On 5 December 2011 17:56, Mathias Warnke
<[log in to unmask]> wrote:
Hi Experts.
I'm working on a new fMRI paradigm, and i would really appreciate any help concerned this topic.
We are trying to assess impulsiveness in risk/non-risk subjects. We have designed a 'play cards' paradigm were after a first 'cue' card presentation the subjects have determine whether the next card (let's say feedback card) is higher or lower than the first one by pressing a button. We have set the same proportion between certain cue cards (2,3,4,8,9 or 10) and uncertain cue cards (5,6, or7) in a random presentation using an event related paradigm.
I have separated the analysis of 'cue' cards and the 'feedback cards' using two EV's, setting the onsets and durations of the cards presentation. The weighting of the input was set to 1 (under the 3 column format).
-My first question is to know whether it's ok to give the same weighting (1) to every cue card presentation even when they have different meaning (certain/ uncertain).
There's no obvious way to decide on different weightings for different task types. You need to decide if you think they will be similar enough to model them as one type of trial, or analyse them separately.
-How can I determine whether the efficency of my paradigm is ok?
Fill in the design in FEAT and have a look the required effect sizes for different paradigms.
-Is it ok to apply a unpaired t test to make comparissons between both groups of subjects?
yes
Cheers,
Eugene
Thanks in advance.
Best.
