CB-SEM allows for the comparison between observed and proposed covariance matrices, which enables assessment of the overall “fit” of one suggested causal model.
Refer to :
- Lowry et al., "Partial Least Squares (PLS) Structural Equation Modeling (SEM) for Building and Testing Behavioral Causal Theory: When to Choose It and How to Use It", 2014 -
Article Partial Least Squares (PLS) Structural Equation Modeling (SE...
Covariance-based structural equation modeling (CB-SEM) is a statistical technique used to estimate causal relationships between variables. It involves specifying a theoretical model with relationships between latent variables, collecting data on manifest variables that measure the latent constructs, and then estimating the parameters of the model.
Some key characteristics of CB-SEM:
It models latent constructs that are not measured directly but are estimated from observed variables.
It estimates the model parameters by minimizing the difference between the sample covariances and the covariances implied by the theoretical model.
It allows for complex relationships between variables including reciprocity, mediation, moderation, and correlations.
Model fit is assessed using a range of goodness-of-fit statistics to evaluate how well the hypothesized model reproduces the observed covariances.
It assumes the data follows a multivariate normal distribution. Violations can affect the reliability of estimates and statistical tests.
It uses a confirmatory approach where a theory-driven model is specified a priori and then tested against the data.
- Badges
- Science topic