Moderating effect using MGA

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
jklstw
PLS Junior User
Posts: 8
Joined: Mon Jun 28, 2010 5:17 am
Real name and title: Associate Prof. Jyun-Kai Liang

Moderating effect using MGA

Post by jklstw »

Hi everyone,

Please I need your help

I have a model includes 2 independent variables (A and B), a dependent variable (C) and a moderator G(gender).
A ---->B ----> C
A -----------> C

If I conduct GMA, does it mean that G moderating the effect of three paths(A--->B, B--->C and A--->C) simultaneously?

If I suppose G only moderating the B-->C and A--->C, I need to conduct two separate analysis B*G-->C and A*G--->C, right?

Then, if I conduct B*G-->C and A*G--->C in a model simultaneously, what are the differences comparing to the above separate two models?

Greetings
jmbecker
SmartPLS Developer
Posts: 1282
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Moderating effect using MGA

Post by jmbecker »

jklstw wrote: Sat Mar 17, 2018 12:15 am Hi everyone,

Please I need your help

I have a model includes 2 independent variables (A and B), a dependent variable (C) and a moderator G(gender).
A ---->B ----> C
A -----------> C

If I conduct GMA, does it mean that G moderating the effect of three paths(A--->B, B--->C and A--->C) simultaneously?
Yes
jklstw wrote: Sat Mar 17, 2018 12:15 am If I suppose G only moderating the B-->C and A--->C, I need to conduct two separate analysis B*G-->C and A*G--->C, right?
You can also include both interaction effects in the same model. You do not need to do this separately. See below.
jklstw wrote: Sat Mar 17, 2018 12:15 am Then, if I conduct B*G-->C and A*G--->C in a model simultaneously, what are the differences comparing to the above separate two models?

Greetings
It is a different assumption to test them both separately or together. Usually, if you assume both to be necessary you should test them together. However, multicollinearity could be a problem is such a model.
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
jklstw
PLS Junior User
Posts: 8
Joined: Mon Jun 28, 2010 5:17 am
Real name and title: Associate Prof. Jyun-Kai Liang

Re: Moderating effect using MGA

Post by jklstw »

Dr. Jmbecker, thank you so much!
Post Reply