It just depends on how you do the group mean modelling. For example,
if EVs 1 & 2 encode group membership with 0s and 1s (and are therefore
not zero-mean), and EVs 3 & 4 are zero-mean behavioural covariates for
the two groups, then you would not demean the data, and you would have:
[1 -1 0 0] = difference between two groups' means
[0 0 1 -1] = difference between two groups' correlation with behaviour
Cheers.
On 4 Mar 2008, at 19:36, Suzanna Laycock wrote:
> Thanks Steve,
>
> I think I'm even more confused. I got the impression if you were
> interested the correlation of a behavioural score (eg. reaction
> time) with FA between groups, that you demeaned the data and used
> the -D flag. Is this still the case?
>
> Specifically for my case, I have about 10 behavioural tests to
> investigate for two groups, so I planned to look individually.
>
> Thanks for clarifying this
> Suzanna
>
> On 4 Mar 2008, at 18:01, Steve Smith wrote:
>
>> Hi,
>>
>> This depends on what's in your model. If you have a simple unpaired
>> t-test with one EV for each group then definitely do not demean the
>> model or the data....does that cover your scenario?
>>
>> Cheers.
>>
>>
>> On 3 Mar 2008, at 15:05, Suzanna Laycock wrote:
>>
>>> HI,
>>>
>>> Sorry for this simple question. I've been reading today's threads
>>> about randomise and I was wondering if you have two or more
>>> groups, do you demean each group separately or calculate the
>>> average for all the groups?
>>>
>>> Thanks
>>> Suzanna
>>>
>>
>>
>> ---------------------------------------------------------------------------
>> 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
>> ---------------------------------------------------------------------------
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|