>> I am now trying to interpret the output from these analyses and am not sure whether to look at:
>>
>> #1 (Intercept) F

This is the intercept in the model, which is of no interest to most people.

>> #3 Session: Anxiety F

Interaction between Session and Anxiety

>> #8 T1-Anxiety
>> #9 T1-Anxiety Z
>> #10 Session_Anxiety
>> #11 Session_Anxiety Z

You specified these tests through -gltCode, so you should know what you meant to get.

>> Our primary question is how connectivity from T1 to T2 is associated with change in anxiety from T1 to T2.

Is the following what you want:

-gltLabel 4 'Session_Anxiety' -gltCode 4 'Session : -1*T1 1*T2 Anxiety :' \

>> I had entered the mean value for qVarCenters, but since there are important differences between the
>> mean of Anxiety at T1 and T2, should I be entering two separate means here? Or not have the qVarCenters option?

I recommend that you center Anxiety at T1 and T2 separately, and then feed the numbers into the table for 3dLME.

>> in regards to thresholding, what would you recommend as a threshold?

Personally I don't like thresholding because of its arbitrariness. For publication purpose, you can use p-value of 0.01 or below.

Gang