MANCOVA interaction effect

#1
Hello everyone,

I am facing the following problem:

I conducted a MANCOVA to investigate whether a control group significantly differs from 2 other experimental groups with regard to 4 different dependent variables. There was a covariate included which was supposed to moderate the relationship between the experimental situation and three out of the 4 dependent variables.

I found the following: The main effects of the experimental situation have been significant (also without the covariate). When including the covariate, its main effect is not significant, the significance of the first main effect (the experimental situation) does NOT change BUT the interaction term of the experimental situation and the covariate is significant for 2 out of 3 dependents.

What does that mean? Does it mean that the experiment causes significant change but to be able to explain ALL CHANGES caused with the dependent variable one needs to also consider the covariate? I am a bit lost with that one...

Thank you for your help!

Best regards,
Gerrit
 
Last edited: