Hi Alessandra,
Keep in mind that the BOLD response should last at least 12 seconds.
If your stimulus is 1.65 seconds, maybe you should allow for at least
13 seconds. The TENT functions should model the entire duration, not
just the peak response period.
Do you want the TENTs to be TR-locked? If so, TENT(0,13.2,12) would
be 11 intervals of 1.2 seconds.
If you want the TENTs to come every other TR, maybe TENT(0,14.4,7)
would work out, as 6 intervals of 2.4 seconds.
---
Regarding your script, polort 1 is probably too small for a 400 second
run. Cubic is recommended for that duration. It might also be good
to include motion parameters as regressors of no interest.
- rick