Hi all,
in assessing validity for my measurement scales i came across the nomological validity. In several articles and books it is suggested to assess nomological validity just for formative constructs; a similar suggestion for reflective constructs is missing. What is the reason for this? Why is nomological validity in PLS not assessed for reflective constructs as it is done in covariance based SEM?
And a second question: Is there a difference between nomological validity in measurement model vs. structural model? Do I have to check for nomological validity in structural models (as it is suggested to check, whether the constructs behave in the way predicted by theory) when i only use reflective constructs? If yes, how do i do this? if no, why not?
Thanks for any hints and references in advance!
Sam