 |
 |
So no hints or experience with similar design from anybody? What could I expect if I use extreme higpass filter length (like 1200s) - would it have any sense and give chance for reasonable results? I would be grateful for any comments Marek Dear SPMers, I have a question related to fMRI paradigms with context manipulation and the block length. I know that most of people will tell that the blocks should be rather short (order of 30s), and I know that the highpass filter in the 1st level analysis should be four times of the block (+between block interval) in length. However I would like to repeat in the fMRI (in the mixed block-event related design) the study we did behaviorally – namely probabilistic reversal learning task (PRLT) with context manipulation. In our behavioral study we had 50 trials per block, and the block lasted about 5 minutes (stimuli were displayed for 2 sec (and it was the time to choose), then there was 1 sec feedback, and then randomized ITI). And we had two different kinds of blocks differing in context. Some blocks were rewarding (if you choose right, you get monetary reword, otherwise not), and the other were punishing (if you choose correctly you are not punished, otherwise you loose a fraction of money). And we had two groups and one of the groups seem to be better in the punishment condition. Can I (and how) reasonably transfer it to fMRI? To have the possibility to do some reversals of probability in the PRLT we need reasonable number of trials (let’s say the order of 50), so I rather do not see a possibility to substantially shorten the blocks. Or maybe there are some non-standard ways to analyze long-block-paradigms to have possibility to do between-block comparisons? Or maybe it is just impossible to reasonably repeat it in fMRI? I would really appreciate any hints, comments and suggestions. Best regards Marek