Dear all,

       I am interested in running a second level analysis with a covariate. I saw some examples in the list but mostly was using one sample t-test.

       I am doing a resting state protocol before and after drug ingestion. For each subject, I have two first levels (before and after) in which I run a connectivity analysis with a VOI time course as a regressor. Then, I compare the connectivity before and after the drug just in a second level paired t-test ( [zeros(1, nsubjects) 1 -1], for the before>after contrast and [ zeros(1, nsubjects) -1 1], for after>before). 

       Now, I would like to check if the connectivity changes are correlated with changes in the psychological evaluations we also did before and after the drug ingestion. I thought I could simply add my psychological delta values (after-before) as a covariate in the second level analysis. However, I am not sure if I could do this and if yes, how would be the contrast vector to see, for example, if the increased connectivity (after>before) in some regions is correlated with my covariate, as I have one more column in my design matrix.

       I believe I could run a second level with one sample t-test if I had the contrasts before>after and after>before already in the first level. However, I would need to run all my first levels again with two sessions for each subject in order to have these contrasts. I would like to know if there is an easiest way to do this.

      Thank you in advance for any help,

      Best regards, Kátia.