I'll try and give a little more information on my design.

In my original TBSS design, I was testing to see if FA (or MD) correlates with language ability on group of typically developing subjects. I had a continuous score from a  language test, and was controlling for gender, and socioeconomic status. Here is my design and contrast files for that. 

/NumWaves 3

/NumPoints 10

/PPheights 14 1 10 

/Matrix

12 0 6

10 1 0

8 1 3

9 1 6

8 0 0

8 1 10

11 1 4

7 0 2

14 1 7

13 0 6

/ContrastName1 Language

/NumWaves 3

/NumContrasts 2

/Matrix

+1 0 0

-1 0 0 

I now want to see if left or right hemisphere lateralization has any relationship with the language scores (again controlling for gender and SES). I ran TBSS_sym, so would I be able to use the same design matrix and contrast files and just rerun randomize with the the left_minus_right file? Would the first contrast give me left hemisphere lateralization, and second contrast give me right hemisphere lateralization?

Sorry for all the questions. I just want to be sure I am interpreting my results correctly.

Thanks for all your help,

Matt

On Sat, Oct 24, 2015 at 4:30 AM, Anderson M. Winkler <[log in to unmask]> wrote:

Hi Matt,

I don't understand this design, but yes, the only thing that needs to be changed is swap the sign. If the model is correct in one direction, it will be correct the other too.

Also, yes, works same way irrespective to the imaging modality (FA, MD, etc).

All the best,

Anderson

On 23 October 2015 at 18:35, Matthew Walton <[log in to unmask]> wrote:

Excellent thanks for your help, 

So if I want to compare leftward FA asymmetry to some measure of interest with a few controls would this work?

Measure | Control1|Control2|

+1                   0              0

And then this for rightward?

Measure | Control1|Control2|

-1                   0              0

And this would obviously be the same for any other DTI measure (MD,AD,RD)?

Thanks,

Matt

On Fri, Oct 23, 2015 at 1:19 AM, Anderson M. Winkler <[log in to unmask]> wrote:

Hi Matt,

There are various ways, and perhaps the simplest is simply swap the sign of the contrast so you don't have to actually make changes to the images. In other words, in the contrasts, where values are positive, replace for negative, and vice-versa.

All the best,

Anderson

On 22 October 2015 at 21:55, Matthew Walton <[log in to unmask]> wrote:

Hello,

I've run through the basic TBSS pipeline, as well as tbss_sym. I'm just a little confused as to how to go about changing symmetry results to test for R>L (or rightward asymmetry if you will). 

The user guide says to "invert the data and rerun randomize".

What data is this referring to? Is there any other steps involved?

Thanks for your help,

Matt

On Thu, Oct 22, 2015 at 12:27 PM, Matthew Walton <[log in to unmask]> wrote:

Hello,

I've run through the basic TBSS pipeline, as well as tbss_sym. I'm just a little confused as to how to go about changing symmetry results to test for R>L (or rightward asymmetry if you will). 

The user guide says to "invert the data and rerun randomize".

What data is this referring to? Is there any other steps involved?

Thanks for your help,

Matt