Thank you for your quick and very illuminating help!
Regarding the problem:
I understand the zero-mean is very helpful when one is interested in a parametric modulation of overall equally distributed stimuli effects with overall equally distributed parameters.
Would it be advisable in this instance to switch off the zero mean (or would it create any unwanted sideffects)?
If so, does any one know how to do it in the proper routine?
Thanks a lot!
Nico
-------- Original-Nachricht --------
> Datum: Wed, 17 Nov 2010 11:02:47 -0500
> Von: "MCLAREN, Donald" <[log in to unmask]>
> An: [log in to unmask]
> Betreff: Re: [SPM] Parametric Modulation compared SPM8
> Model A: stimulus A and stimulus B are allowed to have different effects,
> stimulusA*rating and stimulusB*rating are allowed to have different
> effects.
>
> Model B: stimulusAB have the same effect, and stimulusAB*rating.
>
> Difference 1: If A and B are different, the error term increases; the
> statistics will decrease even though the con values for the average are
> the
> same. Same is true for the modulation.
>
> Difference 2: The ratings are zero meaned, so if there is any difference
> in
> the mean ratings between stimuli; the stimulus*rating will be different.
> For
> example, A A B B A B and the ratings were 1 1 4 4 7 1, then you get -2 -2
> 1
> 1 4 -2 and one column versus 2 would be the same IF A and B had the same
> response AND same parametric effect. If the ratings were 2 2 3 2 8 1, then
> you'd get -1 -1 0 -1 5 -2 OR -2 -2 0 0 4 0 and 0 0 1 0 0 -1 as the two
> modulation strings.
>
> Best Regards, Donald McLaren
> =================
> D.G. McLaren, Ph.D.
> Postdoctoral Research Fellow, GRECC, Bedford VA
> Research Fellow, Department of Neurology, Massachusetts General Hospital
> and
> Harvard Medical School
> Office: (773) 406-2464
> =====================
> This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
> HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
> intended only for the use of the individual or entity named above. If the
> reader of the e-mail is not the intended recipient or the employee or
> agent
> responsible for delivering it to the intended recipient, you are hereby
> notified that you are in possession of confidential and privileged
> information. Any unauthorized use, disclosure, copying or the taking of
> any
> action in reliance on the contents of this information is strictly
> prohibited and may be unlawful. If you have received this e-mail
> unintentionally, please immediately notify the sender via telephone at
> (773)
> 406-2464 or email.
>
>
> On Wed, Nov 17, 2010 at 10:14 AM, Nico <[log in to unmask]> wrote:
>
> > Dear SPM List,
> >
> > I have a question concerning why the result of the parametric modulation
> is
> > different in two (very similar/identical?) ways.
> >
> > Experiment:
> > Subjects receive Stimulus A and B. (e.g. two different odors - duration
> >
> > 6s). Subsequently they have to rate them on a scale. This operation is
> > repeated oftenly (also with different kinds of manipulation).
> >
> > Now, I want to parametrically model my ratings onto the Stimulus A and B
> to
> > see if there are regions correlating with the ratings.
> >
> > As far as i understand I have two options in designing the matrix.
> >
> > 1. I model Stimulus onsets and durations from stimulus A and B
> separately
> > in two differnt columns each combined with the corresponding rating
> > (paramter) -> I get a column for Stimulus A*rating and a column for
> Stimulus
> > B*rating.
> > Contrast: (1 1) -> T-test vs 0
> >
> > 2. I model Stimulus onsets and durations from stimulus A and B together
> in
> > one column - again combined with the corresponding rating (parameter). I
> get
> > one column Stimulus*rating
> > Contrast (1) -> T-test vs 0
> >
> >
> > I thought the results from both models should be almost identical but
> they
> > are not. In fact model 1 seems to fit better for some reason. Could it
> be
> > that putting them alltogether in one matrix inceases the variance to
> strong?
> > I mean there is a strong difference between Stimulus A and B(effective
> as
> > well as rated). Or am I completly wrong and something else might be
> > affecting the data.
> >
> > Any help is appreciated!
> > Nico
> >
> >
> >
--
GRATIS! Movie-FLAT mit über 300 Videos.
Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome
|