Hierarchical problems

Ricardofrio · Post by **Ricardofrio** » Fri Jan 24, 2014 10:18 pm

I have a question regarding the PLS.

I'm working with a formative and third-order hierarchical model. After testing the hierarchical path coefficient in both second-order constructs, the results were significant. However, after performing the hypothesis test adding two endogenous variables to the third-order variable, the significance and the path coeficient of a second-order construct changed in relation to the third-order construct.
I was wondering if you have an explanation for that fact or if you could recommend any literature on the subject.
Thank you
Yours truly,

Hengkov · Post by **Hengkov** » Sat Jan 25, 2014 7:09 am

Hi,

May be you have suppression problem. For check it, used WarpPLS 4.0. For literature support, my book WarpPLS 4.0 (2nd) in press now.

Best Regards,