oops! the file should look like this (as described):

/NumWaves 4
/NumContrasts 2
/Matrix

0 1 0 0
0 -1 0 0

On 8 November 2013 12:56, Anderson M. Winkler <[log in to unmask]> wrote:

Hi Elisa,

If anxiety is the 2nd column, and it's the one you're interested in, the contrasts should be:

0 1 0 0
0 -1 0 0

You also need to change the /NumContrasts to 2, instead of 3. The /PPheights isn't used and you can delete this line. The file should then look like:

/NumWaves 4
/NumContrasts 2

/Matrix
1 0 0
-1 0 0

Alternatively, use the command Glm (or Glm_gui in a Mac) to generate these files with visual feedback and error checking.

All the best,

Anderson

On 7 November 2013 14:21, elisa veronese <[log in to unmask]> wrote:

--

Elisa Veronese - PhD

Research Unit on Brain Imaging and Neuropsychology (RUBIN)

Inter-University Center for Behavioral Neurosciences (ICBN)

University of Udine and University of Verona, Italy

ICBN Website: http://icbn.uniud.it/tiki-index.php

Dear FSL experts, 

 I am running a TBSS analysis on a group of healthy subjects. I have their FA images, along with their age, sex, and anxiety scale.

Age and sex are covariate of no interest, and I want to check if the anxiety scale has a correlation with FA.

My design.mat file is something like this

/NumWaves 4

/NumPoints 21

/PPheights 1 1

/Matrix

1 -16,1428571428571 2,74238095238096 -0,333333333333333

1 -13,1428571428571 1,99238095238096 -0,333333333333333

1 -11,1428571428571 3,84238095238096 -0,333333333333333

1 40,8571428571429 -1,40761904761904 -0,333333333333333

...

where the first column is the group (1=healthy subjects, they are 21), the second column is the demeaned anxiety scale, the third column is the demeaned age, and the fourth column is the demeaned sex. 

My problem is in the definition of NumWaves in the design.con file. Is it correct to write it like this? And what about the NumContrasts?

/NumWaves 4

/NumContrasts 3

/PPheights 1 1

/Matrix

1 0 0

-1 0 0

Thank you in advance,

elisa