验证性因子分析模型的潜变量得分及其应用  

LATENT VARIABLE SCORES OF CONFIRMATORY FACTOR ANALYSIS AND ITS APPLICATIONS.

作　　者：张岩波[1] 刘桂芬[1] 郑建中[1] 徐秀娟[1] 

机构地区：[1]山西医科大学卫生统计教研室,太原030001

出　　处：《现代预防医学》2005年第4期285-286,299,共3页Modern Preventive Medicine

基　　金：国家自然科学基金资助项目 (30 2 0 0 2 36 )

摘　　要：目的:当某些情况下不宜采用结构方程模型分析时,通过估计获得潜变量得分,将之作为中间产物,便于进一步做其它的统计分析。方法:探讨验证性因子分析中潜变量得分的估计原理,潜变量得分的估计采用LISREL软件编程实现。结果:利用一组职业紧张调查数据例证了潜变量得分的估计方法,通过对职业紧张量表(OSI -R)测量结果分析,结果符合专业解释。结论:潜变量得分因克服了测量误差的存在。Objective:The estimated latent variable scores were obtained and convenient for other statistical analysis when Structural Equation Model was not suitable in some conditions.Methods:The latent variable scores theory of confirmatory factor analysis was discussed and latent variable scores were estimated with LISREL program.Results:Occupational stress survey data were utilized to exemplify the latent variable scores estimated method and the analysis results OSI-R coincided with professional interpretation.Conclusion:Latent variable scores were more valid than sum scores for it counteracted the influence of measurement error.

关 键 词：潜变量 分析模型 验证性因子分析 测量结果分析 职业紧张 模型分析 中间产物 统计分析 编程实现 估计方法 测量误差 

分 类 号：R195.1[医药卫生—卫生统计学] R532.21[医药卫生—卫生事业管理]