Specifically while executing structural modelling in R programming, nobody mentions about Indicator reliability test, composite reliability test , convergent validity test, Discriminant validity test which is necessary for CB-SEM. Any insights on this will be appreciated.
Thank you
Arun Das O.P These things tend to not be part of the "default" output of covariance-based SEM programs such as lavaan in R or Mplus, but they can certainly be computed / produced / examined based on the estimated model parameters. For example, model-based indicator reliability (R squared) can be computed for a simple structure CFA model as the squared standardized factor loading. Composite reliability indexes (e.g., McDonald's omega or Cronbach's alpha) can be computed based on the loadings and/or other parameters (depending on the type of factor model). Convergent and discriminant validity are often examined based on factor correlations (available from the completely standardized solution).
Dear Arun Das O.P
The SEM package in R offers fundamental structural equation modeling features, such as the ability to fit structural equations in observed variable models using two-stage least squares and to fit latent variable models using full information maximum likelihood while accounting for multinormality.
Best Regards
Dr. Fatemeh Khozaei
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis