【正文】
SEM是統(tǒng)計(jì)技術(shù)– SEM是方法學(xué)n SEM一次量化技術(shù)的大整合,也是量化方法的典範(fàn)大革命3Specialty of SEMnLatent variablesnMeasurement errornTheory testingnMultivariate statistical analysisnControlling for the errors due to statistical decisionmaking4Historical roots of SEMn 心理計(jì)量根源– Galton( relationship in quantitative study )– Spearman: 因素分析( factor analysis)– Thurston:最簡(jiǎn)化結(jié)構(gòu)( simple structure) – J246。 12Modeling of SEMn Basic elements:– 觀察變項(xiàng) (observed variable) or 測(cè)量變項(xiàng)(measured variable)– 潛在變項(xiàng) (latent variable)n Model specification– 測(cè)量模式 (measurement model) ? 指實(shí)際觀察值與其背後的潛在特質(zhì)的相互關(guān)係– 結(jié)構(gòu)模式 (structural model)? 顯示潛在變項(xiàng)之間的關(guān)係n Effects– Direct effect– Indirect effect– Total effect13Procedures of SEM14技術(shù)原理:參數(shù)估計(jì)? Matrix15Parameters in SEMn Fixed parameters – assigned specific v