This is just an opinion rather than evidence-based dogma and I'd be
interested in what others thought of it.

With 2 rare patients, clearly one is going to want to separate what is
unique to each case (noise) from what is common to both (where the
latter is presumably what you are interested in).

Obviously that is going to be difficult in a 2 populaion comparison
where one 'population' is n=2. My opinion is that the most trasnparent
way of expressing the findings would be to compare each individual to
the control population separately. Then discuss the results along the
lines of "each patient had significant alterations in regions x and y
compared to controls etc."

There is certainly precedence for analysing individuals in the
literature (though it depends a bit on what kind of imaging you are
doing).

Peter Nestor

On 18 Apr 2007, at 09:48, Christian Keysers wrote:

> Dear SPM mailing list,
>
> we have conducted a study comparing 2 rare patients against 16
> controls. We have the same contrast (lets call it C) in all 18
> subjects at the first level. At the second level I performed a
> 2-sample t-test (voxelwise) with 2 subjects in group 1 and 16 subjects
> in group 2. One of the reviewers states:
>
> Regarding the 2 sample t-test.  I wonder if SPM2 is able to compare 2
> subjects versus 16.  I believe this is not the case and that one must
> use a different methodology to make this comparison.  I would like to
> see a citation to a study that uses a similar approach that might
> alleviate this concern.
>
>
> Is anyone aware of other studies using a similar approach, or a
> comment from the makers of SPM regarding this issue?
>
> ck
> --
>
> * *
>
> *Prof Christian Keysers, PhD
> *Scientific director neuroimaging center
> Professor for the neural basis of empathy
> http://www.bcn-nic.nl/index_research_empathy.html
>
> University Medical Center Groningen, University of Groningen
> A. Deusinglaan 2,  9713AW Groningen, The Netherlands
>
>
> Phone: +31 50 3638794, Fax: +31 50 3638875
>
>