SEM is ideal for simultaneous run of multiple constructs.

Hope this helps,

Matt

SEM is ideal for simultaneous run of multiple constructs.

Hope this helps,

Matt

Thaneswary, generally, the constructs are correlated. If this is the case, separate analyses would not provide information about the unique contributions of the variables on the dependent variable. Therefore, a single model would be preferable. Of course you need to decide about the relationships between the mediator variables. Does your theory support directed effects or not? If you do not have a theoretical basis for deciding on the relationships between the mediator variables, then you need to let the unexplained parts of these variables be correlated.

Kind regards,

Karin

you can make a single SEM for all constrcuts simultaneously to test the hypotheses

Thank you very much Mohamed, Matthew and Karin for you advice. really useful to proceed my work without confusion.