More than one moderator- each on two relationships

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
MaFa
PLS User
Posts: 12
Joined: Fri Aug 23, 2019 11:32 pm
Real name and title: Ana Foster, PhD student

More than one moderator- each on two relationships

Post by MaFa »

Hello all,
My research model has two moderators, each of them moderating two relationships. As I read posts, I need to test each moderator at a time, the question is should I test each on each relationship at a time or each moderator on both relationships? I think the second way (testing each moderator on both relationships) and running two models make more sense. But the issue is when I test each moderation separately and running 4 separate models, I get better results, but testing a moderator on both relationships does not give the results I can get by testing each separately. Any suggestion?

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

Re: More than one moderator- each on two relationships

Post by jmbecker »

Generally, testing everything in one model would be the best way to do. Only if interaction effects are not significant they may be excluded from a model. Otherwise you may be testing artifacts.
However, given that interaction effects usually introduce some collinearity into the model and several moderating effects even worsen this problem, people often advise to test them sequentially. If the two relationships go into the same construct (same DV) then this recommendations / problem also holds.
If you have the two relationships into different constructs i would certainly test them together in one 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
Post Reply