[Eeglablist] Random selection of trials

[Philip Michael Zeman]

Hello Kris

This is an issue that I've encountered myself and I've had to come up with 
some logical reasoning.  I am also very interested in hearing what other 
people have to say on this topic so that I know what options exist.

My reasoning is as follows:

If there is any chance of learning or adaptation of the neural systems 
involved in the paradigm to the stimuli or the context of the stimuli, then 
I would chose to use only the first 150 trials of each of the conditions you 
are examining.  There is plenty of ERP research out there that shows that 
the second-half vs. the first have of an experiment has characteristically 
different waveforms. (See, face recognition/unknown face ERP literature, 
e.g., Tanaka - I think that participants begin to learn what could be 
considered the 'odd-ball' in this experiment).

Your concern about averaging a different number of trials for each 
condition:
*Yes, this is a concern.  My philosophy is to always compare apples to 
apples --- always try to do the same processing in condition 1 and in 
condition 2 if you are going to ask the question, "are these 2 conditions 
different?" using a statistical test.  If in condition 1 there are 150 
trials and in condition 2 there are 1000 trials (and you use them all) then 
when you average the data within each participant and then look for a 
difference in the distribution across participants you already know there is 
a possible difference simply due to what you did to the data.  Depending on 
the signal-to-noise characteristics (and what we consider to be noise, e.g., 
variability in behaviour, uncorrelated sensor noise, learning and 
adaptation), you might have a distribution of values in condition 2 that are 
tightly distributed around the mean for the condition compared to in 
condition 1 that has a distribution of scores that are spread by a large 
amount around the mean of the condition.

If you are not concerned about learning and adaptation in your experiment, 
(and I think this goes against standard research hypothesis testing rules) 
what I would suggest is doing your hypothesis test in multiple comparisons 
so that you can learn something about your data.  I would answer my 
hypothesis using (1) and then go further by investigating (2) and (3) to see 
what I get.
(1) compare condition 1, trials 1 to 150 vs. condition 2, trials 1 to 150
(2) compare condition 1, trials 1 to 150 vs. condition 2, trials N-150 to N 
(where N is the number of trials in the large set)
(3) compare condition 1, trials 1 to 150 vs. condition 2, random blocks of 
150 trails (and do this multiple times on different random blocks)

Keep me updated as to what you decided to do.


Regards,

Phil

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Philip Michael Zeman B.Eng, Ph.D.
Applied Brain and Vision Sciences Inc.
Brain Function Analysis for Novel Paradigms and Serious Games
Analysis of Pharmaceutical Effects on Brain Function
http://www.abvsciences.com
Latest Brain Research Result:
http://www.spatialbrain.com
Email:   pzeman at alumni.uvic.ca
Phone: +1-250-589-4234
LinkedIn Profile: http://ca.linkedin.com/in/philipmichaelzeman
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

----- Original Message ----- 
From: "Kris Baetens" <Kris.Baetens at vub.ac.be>
To: <eeglablist at sccn.ucsd.edu>
Sent: Wednesday, March 31, 2010 8:04 AM
Subject: [Eeglablist] Random selection of trials


> Dear colleagues,
>
> We employ a paradigm which inherently leads to a different number of 
> trials in both our conditions (oddball-like). We have two conditions, one 
> with an average of about 150 trials, the other with about 1500 
> (artefact-free).
>
> - Does anybody have research to support my concern that comparing both 
> conditions with the total number of trials may lead to artificial effects 
> due to the different number of trials (and associated variance and 
> "cleanliness" of the gavg's)? (I have seen such things published before.)
> - Does anybody know of an easy way to make a random selection of a 
> predetermined number of trials out of the total number in EEGLAB or 
> MATLAB? (Which would allow for selecting an equal number of trials in both 
> conditions.) Obviously, we don't simply want to take the first or last 150 
> regular trials, since this would possibly lead to erronous conclusions.
>
> Thank you very much in advance,
>
> Kris Baetens
> Ph.D. fellow of the Research Foundation - Flanders (FWO)
> Dept. Experimental and Applied Psychology
> Faculty of Psychology and Educational Sciences
> Vrije Universiteit Brussel
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to 
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to 
> eeglablist-request at sccn.ucsd.edu