Hi,
On 21 May 2007, at 23:54, Vishwadeep Ahluwalia wrote:
> Hi Steve,
> Thanks for replying! A few more questions on the same topic..
> 1. From your suggestion,for the second level analysis, i entered the 4
> conditions as cope images.
> group EV1 EV2 EV3 EV4
> input1(A) 1 1 0 0 0
> input2(B) 1 0 1 0 0
> input3(C) 1 0 0 1 0
> input4(D) 1 0 0 0 1
>
> contrast
> Mean-D 0.33 0.33 0.33 -1
>
> Is this the correct way to set the contrast?
Yes, and use the FE (fixed effects) option for this 2nd level analysis.
> 2. Can we setup a 2nd level model that includes all subjects and
> then do a
> group mean in a 3rd level analysis(a bit like the multisession
> multisubject
> example on the FEAT webpage)
> (See attached design.gif)where
> input1 .33 0 0 .... 0
> input2 .33 0 0 .... 0
> input3 .33 0 0 .... 0
> input4 -1 0 0 .... 0 for each subject
I would keep it simple, just do a different 2nd level anlaysis for
each subject to create the contrast of interest, then do a ME
analysis across subjects at 3rd level. There's no advantage in
combining the 2nd level across subjects when doing FE.
> 3. Instead of t-tests i can use ANOVA to find if any condition has a
> different mean. But how should i specify the contrast to find
> whether D is
> more than or less than the mean(ABC),especially in a repeated
> measures design?
I would avoid ANOVA if possible and explicitly ask the questions you
are interested in. If you want to check A, B, C and D separately from
eachother just pass each up into a separate cross-subject analysis.
Cheers.
> Thanks!
> -vish
------------------------------------------------------------------------
---
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
------------------------------------------------------------------------
---
|