Hi Chloe,
I always found it helpful to use the PET models to do more complex 2nd
level models. For example, you could use 'multi-group: Conditions &
covariates'. This should give you what you want. The only thing is that
you have to answer 'no' to most of the questions about global
normalization, masking etc. To switch off threshold masking, you should
input '-inf'.
Stefan
Chloe Bextor wrote:
> Dear SPMers
>
> How can I perform a mixed ANOVA analysis in SPM at the RFX level? I have
> one BS factor with 2 levels (patients vs controls) and one WS factor
> with 3
> levels (A vs B vs C).
>
> I know how to examine the task-by-group interaction if I only have 2
> levels
> of my WS factor (take one con*.img corresponding to A vs B per subject to
> the 2nd level and perform a two sample t test). How do I do this if I
> have
> 3 levels to my WS factor? So far I have only been able to conduct a
> one-way
> ANOVA with 3 levels within SPM2 ...
>
> Any help is greatly appreciated.
>
> Thanks.
>
> Chloe
>
> _________________________________________________________________
> Find a cheaper internet access deal - choose one to suit you.
> http://www.msn.co.uk/internetaccess
>
>
--
Dr. Stefan Kiebel
Wellcome Dept of Imaging Neuroscience
Institute of Neurology, UCL
12 Queen Square
London WC1N 3BG
Phone: (+44) 20 7833 7478
Fax: (+44) 20 7813 1420
|