DPABI ANCOVA repeated mesures

Submitted by aguinea on

Dear DPABI list,

I'm currently done a repeated measures ANCOVA model with three conditions ("groups"). I want to compare two by two, such as 1 -1, -1 1, 1 0 -1, -1 0 1, etc...
 
Attending how the program produces the predictors and the contrasts (I can see that the program produces two columns for three conditions? (being the last one coded as 0 in the last lines), I'm not sure how should I change the y_GroupAnalysis_Image.m file in order to establish all the possible two by two comparions. How should I compare i.e., the condition 1 vs 3,?
 
Any idea?
 
Thanks in advance.
 
The best
 
andres

YAN Chao-Gan

Wed, 01/14/2015 - 16:40

Hi Andres,

A simpler answer it that I will do paried t-tests between each pair of conditions.

If you want to use y_GroupAnalysis_Image to do that in the ANCOVA mode, for each contrast, I will suggest you to:

set a breakpiont at Line 95 of y_ANCOVA1_Repeated_Image.m

[b_OLS_brain, t_OLS_brain, ANCOVA_F, r_OLS_brain, Header] = y_GroupAnalysis_Image(DependentVolume,Regressors,OutputName,MaskFile,CovariateVolume,Contrast,'F',0,Header); 

And when the program stops here, you can check the Regressors, and then modify the Contrast to the two columns you wanted, and change 'F' to 'T'.

Best,

Chao-Gan