Print

---

Dear Stef,

You could use ImCalc to compute linear combinations of your
within-subjects images. Otherwise, as you mention, you could obtain the
same images from a GLM, even if inefficiently: enter all of your images
in a GLM where the design matrix is the identity matrix (e.g. a one-way
ANOVA with one image per cell, assuming sphericity) and use contrasts to
create the images of interest.

Best regards,
Guillaume.


On 04/01/2019 09:50, Stef Meliss wrote:
> Dear Guillaume,
> 
>  
> 
> I have a (very basic) follow up question.
> 
>  
> 
> I would like to perform VBM analysis in a 2(Group: Experimental vs
> Control) x 2 (Time point: pre vs post) mixed design. How do I specify
> the first level in the context of VBM? Would I simply set up a t-test
> containing the images for each time point separately for each subject
> and then specify the contrasts as described? Or should I use the fMRI
> First Level specification to do that?
> 
>  
> 
> Many thanks in advance,
> 
> Stef
> 
>  
> 
> *From: *"SPM (Statistical Parametric Mapping)" <[log in to unmask]> on
> behalf of "Flandin, Guillaume" <[log in to unmask]>
> *Reply-To: *"Flandin, Guillaume" <[log in to unmask]>
> *Date: *Tuesday, 4 December 2018 at 17:51
> *To: *"[log in to unmask]" <[log in to unmask]>
> *Subject: *Re: [SPM] Flexible factorial design
> 
>  
> 
> Dear Geneviève,
> 
>  
> 
> The document mentioned by the reviewer is actually published as a
> 
> chapter of the SPM book. I also list further references that you might
> 
> find useful.
> 
>  
> 
> Penny, W. & Henson, R.N. (2006). Analysis of Variance. In K. Friston, J.
> 
> Ashburner, S. Kiebel, T. Nichols, and W. Penny (Eds), Statistical
> 
> Parametric Mapping: The analysis of functional brain images. Elsevier,
> 
> London, 2006.pp. 166-177.
> 
> http://www.mrc-cbu.cam.ac.uk//personal/rik.henson/personal/PennyHenson_SPM_06b.pdf
> <http://www.mrc-cbu.cam.ac.uk/personal/rik.henson/personal/PennyHenson_SPM_06b.pdf>
> 
>  
> 
> Henson, R.N (2015) Analysis of Variance (ANOVA). In: Arthur W. Toga,
> 
> editor. Brain Mapping: An Encyclopedic Reference, vol. 1, pp. 477-481.
> 
> Academic Press: Elsevier.
> 
> http://www.mrc-cbu.cam.ac.uk/wp-content/uploads/2015/03/Henson_EN_15_ANOVA.pdf
> 
>  
> 
> Friston, K.J. et al (2002) Classical and Bayesian Inference in
> 
> Neuroimaging: Applications. NeuroImage. 16:484-512.
> 
> https://doi.org/10.1006/nimg.2002.1091
> 
>  
> 
> McFarquhar, M. (2018, April 10). Modelling group-level repeated measures
> 
> of neuroimaging data using the univariate general linear model.
> 
> https://doi.org/10.31234/osf.io/a5469
> 
>  
> 
> Flexible factorial design is indeed SPM terminology and refers to an
> 
> option in the interface to specify the design matrix (and non-sphericity
> 
> assumptions) of a GLM.
> 
> Instructions for a 2x2 mixed ANOVA are given here:
> 
> https://en.wikibooks.org/wiki/SPM/Group_Analysis
> 
>  
> 
> Best wishes,
> 
> Guillaume.
> 
>  
> 
> On 22/11/2018 11:49, Genevieve Allaire-Duquette wrote:
> 
>     Hi,
> 
>     I recently submitted a manuscript where I used a "random effects
>     flexible-factorial design" with one within-subject and one
>     between-subject factor (2X2) in SPM8.
> 
>     One reviewer criticized the approach (see below) and I'd like to
>     know if there is some relevant literature I could use to explain the
>     model better.
> 
>     Many thanks,
> 
>     Geneviève Allaire-D.
> 
>     *
> 
>     The use of "random effects flexible-factorial" is not given a
>     published citation so one looks for supporting literature. The SPM8
>     manual only describes this modeling nonstatistically, in section
>     10.2.8. Henson and Penny's Technical Report (2003? 2005?)
>     (https://www.fil.ion.ucl.ac.uk/~wpenny/publications/rik_anova.pdf)
>     does not make any note of the flexible-factorial concept, and where
>     it is available on the web it seems not to be elucidated in detail.  
> 
>     No other software product appears to use the phrase, and it is not
>     in common parlance. Thus, it will have to be explained to the
>     readership in much finer detail to be interpretable here.  
> 
>     Moreover, a recent critique by Chen et al (Neuroimage
>     2014, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4121851/) shows
>     that SPM is hindered by serious limitations and can readily
>     overestimate significance.
> 
>  
> 
> -- 
> 
> Guillaume Flandin, PhD
> 
> Wellcome Centre for Human Neuroimaging
> 
> UCL Queen Square Institute of Neurology
> 
> London WC1N 3BG
> 
>  
> 

-- 
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
UCL Queen Square Institute of Neurology
London WC1N 3BG