Hello all,
Concerning the normality assumption: Shapiro-Wilk test and Q-Q plots indicate that 7 of 8 dependent variables (DVs) violate the assumption. I applied a logistic transformation, but afterwards, it was still violated. I applied an inverse transformation after which 6 of the 8 DVs meet the normality assumption.
Is it okay to use an inverse transformation to be able to run a MANOVA?
Also, what can I do about the remaining 2 DVs?
When checking the linearity assumption by comparing means and testing for linearity, only 6 of the 28 pairs of DVs for each group meet the assumption. What do I need to do now?
Can I use the Box's M test for checking the homogeneity assumption?
Thank you for your attention. I am looking forward to your German or English replies.
Best, Jo