MANCOVA and MANOVA Discussion > Which statistic to use from my MANCOVA
Marissa,
the differences between an IV and covariate are typically merely conceptual, not mathematical (in most cases). However, to determine if your case is an instance where there are meaningful difference, what is your perspective about what the differences are/should be?
June 22, 2012 |
Jeremy Taylor
Jeremy Taylor
I have performed a MANCOVA and now in Field I see that i should use the 'multivariate tests' output instead of the usual 'test of between-subject effects' output. I've used two covariates. The multivariate test output (like pillai's trace) gives me f and p values for my independent value and my covariates, so now it seems like it's just testing my covariates as if they were independent variables, which is not what I want. I'm confused. which statistic do I need to see if there are any effects on the dependent variables of the independent variable, corrected for the effects of the covariates? and how can i see whether there are differences between my two dependent variables? I know there are, because before when i used to ancova's one of them was significant and one was not.
I hope you can help me!