Sign of a moderator effect

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
Damien
PLS Junior User
Posts: 1
Joined: Tue Dec 16, 2014 9:32 am
Real name and title: Doctorate

Sign of a moderator effect

Post by Damien »

Hi all,
I need assistance to analyse the way a moderator variable impacts a relation between 2 variables. In other words, if the MV impacts positively or negatively a relation. It’s quite complicated for me because I am a novice with SmartPls.

Let me show you my results :
1. IV A to DV C (R² 0.363, t 4.216) -> MV - Group 1 (R² 0.363, t 3.110) and Group 2 (R² 0.759, t 4.630)
2. IV B to DV C (R² -0.002, t 2.054) -> MV - Group 1 (R² -0.002, t 0.042) and Group 2 ( R² 0.073, t 8.889)

Assumption 1 : there is a positive relation between A to C and a negative between B to C.
Assumption 2 : there exists a moderator impact for both this relations (for the first one, I used the Excel sheet)

My questions are the following:
1. Does the MV impact positively or negatively the relation between A and C, and how strong ?
2. Same question for the relation between B and C.
3. Is it possible that the sign of a relation between two variables would be changed due to sign of the MV impact?

I hope I explained clearly the problem and my questioning.
Thanks a lot for your answers.
Post Reply