Hi All,
Sorry to clog up the list but I had a few more questions re: setting up my
design in FEAT.
We're using a simple box-car design for a stroop like task with 16 blocks
of alternating neutral (N) and interference (I) conditions.
I'm interested in activation associated with the interference effect
across the entire task (I>N), as well as how this effect varies with time
spent performing the task (ie., what regions show more activation for the
I>N effect in the first half of the task vs the second, and vice versa.
Not what regions are more active in one half vs the other collpased across
condition).
I've tried to model the 1st half-2nd half differences by using 2 EVs; one
that switches off after the first half (1st 8 blocks only), and another
that skips the first half (last 8 blocks only). However, this precludes
including an EV for the overall effect since it represents a linear
combination of the first 2. Is it valid to sum the two half-task EVs to
obtain activation maps associated with performance across the whole task
(i.e, a 1, 1 contrast)?
Assuming this is OK and I have only these 2 EVs, I assume the following
contrasts would represent (take EV1 as the first-half regressor & EV2 as
the second-half):
EV1 EV2
1 0 I>N for first half only
0 1 I>N for second half only
1 1 I>N for the entire task
Then, if I want to directly compare 1st half to second half (or vice
versa) with 1 -1, would this represent comparing activation collapsed
across I and N conditions (simply, what is greater in the first vs the
second half, irrespective of interference-neutral), or would it be the
difference in the I>N effect across the first and second half (ie., what
regions are more active for the I>N contrast in the first half vs the
second)? I'm only interested in the second question. I think that the
above is not the right way to go about it and that I may need to add an
interaction term to produce the relevant map, but am unsure of how to do
this.
Finally, is the choice of HRF convolution merely one of preference?
Are their optimal convolutions for particular designs?
Does it matter in a boxcar design?
Many many many many thanks,
Alex
|