的潜在结构为简化的五因素二阶模型,表明问卷总体上具有较好的结构效度和测量稳定性。

Objective To assess construct validity of measurement tool through a probe into latent factorial structure of physicians job satisfaction from public medical institutions. Methods Multi-sample, multi-order confirmatory factor analysis （CFA） was used in inspection of hypothesis models constructed according to the result of exploratory factor analysis（EFA）. Formal data of three cities were surveyed by self-compiled PJS-PMI questionnaire. Results Though the five-factor model A＆A made of fifty-eight original indicators wasn t fitted well, the simplified model B （ including thirteen indicators ） through item parceling could be supported ideally by entire sample data（ GFI,NFI, CFI,IFI 〉 0. 9, RMSEA = 0. 075, RMR = 0. 024 ）, and the same as by three single samples. Moreover,the second-order model Craised on the basis of model B ,could also reach the goodness-of-fit standards （ RMSEA = 0. 086, RMR = 0. 025, γ≥ 0. 64, ΔCMIN = 10. 163, ΔDF = 5, P 〉 0.05 ） and be accepted ultimately according to the brief principle. Conclusion Multi-sample cross inspection through the combination of EFA and CFA indicated that the latent structure of physicians job satisfaction from public medical institutions was the simplified second-order model consisted of five factors, and the PJS-PMI questionnaire enjoyed good construct validity and measurement stability.