Print

Print
Hi

On 11 Sep 2014, at 17:06, Paul Beach <[log in to unmask]> wrote:

FSL folks,

So the VBM wiki talks about making sure that one has equal parts controls and patients when doing early VBM steps. I understand the rationale for this in terms of making an unbiased template. 

However, are you supposed to always keep things like this (i.e. completely equal, number-wise) for the entirety of the VBM analysis?

No - later stages are fine - e.g. a two-group t-test is unbiased by different group sizes.

[ If *all* stages demanded equal group sizes, we wouldn't have had to add in the option of early stages only using subsections of the input data..... ]

Cheers.





It seems like you could add back the "extra" subjects in whatever group at step 3 and beyond. Is that correct, or are you completely limited by whatever group has the lowest 'N'?

Thanks
--
Paul Beach
DO/PhD candidate - Year VI
Michigan State University
- College of Osteopathic Medicine
- Neuroscience Program
 - MSU Cognitive and Geriatric Neurology Team (CoGeNT)


---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------

Stop the cultural destruction of Tibet