panda panda wrote:
> Hello,
>
> We have a 4 condition block-design experiment, in which we would like to
> look at several contrasts. Our questions are as follows:
> 1. Which is the best contrsat in order to see a difference between two
> conditions, e.g., the 1st and 2nd:
> t contrast: 1-100
> or F contrast: 1000;0-100 (Is that a valid contrast for testing a
> directional difference?)
>
> We prefer using F contrasts as it is less restricting than t, however we
> wonder whether it can be used for comparison of 2 conditions only.
>
(a) The t-contrast [1 -1 0 0] looks for voxels where the response is significantly *larger*
on condition 1 than condition 2.
(b) The F-contrast [1 -1 0 0] looks for voxels where the response is significantly
*larger or smaller* in cond 1 than cond 2. This is not directional.
(c) The F-contrast [1 0 0 0; 0 -1 0 0], or equivalently [1 0 0 0; 0 1 0 0], looks
for voxels where either conditions 1, or 2, or a linear combination thereof,
are significantly zero. It does'nt look for differences between conditions.
And its not directional.
If you just want to compare 2 conditions (irrespective of directionality) then
(b) would be adequate.
> 2. Our experiment is based on a parametric hypothesis, according to which
> cond1<cond2<cond3<cond4. What would be the best way to model this
> hypothesis
> in a contrast?
>
It may be best to model this parametrically, looking for linear and nonlinear
increases in response with condition.
Best, Will.
> Thank you for a rapid response,
> Einat and Dafna
>
> _________________________________________________________________
> Express yourself instantly with MSN Messenger! Download today - it's FREE!
> http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
>
>
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7475
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|