PLS Latent variables interactions

[avocino]

In PLS, is it possible to measure the moderationg effect of a continuos observed variable on a latent variable?

Thanks.

A.

[joerghenseler]

Yes.
I recommend reading the article written by Chin/Marcolin/Newsted (2003).

viewtopic.php?t=21

[avocino]

Thanks very much ...

Yes.
I recommend reading the article written by Chin/Marcolin/Newsted (2003).

[ghozali]

Yes, you can run interaction latent variable using SmartPLS. The interaction latent variable is product of predictor indicators and moderator indicator. Those two indicator should be centered or standardized (you can set the data metric to mean=0 and variance=1 or choose centroid). You can compare the result between the main effect only and included the moderating variable. If the R-square change is higher with the moderating variable then the interaction effect is fine