gltCode 3dMVM

AFNI Message Board
1

I am getting output from t tests I specify in 3dMVM that I expect to be different from each other, but they are not, and I could use some help understanding why. There are a few variables in my model, but right now I would like to get a t test result for the within subject variable Drug for each sex separately. With the following 3dMVM script, the output from each glt line that I think should give me those separate tests (glt 2 and 3) yields datasets that are identical. I am not sure what i am doing wrong. Help?

3dMVM_rcc -prefix macs.sex -jobs 64
-bsVars “Sex+Paired_BG”
-wsVars "TimeDrug"
-qVars ‘Age’
-num_glt 3
-gltLabel 1 ‘Sex’ -gltCode 1 'Sex : 1G -1B’
-gltLabel 2 ‘MaleXPairedBG’ -gltCode 2 'Paired_BG : 1O -1F Sex : 1B ’
-gltLabel 3 ‘FemaleXPairedBG’ -gltCode 3 ‘Paired_BG : 1O -1F Sex : 1*G ’
-dataTable
Subj Age Sex Paired_BG Drug Time Like_MA InputFile
MACS002 19 G O MA A 83.5 stats.S002+tlrc’[67]’
MACS002 19 G O MA B 83.5 stats.S002+tlrc’[70]’
MACS002 19 G O PB A 83.5 stats.S002+tlrc’[73]’
MACS002 19 G O PB B 83.5 stats.S002+tlrc’[76]’
MACS004 23 G F MA A 27 stats.S004+tlrc’[67]’
MACS004 23 G F MA B 27 stats.S004+tlrc’[70]’
MACS004 23 G F PB A 27 stats.S004+tlrc’[73]’
MACS004 23 G F PB B 27 stats.S004+tlrc’[76]’
MACS005 20 B O MA A 85.5 stats.S005+tlrc’[67]’
MACS005 20 B O MA B 85.5 stats.S005+tlrc’[70]’
MACS005 20 B O PB A 85.5 stats.S005+tlrc’[73]’
MACS005 20 B O PB B 85.5 stats.S005+tlrc’[76]’
…

2

The reason you got the same results for post hoc tests #2 and #3 is that you assumed there is no interaction between Paired_BG and Sex in your model specification:

-bsVars “Sex+Paired_BG” \

It sounds like you want

-bsVars “Sex*Paired_BG” \

In addition, you labeled “Age” as a quantitative variable:

-qVars ‘Age’ \

However, you didn’t include ‘Age’ in your model. Maybe you want something like

-bsVars “Sex*Paired_BG+Age” \

Also check out this: https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/STATISTICS/center.html