Hi Eva,
Yes - that looks fine. Presumably your C15 (C mean) should have been
[0 0 1 0]?
You could also have another contrast [1 1 1 1] to look for a
significant mean averaged across all groups.
In general f-tests look at the combined explanatory power of a subset
of specified contrasts. So an f-test could be used, for example, to
look for significance in any of the groups (like an omnibus test, or a
test for a main effect of group) by doing an f-test combining
contrasts C13,14,15,16.
Cheers, Mark.
----
Dr Mark Woolrich
EPSRC Advanced Research Fellow University Research Lecturer
Oxford University Centre for Functional MRI of the Brain (FMRIB),
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK.
Tel: (+44)1865-222782 Homepage: http://www.fmrib.ox.ac.uk/~woolrich
On 11 Aug 2009, at 12:20, Eva Kenny wrote:
> Hi,
>
> I’d like to run higher-level FEAT analysis on 4 subject groups. Is
> the model
> below correct for investigating differences between groups and group
> averages?
> Also, what would be the advantage of running F-tests also?
>
> Many thanks,
> Eva
>
>
> Number of EVs = 4
>
> Group EV1 EV2 EV3 EV4
> 1 1 0 0 0
> 1 1 0 0 0
> 2 0 1 0 0
> 2 0 1 0 0
> 3 0 0 1 0
> 3 0 0 1 0
> 4 0 0 0 1
> 4 0 0 0 1
>
> Contrasts = 16 F-test = 0
> Title EV1 EV2 EV3 EV4
> C1 A>B 1 -1 0 0
> C2 B>A -1 1 0 0
> C3 A>C 1 0 -1 0
> C4 C>A -1 0 1 0
> C5 A>D 1 0 0 -1
> C6 D>A -1 0 0 1
> C7 B>C 0 1 -1 0
> C8 C>B 0 -1 1 0
> C9 B>D 0 1 0 -1
> C10 D>B 0 -1 0 1
> C11 C>D 0 0 1 -1
> C12 D>C 0 0 -1 1
> C13 A Mean 1 0 0 0
> C14 B Mean 0 1 0 0
> C15 C Mean 1 0 1 0
> C16 D Mean 0 0 0 1
>
|