control variables / cfg.cvar

[Michael Wibral]

Dear Eric,

Thanks for the answer, which I understood.  Your design is in both cases a between units of observation design. my exmaple (and actually my data) are a within units of observation design:
 
The independent variable in my example is condition (1,2). The units of observation are subjects, i.e. it would be a within unit of observation design. However, to exclude motor effects in the analysis, response hand assigments were balanced over subjects, as is done in many studies. Hence, we have a control variable with two levels. These levels could (naively) be the assigment rules (cond1-hand1/cond2-hand2 and vice versa) however then the nuisance effect is not linked to the control variable but to the interaction control-variable*independent variable. Hence, the copntrol variable should be the response hand itself (hand1, hand2). In this case, each of my units of observation has both levels of the control variable. To permute only within levels of the control variable I would then have to permute between units of observation, i.e. permute cond1 with cond2 between two subjects that had both left hand responses in cond1 and both right hand responses in cond 2 for example.
This way the control variable should work, but i loose the power related to the within UO design.

Michael


-----Ursprüngliche Nachricht-----
Von: Eric Maris <e.maris at DONDERS.RU.NL>
Gesendet: Jul 23, 2010 1:28:59 AM
An: FIELDTRIP at NIC.SURFNET.NL
Betreff: Re: [FIELDTRIP] control variables / cfg.cvar

>Dear Michael,
>
>
>
>> So in my design I have two levels of the control variable: L1 are
>> subjects with c1-R/c2-L pairing and L2 are subjects with c1-L/c2-R
>> pairing (?). If I now permute only within the levels (which I would do
>> anyway if I used a dependend samples test?) then still there are
>> perumtations where response hands are sorted on the two sets to be
>> compared, giving rise to large unwanted clusters in this particular
>> permutation and other that are similar to it.  Sure, this will not
>> result in false positives, but decrease the sensitivity of my
>> experiment.
>> 
>> But maybe I misunderstand the meaning of  "permuting the data sets
>> (single trials or subject averages) within each of the  levels of the
>> control variable" . Does a permutation within the levels mean, I
>> permute left hand responses only with left hand repsonses and right
>> hand responses only with right hand responses? In this case I would
>> have to give up my dependend samples testing (within subjcets) and use
>> independent samples testing (across) subjects, because I can only
>> exchange left hand responses in condition 1 and left hand responses in
>> condition 2 between subjects, correct?
>
>
>An example should be able to clarify things. Control variables are especially useful in between-subject and (single subject) between-trials studies. Let's consider a between-subjects study in which our interest is in assessing the effect of some individual difference variable, such as two different alleles of some gene. Assume that dependent variable strongly depends on age. In that case, sensitivity wrt identifying the genetic effect could be increased (as compared with an analysis without the age variable) by making a number of fairly homogenous age groups, and to perform a permutation test by randomly permuting the two alleles WITHIN each of these age groups. 
>
>I found it hard to use your example to illustrate the usefulness of permutation within the levels of a control variable. Probably, I'm missing a point here. Maybe you can help me by pointing out what is the independent variable whose effect you want to assess and which other variable (the control variable) is also responsible for variance in the dependent variable, but in whose effect you are not interested.
>
>Best,
>
>Eric
>
>
>
>
>
>
>
>> 
>> Thanks for your help on this,
>> Michael
>> 
>> ----------------------------------
>> 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.
>
>----------------------------------
>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.

----------------------------------
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 --------------
A non-text attachment was scrubbed...
Name: Michael Wibral.vcf
Type: text/x-vcard
Size: 628 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100723/5e7c2030/attachment-0002.vcf>