Structural Equation Modeling 1 What is SEM SEMStructural

  • Slides: 28
Download presentation
Structural Equation Modeling 結構方程模式的基本概念 與操作技術 1

Structural Equation Modeling 結構方程模式的基本概念 與操作技術 1

What is SEM? SEM是Structural Equation Modeling的縮寫 SEM又稱為causal modeling, causal analysis, simultaneous equation modeling, analysis

What is SEM? SEM是Structural Equation Modeling的縮寫 SEM又稱為causal modeling, causal analysis, simultaneous equation modeling, analysis of covariance structures, path analysis, confirmatory factor analysis SEM是一種統計方法學(statistical methodology) (Byrne, 1994) SEM是統計技術 SEM是方法學 SEM一次量化技術的大整合,也是量化方法的典範 大革命 2

Specialty of SEM Latent variables Measurement error Theory testing Multivariate statistical analysis Controlling for

Specialty of SEM Latent variables Measurement error Theory testing Multivariate statistical analysis Controlling for the errors due to statistical decision-making 3

Applications of SEM Confirmatory factor analysis (M 1) Path analysis (M 2) Structural regression

Applications of SEM Confirmatory factor analysis (M 1) Path analysis (M 2) Structural regression analysis (M 3) Time-dependent/longitudinal data (M 4) Recursive and non-recursive models for cross-sectional data Covariance structure models Multi-sample analysis Multi-level analysis 6

Four Basic Model of SEM I CFA model 7

Four Basic Model of SEM I CFA model 7

Four Basic Model of SEM II Path Analysis model 8

Four Basic Model of SEM II Path Analysis model 8

Four Basic Model of SEM III Structural Regression model 9

Four Basic Model of SEM III Structural Regression model 9

Four Basic Model of SEM IV Latent Change model 10

Four Basic Model of SEM IV Latent Change model 10

Procedures of SEM 12

Procedures of SEM 12

Modeling of SEM Basic elements: 觀察變項(observed variable) or 測量變項(measured variable) 潛在變項(latent variable) Model specification

Modeling of SEM Basic elements: 觀察變項(observed variable) or 測量變項(measured variable) 潛在變項(latent variable) Model specification Effects Direct effect Indirect effect Total effect 測量模式(measurement model) 指實際觀察值與其背後的潛在特質的相互關係 結構模式(structural model) 顯示潛在變項之間的關係 17

LISREL語法 1. TITLE Latent variable Path Model 2. DATA NI=6 NO=289 3. CM SY

LISREL語法 1. TITLE Latent variable Path Model 2. DATA NI=6 NO=289 3. CM SY FI=groupall. cov 4. LA; out 1 out 2 out 3 co 1 co 2 co 3 5. MODEL NY=3 NE=1 NK=1 NX=3 PS=DI, FR GA=FU, FI 6. LE; 員 績效 7. LK; 組織承諾 8. FREE LX(2, 1) LX(3, 1) LY(2, 1) LY(3, 1) GA(1, 1) 9. VALUE 1 LX(1, 1) LY(1, 1) 10. PD 11. OUTPUT SE TV RS AM FS EF SS SC MI 18

SIMPLIS語法 1. Latent variable Path Model 2. Observed variables: out 1 out 2 out

SIMPLIS語法 1. Latent variable Path Model 2. Observed variables: out 1 out 2 out 3 co 1 co 2 co 3 3. Covariance Matrix from file groupall. cov 4. Sample size: 289 5. Latent variables: perform comm 6. Relationships: 7. out 1= 1*perform 8. out 2= perform 9. out 3= perform 10. co 1= 1*comm 11. co 2= comm 12. co 3= comm 13. Paths: 14. Comm -> perform 15. Path Diagram 16. LISREL OUTPUT SE TV RS AM FS EF SS SC MI 19

23

23

26

26