Dear Alfonso,

Thanks for your quick reply!

I do extract correlations for each individual, since my "intrusion" (ptsd symptom) variable is highly skewed, and I'm running some bayesian analyses. Thus in the Second-level Results tab I
select 'All Subjects' [1]
select rest1 and rest2 in the 'conditions' list, and enter a [-1 1] between-conditions contrast.
Then I import the values to R, and perform my bayesian regression analyses there.

My main question however concerns directionality, more precisely:

Suppose my delta (change-score from rest1 to rest2) resting-state connectivity values (x) predict more intrusions in individuals (y), thus resulting in a positive regression coefficient. However, what does this positive regression coefficient really mean?

1) individuals with a higher increase of connectivity from rest1 to rest2 show more intrusions
but is this "increase of connectivity" positive or negative? I.e., do individuals show
a) higher positive correlation between target and seed region in rest2 as compared to rest1, or
b) higher anticorrelation between target and seed region in rest2 as compared to rest1?

Intuitively I would say
a) - if the connectivity change score from rest1 to rest2 is positive, then individuals increased in positive connectivity between seed and target region: According to a regression logic, more x (higher positive change scores) result in more y (more intrusions)?

Yet, I am unsure whether my assumptions about directionality (positive vs. negative correlations) are correct?

Thank you a lot in advance!!
Best,
Laila