Hello everyone.
Recently, when I searched the literature about SEM, I found some problems that I couldn't solve.
I found that the models established in some studies are all observation variables (there are no latent variables, all from the results of the questionnaire). But such a model cannot be implemented in AMOS.
Can anyone help me? Thank you very much.
Could you please explain more about " no latent variables" or send a paper which you read about it .
Best,
The reasons you only saw the observable variables are:
I. The variables had only one item to measure the constructs, or
Ii. The constructs are measured in a preceding section after following all the requisite procedure of the measurement model as we do in CFA. Once the latent constructs are imputed you can use them straight in the Amos -structural models without providing all the indicators/items. However, if the variables are directly measurable through observable variables then off course there is no need to do the CFA and measurement model estimation. You could share further details to see how and why these observable variables were included in the Amos model estimation.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling