基于二阶验证性因子分析的SF-36量表结构效度的分析  

Construct Validity Analysis of SF-36 Scale Based on Secondorder Confirmatory Factor Analysis

作　　者：徐秀娟[1] 倪进东[1] 王效军[1] 

机构地区：[1]广东医学院公共卫生学院523808

出　　处：《中国卫生统计》2013年第6期846-848,共3页Chinese Journal of Health Statistics

基　　金：流行病与卫生统计学校级重点学科(XZ1105)

摘　　要：目的采用二阶验证性因子分析的方法对SF-36量表的结构效度进行分析。方法分别构建SF-36量表的一阶验证性因子分析模型和二阶验证性因子分析模型进行等同模型的对比研究,并对拟合结果进行分析。结果两个模型的RMSEA均小于0.08,NNFI和CFI均大于0.90,均拟合较好。二阶验证性因子分析模型相对一阶验证性因子分析模型的Δχ2=7.49<χ2(4)=9.49,P>0.05,不具有统计学意义。二阶验证性因子分析模型的一阶因子载荷和二阶因子载荷的标准解均大于0.30,所有因子载荷非标准解t检验的P<0.05,均具有统计学意义。结论二阶验证性因子分析模型可用于SF-36量表结构效度的分析,符合SF-36的理论构想,而且模型相对简约。Objective To analyze construct validity of SF-36 applying second-order confirmatory factor analysis. Methods Both first- order confirmatory factor analysis model and second-order confirmatory fac- tor analysis model were constructed of the SF-36 scale and the results ofgoodness of fit as equivalent model were compared and analyzed. Results RMSEA of the two models are less than 0. 08, NNFI and CFI are greater than 0. 90, both models fitted better. Relative to the first-order confirmatory factor analysis model, Second-order confirmatory factor analysis model adds ~0（2 = 7.49 〈X^4~ = 9. 49, P 〉 0.05. Standard solutions of the fwst-order factor loadings and second-order factor loadings of second order confirma- tory factor analysis model are greater than 0. 30. T-test of all nonstandard factor loadings have P 〈 0. 05. Conclusion Second-order confm-natory factor analysis model can be used to analyze construct validity of SF-36, and consistent with the theoretical conception of the SF-36, and model is relatively simple.

关 键 词：SF-36量表 二阶验证性因子分析 结构效度 

分 类 号：R195[医药卫生—卫生统计学]