Hello,
I'm designing a new experiment, and I'd like to make sure that what I'm
doing makes any sense. I hope someone can comment (or, preferably, reassure
me..).
It's an AX-CPT task (similar to GonoGo). The task consists of 4 trial
types: 1 target (70% of trials), 3 types of non-targets (30% of trials). I
would like to use a mixed design (only-targets vs. mixed
targets/nontargets). For the mixed blocks, I was thinking of presenting the
stimuli in a "rapid event-related" way with randomized order of
target/nontargets and a fixed ITI of 2.5 seconds (each trial consisting of:
Prime - 1s SOA - Target - 1s SOA) .
I'm mainly interested in the differential effect between nontargets>target,
not the common effect of stimuli>baseline or the shape of the HDR. Does
this design make sense then?
And would it be a good idea to randomize the ITI or optimize it by the
inclusion of null events (using eg Optseq)? So far I decided against this,
because a fixed ITI seems to be more behaviorally challenging (eg harder to
inhibit a response) and I thought null events don't really contribute to
the detection of differential effects. Please tell me if I'm wrong..
I would greatly appreciate any help!
Elseline
|