Dear Alfonso,

We are running an analysis in which we examine the connectivity of a set of seed regions to the whole brain at rest. Subsequently, we are interested in contrasting these seeds to observe connectivity patterns. However, I’m not clear about the precise details of how CONN generates our results and would appreciate some guidance.

I’d find it helpful to understand the structure of design matrices at first and second levels, contrasts applied at the first level, the files that pass from first level to second level, the second level tests, and finally the way in which contrasts are applied at the second level. 

For applying contrasts, there are two scenarios that I’ve tried to distinguish between that could be occurring in CONN. One possibility is that [1] [1] contrasts are applied for both seeds at the first level, and images from each contrast are then taken to the second level, after which [1 -1] contrasts are applied. 

Alternatively, perhaps [1 -1] contrasts were applied at the first level, and the resulting difference images are taken to second level, and a simple test is conducted at that level (perhaps a one sample t-test?). This appears to be more likely given that SPM.xX.X contains a single vector in our analysis.

I’d be grateful if you could provide some help in understanding these processes in CONN.  

Thank you in advance for your time and assistance. Your guidance will be really useful.

Best regards.