I have collected data on a set of student variables in an undergraduate course (GPA, frequency of participation in class discussion, homework completion, exam performance, and year in school). My main hypotheses pertained to exam performance and homework completion (using different reinforcement contingencies), however, prior to conducting that analysis I wanted to first determine if there were significant differences in the above-mentioned variables across year in school (e.g., freshman, sophomore, junior...). If there were significant differences, then I would control for those differences in subsequent analyses. My sample size is around 80 students for this study.

To test for differences across year in school, I used a MANOVA as I thought it was appropriate for multiple DVs with only one IV (year in school). Does that seem like the appropriate test to use or should I have used multiple ANOVAs (as others have advised I should have done)?

Similarly, I wanted to determine if students might exhibit significant differences on multiple DVs (exam performance, homework completion, and participation in class discussion) based on a combination of their critical thinking ability and GPA. Thus multiple DVs and multiple IVs. I ran a Multivariate GLM through SPSS, which indicated it was a MANOVA and produced a Wilk's Lambda score along with subsequent ANOVA post-hocs. Should this have been a MANCOVA instead of a MANOVA?

I sincerely appreciate any insight and advice that can be provided. I feel like I've been staring at it all too long to objectively make sense of it.
Hi there

In order to use MANOVA your DVs must be correlated, If there arent you shouldnt use MANOVA but different ANOVAs' for each DV. As far i can understand your DVs arent correlate.....
"based on a combination of their critical thinking ability and GPA" you should use ANCOVA because your thinking ability and GPA is covariates (you want test something having those variables in your mind), So I will recommend you differents ANCOVAs'.