MANOVA vs multiple ANOVAs

#1
Hello everyone,

For starters, I just found this site and am wondering where it has been all my life, or all my PhD, they feel like the same thing sometimes.

Anyway, here is the situation. I have a study where I am looking at the effect of one independant variable on about 30 dependant variables. Due to my own ignorance, I evaluated these traits using multiple ANOVAs instead of a MANOVA. I am in pretty far, I have already submitted my thesis to the committee and one of them mentioned this. However, they didnt say to redo it, rather just be prepared to denend my choices at the defense.

Now, after they told me this I ran the data through a MANOVA and the overall model was significant. This being said, am I justified using the individual ANOVAs if the MANOVA is significant?

I guess the real question is, how screwed am I? Any suggestions?

One final thing though, if MANOVA reduces the p-value of each comparison, wouldnt this weaken the sensitivity of the comparis? For example, if one of my ANOVAs was significant, which could have serious real world ramifications, couldnt it be masked by using the MANOVA?

Thanks for any input.
 

WeeG

TS Contributor
#2
usually if you run a MANOVA and it's being significant, than the next phase IS to run multiple ANOVA's. So if in your case the MANOVA was significant, you can simply say that you ran it but didn't put the info about it in your work. You could have been screwed up hard, but maybe in this case you can get away with it....