Regarding avwroi:

It seems that I specify the starting point and the length of the "roi" I want to extract, corrct?

Is the counting in volume number or time (based on the TR in the AVW header?

If the former, is the first volume 0 or 1?

Thanks again,

Michael

-----Original Message-----
From: Christian F. Beckmann [mailto:[log in to unmask]]
Sent: Thursday, February 26, 2004 5:23 PM
To: [log in to unmask]
Subject: Re: [FSL] Partial time series


No, data after the 'stop after' is not ignored but compared to a constant. If there is signal in the second half, this will not be modelled out and thus end up in the GLM residuals, inflating the standard error and lowering your stats. You can model signal during the entire length of the experiment using 2 different EVs, the first one for the first 2 cycles and a second EV for the second 2 cycles. This will, however, not give exactly the same results as when using avwroi to truncate the data as there is a possibly different 'quality' of fit during the first compared to the second half; but it might be good enough

ta
christian


On 26 Feb 2004, at 22:04, Michael Lipton wrote:

> I am interested in only analyzing the first half (2 of 4 cycles) of a
> block design dataset. Wil setting the "stop after" field in stats to
> the correct time accomplish this? The graphic
> of my model  shows and the FEAT web page states that all values in the
> model are set to
> zero after the "stop after" time. Does this mean that the data is
> being compared to a
> constant zero model or is the data beyond the "stop after" time being
> ignored?
>
> If not this approach, is there a way to have FEAT only look at the
> first part of my data without actually editing the analyze datasets?
>
> Thanks,
>
> Michael
>
--
  Christian F. Beckmann
  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