Hakwan Lau wrote:
> Hey Will,
>
> Just a quick question about PPM (in SPM2), with which I have started
> playing only recently. When I try to view a PPM with a certain
> contrast, it asks for a threshold of the effect size. How does SPM
> generate the default value for this? It changes for different analyses,
> and it seems to me that it is somewhat based on the variability over
> voxels, which is used as the prior?
>
That's correct - the default value is based on the estimate of the
prior variance.
> In general, do you have any problem with using 0 as the threshold?
> Afterall, this is what most of us frequently do with classical stats,
> and typically we justify ourselves by saying that we're doing a basic
> science (contra clinical) study so effects of any size, if significant,
> are of interests.
No - 0 would be fine. But the hope is that you use an effect size that
is meaningful to the type of effect you are studying. For example,
evoked sensory responses should be in the range of 1-2% (of global mean
value) whereas subtle cognitive effects are more likely to be in the
region of 0.5% or lower.
Best,
Will.
>
> Hakwan
>
> PS - you can forward your answer to the mailing list if you like
>
>
>
--
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/
|