高级检索
李跃平, 黄子杰. 验证性因子分析在量表结构效度考核中作用[J]. 中国公共卫生, 2007, 23(10): 1198-1199. DOI: 10.11847/zgggws2007-23-10-25
引用本文: 李跃平, 黄子杰. 验证性因子分析在量表结构效度考核中作用[J]. 中国公共卫生, 2007, 23(10): 1198-1199. DOI: 10.11847/zgggws2007-23-10-25
LI Yue-ping, HUANG Zi-jie. Role of confirmatory factor analysis in checking questionnaire’s construct validity[J]. Chinese Journal of Public Health, 2007, 23(10): 1198-1199. DOI: 10.11847/zgggws2007-23-10-25
Citation: LI Yue-ping, HUANG Zi-jie. Role of confirmatory factor analysis in checking questionnaire’s construct validity[J]. Chinese Journal of Public Health, 2007, 23(10): 1198-1199. DOI: 10.11847/zgggws2007-23-10-25

验证性因子分析在量表结构效度考核中作用

Role of confirmatory factor analysis in checking questionnaire’s construct validity

  • 摘要: 目的 探讨验证性因子分析在新编量表结构效度评价中的作用。方法 以乙型肝炎患者生存质量测定量表(QOL-HBV)为工具,用探索性因子分析(主成分分析、方差极大旋转法)和验证性因子分析2种方法评价量表的结构效度。结果 探索性因子分析将量表归为8个因子,每个因子能得到合理解释抽样适度测定值(KMO)0.883,解释的总方差为69.15%,将8个因子再进行因子分析,得到2个因子;用AMOS软件对8个因子进行验证性因子分析,得到二阶七因子模型,该模型的χ2/DF为1.310,拟合优度指数(GFI),调整拟合优度指数(AGFI)均>0.90。结论 探索性因子分析得出的因子间尚存在某种潜在的联系,且不能用探索性因子分析解释,用验证性因子分析可以弥补这缺陷,因此最好采用探索性因子分析与验证性因子分析相结合的方法来评价量表的结构效度。

     

    Abstract: Objective To explore the role of confirmatory factor analysis in checking the new questionnaire's construct validity.Methods Using QOL-HBV as a tool,the questionnaire's construct validity was estimated by exploratory factor analysis(principal component analysis,varimax rotation)and confirmatory factor analysis.Results Exploratory factor analysis made QOL-HBV including eight factors,every factor had reasonable explanation(KMO value was 01883,explanatory total variance was 69.15%),then using factor analysis got two factors;employing confirmatory factor analysis on the eight factors by AMOS got the two-orders and seven factors model.χ2/DF was 11310,GFI>0.90,A GFI>0.90.Conclusion The factors using exploratory factor analysis are still having some potential connections and can not be explained.Confirmatory factor analysis can redeem the defect,so it is the best way to combine confirmatory factor analysis with exploratory factor analysis to estimate the questionnaire's construct validity.

     

/

返回文章
返回