How to set Control Variables correctly

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
dasa10
PLS Junior User
Posts: 1
Joined: Mon Feb 10, 2020 4:58 pm
Real name and title: Fabian KH

How to set Control Variables correctly

Post by dasa10 »

Hello everyone,
i am currently working on a project and using for the first time smartpls. I have a model with 3 IV and one DV. Furthermore 3 moderator variables (which describe different company measures) affect the relationships between IV and DV. I implemented two Control Variables: One describes the past behavior and is directly connected to DV and the other one describes the people knowledge/awareness of the company actions and should control the moderators.

The second CV is directly linked to the moderators. Bootstrapping shows a significant effect on all moderators, but besides that there is no further effect on the model and when i analyse the model fit -> the estimated model fit increase from 0,045 to 0.221. If I link it directly to the DV, there is no effect or significance.

My questions is: Did I linked the "moderator" CV wrong? Or is awareness of the company*s action steps just not a suitable control variable and I should take it out of the model?

I would be glad about any feedback!
Attachments
smartpls.png
smartpls.png (84.15 KiB) Viewed 22069 times
Post Reply