Selecting trialtypes from a larger set without replacement

[Ben Robinson]

hi marc.
here's how i would do this.  make a new list object.
first row: blockproc.
second row: breakproc.
3rd row: blockproc.
then breakproc
then blockproc
then breakproc
then blockproc.
that gives you 4 blocks, with a break procedure between each.
on the blockproc timeline set your totaltriallist.  set the totaltriallist to Random Selection, and to Exit List after 36 trials, and to Reset Sampling until all samples have been used.  this will randomly select 36 of your 144 trials, then exit the list.  the next time the list is run, during the next blockproc, the sampling will resume where it left-off - never again pulling any of the 36 already-sampled trials.

ben robinson

>>> Marc Berman <bermanm at umich.edu> 9/4/2007 2:51 PM >>>
Hi,
 
   I'm currently coding a study where I have a large set of trials that 
I want to present (144 many of which are unique), but only want to have 
blocks of 36 trials with rest breaks in between.  I'm wondering if I 
nest the blocklist, into my totaltrial list, if I will be able to sample 
from the 144 trials, 4 times, and get all of the 144 trials without 
replacement.  If this will not work, is there some other way to do it?  
Thanks so much and hopefully, this is clear.

Thanks in advance,
-Marc