Print

Print
Hi Joy,


From what I understand, the number of columns in your design.fts file must match the number of rows in your design.con file and placing a 1 in a row of your fts file is equivalent to clicking the f-test boxes in the contrast GUI for that particular f-test. Is my set up correct?


Yes, that's correct.

 
Also, is this f-test equivalent to doing an ANCOVA to get the main effect of your groups on the dependent variable?

There are multiple different F-tests that you could construct, and if you'd like to compare the 3 disease groups with no regard for the controls, then I think the following would be more appropriate (using your current design.con file):

design.fts
0 0 0 1 1 0 0 0 0 0 0 0 0

If you want to include the controls, then I think you'd want something as this:

design.fts
1 1 1 0 0 0 0 0 0 0 0 0 0

That said, you need to modify your design: remove the intercept, because it's a linear combination of the 4 groups, then redefine the t-contrasts and then define the f-tests. Also, although for your current contrasts it won't make a difference, you may want to mean center sex, just as you did for age and education. If later you decide to create a contrast to look at the mean of individual groups, you'll already have the design properly set up.

And yes, look at different groups/levels with an F-test in the presence of continuous nuisance variables is equivalent to an ANCOVA.

Hope this helps!

All the best,

Anderson