Hi,

I have a question regarding the output of a t-test on 1 group with 3 regressors, contrast [0 1 0], method NBS. In the output I get the NBS.test_stat variable with an 'N x N array of test-statistic values for each edge' but these values do not seem to be thresholded. I get a (presumably t-) value for every point in my 100 by 100 matrix (except for the diagonal) ranging from -7 to 8 including values close to 0. I only want to show in my matrix the thresholded results from the NBS permutation testing, so only the 829 significant edges. How can I find that threshold to apply to the output matrix?

Many thanks,
Selma

Hi Selma, 


nbs.NBS.con_mat{1} will give you the matrix of significant connections. 

So to threshold the t-statistic values so that only the significant connections remain non-zero, consider using:

nbs.NBS.con_mat{1}.*nbs.NBS.test_stat

I hope that helps, 
Andrew


Originally posted by Selma Lugtmeijer:

Thanks a lot!