Which ANOVA to use - sum of squares,model,n-way,one-way?

#1
Hi all,

I want to analyse some data using an Anova. The data are one continuous variable (the thing I was measuring) and one categorical variable (the 4 different group numbers), I also have 3 more continuous data which I want to test for effects on my dependent variable.

I really just want to run an Anova on my data and my grouping variable, with 3 more anova's looking at the other variables, I'm not interested in any interactions between them.

I've been running an N-way anova but I'm not convinced I know what I'm doing. If I run a one-way anova on my data and interesting grouping variable I get different values to those I get when I run an N-way anova on all the groups at once.

Is that normal? What is the n-way anova doing that is different? As far as I know it shouldn't be adjusting anything (that's what an Ancova is for). At this point I'm unsure whether I'm using the correct model or sum of squares for what I want to achieve...

Any help would be greatly appreciated. I have tried to work this out but I'm not a statistician and unfortunately I don't know any.

Roddy.

EDIT: I should add that this is not some kind of homework, I'm a neuroscience PhD student and I'm doing analyses in Matlab in case that changes things.
 
Last edited: