Questions about MANOVA

#1
Hello,

Appreciate if someone can help me. I am a STAT dummy that tries hard to learn statistics.

I have two questions about MANOVA.
I ran a model of 2 IVs (FACTOR A--6 levels and FACTOR B-2 levels) and two DVs (VARIABLE C and D: 1-7 likert scale) and had some confusions.

1. Factor A appears to be significant in overall MANOVA but when examining the SPSS output of "tests of between subjects effects), factor A is not significant on each individual DV. In this case, is it still legitimate to claim A is significant in the model?

2. My interaction term (A*B) is not significant in the overall model but the graphic plot shows that few levels of A indeed has interaction with B. Is the insignificance in the model can be due to to that I have too many levels (6) of A and not each has interaction? Will increase the cell size help? (right now I have around 18-28 each cell). can I run each level of A respectively to show interaction?

Thank you so much.
Chloe
 
Last edited: