Hi Manish
I did something similar - simply use the "three-column-textfile" approach
delineated in the online help.
Briefly put, for each regressor of interest / explanatory variable (EV),
create a text file with three columns:
1) onset time in seconds
2) duration in seconds (mustn't be zero)
3) "weighting factor" (be careful with this one - this is for specific
design demands.)
Then, in FEAT, simply indicate which *.txt-file to use for each
explanatory variable (EV) and off you go...FEAT will tel you if you
specified to wrong data and if there are serious correlations between your
EV's...but for more details go and see:
http://www.fmrib.ox.ac.uk/fsl/feat5/detail.html#stats
Hope that helped a bit!
Cheers
Cornelius
On Mon, 03 Jan 2005 08:54:37 -0800, Manish Dalwani
<[log in to unmask]> wrote:
> Hi FSL group,
>
> I have a design (userinteractive) which gives me
> blocks of variable size/durations. Is it possible to
> analyze such a data? If yes, can anyone please suggest
> how to analyze such a data?
>
> Thanks,
> Manish Dalwani
>
>
>
>
> __________________________________
> Do you Yahoo!?
> The all-new My Yahoo! - Get yours free!
> http://my.yahoo.com
>
--
Cornelius Werner
Institut fuer Medizin (IME)
AG Kognitive Neurologie
Forschungszentrum Juelich
52425 Juelich
Germany
Tel. +49-(0)2461-61-8609
|