History of AFNI updates  

|
December 19, 2008 07:02PM
Hello AFNIers,
I have a quick question about setting up contrasts using the -aBcontr function in 3dAVOVA3. I have a 3x3 ANOVA design, and I would like to test for the effects of the first variable at the first level of the second.

This would be a contrast of:
1 0 -1
1 -2 1

or some version of that.

I can set up two separate -aBcontr lines:
-aBcontr 1 0 -1 : 1 con1 \
-aBcontr 1 -2 1: 1 con2 \

but that is not necessarily what I want. I would like to estimate each simultaneously. Is there a way of setting up the contrast [1 0 -1; 1 -2 1] : 1?

This will also be useful for tricking AFNI into larger designs (such as a 5 way ANOVA).

Best,
Wil
Subject Author Posted

3dANOVA3 -aBcontr w/ 3 levels

Wil Cunningham December 19, 2008 07:02PM

Re: 3dANOVA3 -aBcontr w/ 3 levels

Gang Chen December 20, 2008 02:08PM

Re: 3dANOVA3 -aBcontr w/ 3 levels

Wil Cunningham December 20, 2008 02:26PM

Re: 3dANOVA3 -aBcontr w/ 3 levels

Wil Cunningham December 20, 2008 02:31PM

Re: 3dANOVA3 -aBcontr w/ 3 levels

Gang Chen December 20, 2008 04:27PM

Re: 3dANOVA3 -aBcontr w/ 3 levels

Wil Cunningham December 20, 2008 04:32PM