Hi Conny
unlike the previous case (where there was no rest period and only A,B)
the regressors in this case do not seem to be linearly dependent -
unless the other 2 EVs that were missing in your example do actually
'cover' all the remaining rest periods.
If so, you can still create 4 regressors (instead of five) and create
contrasts based on those - which ones to model as individual EVs will
all depend on the context though...
ta
Christian
Conny Schmidt wrote:
> Hi Christian,
>
> thanks a lot for your mail !!!
>
> The first problem is, that I wasn't the one who designed the paradigm, I just
> have to analyze the data.... I'm also somehow sceptic whether you can get
> some useful information out of that. Especially as the paradigm that was used
> was even more complicated, in the sense that more conditions had been put
> into (in total 5!) I just wanted to make it simpler an described the main
> features in my last mail.
>
> Your suggestion to use 3 EV is my problem, as those would be linear dependent
> (because they are like time-shifted copies) and the analysis wouldn't
> work...... ( I mean, according to the paradigm I would have to use for
>
> condition A = EV1 on: 10 off: 90 phase: 90
> condition B = EV2 on: 10 off: 90 phase: 40
> condition C = EV3 on: 10 off: 90 phase: 70 )
>
> Probably there is a way of an other description which I cannot see .........
>
> Thanks a lot again for your help!!!!
> Conny
>
>
--
Christian F. Beckmann
Address: Oxford University Centre for Functional
Magnetic Resonance Imaging of the Brain,
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
Email: [log in to unmask] http://www.fmrib.ox.ac.uk/~beckmann/
Phone: +44(0)1865 222782 Fax: +44(0)1865 222717 Mob: +44(0)7980 691852
|