Dear Christian,
>I am trying to get a visual activation in an event-related design using
>flashes as stimuli. The participants have to fixate a cross on a screen. The
>flashes are presented for 3 seconds and one volume is recorded 6 seconds
>after stimulus onset. In the rest condition, the participants just have to
>fixate the cross.
Hmm. That sounds a bit weird. Are you using fMRI? If so, I wonder
why you are only sampling brain activity once per event. You must
need to have a vast number of repetitions to get any power, as this
is a very inefficient design. Presumably your events have to be very
sparse, to allow the BOLD signal to recover (which takes 30 sec or
more) between each stimulus.
>For statistics I use the t-Test or ANOVA in Basic Models.
This simple approach won't, of course, take account of any confounds
(such as signal drift during the experiment).
>While I have no problem showing a V1-activation in a block-design, this
>event-related procedure doesn't lead to satisfying results. Is there a
>mistake in the experimental design or did I do something wrong with the
>statistics?
Neither is wrong in principle, but you have reduced your chance of
getting any significant results by
1. only sampling brain activity once after each event;
2. not modelling statistical confounds (as you can with a GLM approach);
3. presumably spacing your events very sparsely (whereas the trend in
our department is to pack as many events in as possible).
If I was going to use fMRI to try to do this, I would use a more
efficient design. The approach which is generally used in our
department is explained in:
Josephs et al. (1997). Event-related fMRI Human Brain Mapping 5: 243-248.
Friston et al. (1998). Event-related fMRI: characterizing
differential responses. NeuroImage 7, 30-40.
Then there are some other papers by Rik Henson and various others
which deal with the issue of design efficiency etc. You'll find some
relevant references on Rik's web site at
http://www.fil.ion.ucl.ac.uk/~rhenson/refs.html.
>Thanks in advance,
No problem. Best of luck.
Richard.
--
from: Dr Richard Perry,
Clinical Lecturer, Wellcome Department of Cognitive Neurology,
Institute of Neurology, Darwin Building, University College London,
Gower Street, London WC1E 6BT.
Tel: 0207 679 2187; e mail: [log in to unmask]
|