I'm planning on conducting a linear regression analysis with 5 moderator variables (Big 5 personality traits) but couldn't find any example. I know that it's theoretically possible to include that many moderators, though it would be very helpful to read how other people dealt with it.
Any publication (in any field of research) with more than 2 moderators would be helpful! Thank you :-)
Hi Sophia,
unfortunately, I can not refer to any paper but I would recommend to test one product term at the time (however having all other moderators in any model). Adding several product terms at once will further increase the multicolinearity that is essential in a moderator analysis anyway. Yes, you increase the danger of finding spurious moderators but as in any case the question is which error you want to make (type I or II).
Hope that helps
Holger
I agree with Holger Steinmetz. Here is another way to think of it. If the 5 moderator variables are inter-correlated, so you probably wouldn't want to run them all in the same moderator analysis due to multi-colinearity. Alternatively, if they aren't strongly correlated, then you might as well run them separately because they won't affect each other.
- Badges
- Science topic
- Similar topics
- Mathematics
- Algebra
- Matrix