Hi; in this case you would need to put more of the lower-level copes
together in larger second-level analyses. You will need to combine several
features from the different examples on the web page; for example,
combining paired t-test and the 2-factor-2-level ANOVA design.
Good luck! Steve.
On Thu, 31 Jul 2003, Anda van Stegeren wrote:
> clear. All EV's conform the example on the website and that x 4.
>
> But what if I do want to look at interactions between the 4 categories and conditions?
> and at interaction between (the between-subjects) FactorC and conditions?
> and at interaction between FactorC and activations in the 4 categories?
>
> .....
> Anda
>
> -----Oorspronkelijk bericht-----
> Van: Stephen Smith [mailto:[log in to unmask]]
> Verzonden: wo 30-7-2003 4:49
> Aan: [log in to unmask]
> CC:
> Onderwerp: Re: [FSL] Within and between subjects design Higher Level Analysis
>
>
>
> Hi. unless you want to do interactions between categories and conditions,
> you may as well analyze each first-level contract separately at second
> level.
>
> Your first EV should probably be the mean across all subjects and
> categories, and the second can be the difference (A vs B). You probably
> want the paired-t-test setup; see
> http://www.fmrib.ox.ac.uk/fsl/feat5/detail.html#PairedTwoGroupDifference
>
> To find C, take the C-factor measures, de-mean them and put them in a
> separate EV to get this.
>
> For your final question, (cat1-BL)-(cat2-BL) is the same as (cat1-cat2).
>
> Thanks, Steve.
>
>
>
> On Wed, 30 Jul 2003, Anda van Stegeren wrote:
>
> > Hi,
> > 2 Questions:
> >
> > In a quite complicated design where we finished all first level analyses, I created (among others) 7 contrasts that produced the accompanying COPE images.
> >
> > Contrast 1 - 4 consist of contrasting 4 different categories of stimuli presented to a repeating baseline condition in which grey screens were presented. (cat1>BL, cat2>BL etc.)
> > Contrast 5-7 consist of contrasting category1 to cat2, 3 and 4 respecively. (cat2>1, Cat3>1, cat4>1)
> >
> > All subjects had ConditionA during session 1 and ConditionB during session 2 or vice versa (crossed-over of course).
> >
> > 1) We first want to find out if fMRI activitity during ConditionA differs from ConditionB
> > 2) If the effect of ConditionA versus ConditionB varies across the categories.
> > 3) Subjects differ (between) on one other personal variable (factor C): what is the effect on main effect or on the interaction effect of this factor
> >
> > We now plan for a MAIN effect analysis and want to
> > 1) enter all cope images from contrasts 1-4 of all subjects vertically
> > 2) Create EV 1: to create a difference between Condition A versus B
> > 3) Create EV 2-5 : to check if Condition is interacting with different categories
> >
> > QUESTION 1: How do we enter the Between-factorC to make a within subjects comparison of Condition and Between subjects comparison of Factor C + interaction effects; how can we proceed ?
> >
> > QUESTION 2: Can we expect a difference if we first analyse contrasts 1-4 and then create new contrasts between each of them,
> > from a second procedure: directly entering contrasts 5-7 in the higher level analysis.
> > Or: is the contrast to a baseline (grey screens) condition with later 'subtraction' of each 'df category - baseline' different from comparing the categories directly with each other ?
> > how are variances etc. involved here ?
> >
> > Thanks,
> >
> > Anda van Stegeren
> >
> >
> >
> > Drs. Anda van Stegeren, VUMC &University of Amsterdam
> > Department of Clinical Psychology
> > Roetersstraat 15
> > 1018 WB Amsterdam
> > The Netherlands
> > tel. ++ 31 20 525 6799
> > fax. ++31 20 639 1369
> > E-mail: [log in to unmask]
> >
> >
> >
> >
> >
>
> Stephen M. Smith MA DPhil CEng MIEE
> Associate Director, FMRIB and Analysis Research Coordinator
>
> Oxford University Centre for Functional MRI of the Brain
> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> +44 (0) 1865 222726 (fax 222717)
>
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
>
>
>
Stephen M. Smith MA DPhil CEng MIEE
Associate Director, FMRIB and Analysis Research Coordinator
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|