Multivariate ANOVA?

My proposal has some merits

  • Yes

    Votes: 0 0.0%
  • No

    Votes: 0 0.0%
  • I can help - let me respond to your post.

    Votes: 0 0.0%

  • Total voters
Suppose I have 7 outcomes of interest (all population means) and I have 3 groups (control, group 2, group 3). I am interested in comparing the means for group 2 and group 3 against the control, but not against each other. There are no other covariates of interest at this time (just the 3 groups). I feel like I can do a MANOVA to investigate the outcomes of interest across the groups. But, I don't know what assumptions I am making because I've never had an opportunity to use/study manova. Essentially, what was proposed was 7 x 3 independent one sample t-tests. But, there are several problems with this. Namely, there are no direct comparisons between the groups and control. And the family-wise error rate of 7 x 3 tests is HUGE. Can someone please share some insight. P.S. I completed my Masters degree and I'm working in industry.