SMARTPLS 4 WORKSHOP (PART 1):
BASIC STRUCTURAL EQUATION MODELING
(PLS-SEM) DATA ANALYSIS USING SMARTPLS 4
Prof Dr Chua Yan Piaw
University of Malaya
Content
Four data analysis methods in SmartPLS 4
1. PLS-SEM
2. CB-SEM
3. PROCESS
4. REGRESSION
1
1. PLS-SEM
Non-parametric SEM analysis
Variance-based SEM (VB-SEM)
No requirement of normality of data
2
2. CB-SEM
Parametric SEM analysis
Covariance-based SEM (CB-SEM)
Requirement: normality of data
3
3. PROCESS
Parametric SEM analysis
Assumbed equal weights among indicators
May control external variables (confounding
variables that influence the dependent variables)
4
4. REGRESSION
a. Parametric analysis
Requirement: normality of data
Like a normal multiple linear
regression analysis in SPSS
Click here for explanation for Intercept.
4. REGRESSION
b. Non-parametric analysis
No requirement of normality of data
Logistic regression analysis in SPSS
Dependent variable: Dichotomy data
10
5
Choosing a data analysis method in
SmartPLS depends on:
The scales of measurement of the
data
Normality of Data
The objective of the analysis
11
RESEARCH STATISTICS
DR CHUA YAN PIAW (UM)
12