Thanks for your helpful comment, Rick

I have a very basic question about the structure of the stimulus timing file (in -stim_times): where to start counting. My experiment is a rapid event-related, with 8 blocks. There are 9 conditions randomized in each block, presented 4 trilas/condition/block. However, we are interested in analyzing only correct trials. Each block is 180TRs long, and TR = 2 sec.

My stimulus timing file looks like (in .txt format, only correct trials)
39 149
7 79 241 309
79 109 131 249
237 271 297 335
15 265
27 67 245 273
71 89 291
265

where each row represents one block, and within that, I have timings of stimulus onset in sec. e.g.39 (1st block, 1st trial) means the onset of the stimulus at 39th sec (onset of the 1st TR is "1"). Is this the right way to count, or does this have to be in TR? I'm not cropping first or last data points.

Thank you very much for your help! Akiko