Question about permutation testing of an interaction in a two by two design

Guillaume Rousselet g.rousselet at PSY.GLA.AC.UK
Tue Nov 24 11:18:02 CET 2009


Dear Stephan,

to build an appropriate null distribution, the resampling procedure
must stick to the structure of your experimental design. In the case
of a group x condition (repeated measure) interaction, the groups are
independent but the conditions are not. To take that constraint into
account, I think you need to resample from each group independently.
However, for each group, because of the paired nature of the design,
when you sample 1 subject, you have to sample both conditions from
that subject (and all the time points and sensors), i.e. you sample
subjects with replacement independently from each group, but keeping
paired data points together.
The strategy described here relies on a sampling with replacement
under H0. In that case H0 is created by centering the data
independently in each of the 4 cells of your design. So you start with
a true null hypothesis and through resampling you can derive a table
of F values for the interaction under H0.

Hope this help,

GAR



On 24 Nov 2009, at 09:19, Stephan Moratti wrote:

> Hi all,
>
> Thanks for the interesing discussion about the interaction with
> respect to
> permutation. I have been also struggling with this question. As we
> want to
> test the interaction (and not the main effect condition), I am not
> sure if
> resampling condition would produce the distribution of our null
> hypothesis.
> Regarding the interaction, we want to check if the mean values across
> conditions follow the same pattern in two groups (if we consider a
> group x
> condition interaction). If we make a line plot an interaction would
> be indicated
> by a line crossing if each line represents a group (ideally). No
> interaction
> would be represented by parallel lines for each group. So I wonder
> if by
> resampling the subject values between the groups keeping the condition
> structure intact, would create our null distribution for the
> interaction. If we
> resample condition, we would destroy the condition structure.
>
> What do you think? I would be happy for any input.
>
> Best,
>
> Stephan
>
> ----------------------------------
> The aim of this list is to facilitate the discussion between users
> of the FieldTrip  toolbox, to share experiences and to discuss new
> ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html
>  and http://www.ru.nl/neuroimaging/fieldtrip.




************************************************************************************
Guillaume A. Rousselet, Ph.D.

Lecturer

Centre for Cognitive Neuroimaging (CCNi)
Department of Psychology
Faculty of Information & Mathematical Sciences (FIMS)
University of Glasgow
58 Hillhead Street
Glasgow, UK
G12 8QB

The University of Glasgow, charity number SC004401

http://web.me.com/rousseg/GARs_website/

Email: g.rousselet at psy.gla.ac.uk
Fax. +44 (0)141 330 4606
Tel. +44 (0)141 330 6652
Cell +44 (0)791 779 7833


"YOU ARE NOT NORMAL. [...] you probably belong to the minority of the
world's population that has a steady job, adequate access to social
security, and enjoys substantial political freedoms. Moreover, you
live on more than $2 a day, and, unlike 860 million others, you can
read. The percentage of humanity that combines all of these attributes
is minuscule. [...]
Statistically, a "normal" human being in today's world is poor, lives
in oppressive physical, social, and political conditions, and is ruled
by unresponsive and corrupt government."

	Moises Naim - 2005 - Editor in chief of Foreign Policy
************************************************************************************






----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20091124/e9b88970/attachment.html>


More information about the fieldtrip mailing list