[FieldTrip] depsamplesF

Balázs Knakker bknakker at gmail.com
Mon May 16 22:22:37 CEST 2016

Dear Alex,

As far as I know, depsamplesFmultivariate corresponds to what used to be
the default option to perform a dependent samples F-test (it was called
depsamplesF), which tolerates nonsphericity when there are more than 2
levels by dealing with the question from a multivariate standpoint. But in
your case, you can also test main effects and interactions in your 2×2
design, and in that case I think these will yield identical results, and
you could as well be using T-tests (comparing averages over the other
factor for the main effects, and comparing the differences along one factor
at both levels of the other factor for the interaction, see also this
). If you would decide to disregard the factor structure of your data and
would do something like a one-way ANOVA with 4 levels, I would stick with
the MANOVA approach. I would be curious to know whether the univarate
version is recommended, apart from the case when one wants results
comparable to a conventional ANOVA F-test.

Balázs Knakker

On Mon, May 16, 2016 at 8:55 PM, Alex Sel <alex.sel at psy.ox.ac.uk> wrote:

> Dear list,
> I would like to run a non-parametric cluster-based permutation analysis.
> My experiment has a 2X2 within-subject design, i.e. I would like to compare
> more than two experimental conditions. I am using the cfg.statistics  =
> ‘depsamplesFmultivariate’. I wonder if this would be the correct option
> considering that I only have ONE dependent variable (i.e. ANOVA) and not
> multiple dependent variables (i.e. MANOVA). If this is not the correct
> statistic, would you be able to tell me what is the correct statistic that
> should be apply to tests differences between more than 2 experimental
> conditions in a within subject design?
> Any insight on this would be much appreciated.
> Best wishes,
> Alex Sel, PhD
> Postdoctoral Researcher
> Department of Experimental Psychology,
> University of Oxford,
> 9 South Parks Road,
> OX1 3UD
> Tel: 01865 271 340
> Email: Alex.sel at psy.ox.ac.uk
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20160516/37a9a0fa/attachment-0002.html>

More information about the fieldtrip mailing list