I am preparing a simple food cue reactivity task for BOLD using an event related design. At the moment, the programme presents images from three conditions (50 in each) in a pseudo-random sequence, each image for 3 seconds followed by a fixation cross for a mean ISI of 1 second to provide the jitter. The sequence has been prepared such that all conditions precede and follow each other equally.
But, after reading some messages on here and talking to others, should I have a longer mean ISI, or perhaps include null trials instead (rather than a jittered fixation cross after every image)? I would like to optimise the sequence to be able to detect and discriminate between the events in three conditions (contrasting each condition with the others).
I would be grateful for advice regarding the optimal ISI in a design such as this.