Please enable JavaScript.
Coggle requires JavaScript to display documents.
PLS - Coggle Diagram
PLS
-
Process
Variance based SEM
Exploratory approach:
- Partial least squares path modeling (PLS)
- Path analysis with sum scores
- Generalized structured component analysis
Covariance based SEM
Confirmatory approach:
- Max. likelihood (ML)
- Weighted least squares (WLS)
- Unweighted least squared (ULS)
- Generalized least squares (GLS)
- Asymptotically distribution free (ADF)
-
-
Specify structural model
-
-
Which relationships between these variables:
- Linear vs non-linear
- Moderation
- Mediation
- No relationship
Intro
Summary why SEM
Modeling causal relationship within nomological net:
- Incl. of partial & total effects of constructs
- Representatioon & testing of entire theories
- Context-sensitive approximation of latent variables
Taking measurement error into account:
- Assessing measurement reliability/validity
- Correcting for measurement error
-
-
Why structural equation modeling (SEM):
- Each technique so far -> examines only single relationship at a time
- SEM: Series of dependence relationships simultaneously & nomological network
Important sequence:
- First, measurement model (corresponds to FA)
- Second, structural model (corresponds to MRA)
-
-