Hello everyone,

I have searched the forum and couldn't find an answer, so I figured I would start a new thread. I am running a 1-way MANOVA for the effect of school program (categorical variable) on 11 likert scale variables about perceived implementation of the program. Sample size is 129 in group 1 and 109 in group 2.

Firstly, upon visual inspection the data is not normally distributed, but the professor wants the analysis done anyway. After running the analysis I found the Box's test was significant (p<0.001) and has unequal groups. I removed outliers based upon Mahalanobis Distance and reran the MANOVA, still p<0.001. I brought up the covariance matric and everything appeared to be "normal". No variable was 3x larger than the equivalent variable in the other group.

I am kinda at a loss on what to do and how to interpret this Box's test. Can I go forward and assume equality of covariance? Any help would be greatly appreciated