nested list items repeating across blocks

[Cecilia Westbrook]

I have an experiment in which I'd like to present the conditions (neg, neu, etc.) in pseudorandom order, but have the actual stimuli randomized between participants. I've accomplished this by creating an order list (neg-neu-neu-neg-neu etc.) which then draws from a nested list for each condition. All of the stimuli in a given condition are in the nested list, with the idea that a given negative stimuli could appear in block 1 for some participants, block 2 for others, etc., but the condition order is the same for everyone.

The nested lists each have the number of stimuli (48) that I want for that condition. The task is broken into six blocks, and each block contains some number of trials of each condition, totaling to 48 across the entire task. The problem is that instead of going through all 48 stimuli, I'm finding that stimuli are repeating. I suspect this is happening because the stimuli lists are nested within the block list, and every time the block list ends, the stimuli lists are resetting too. Is there any way to keep the stimuli lists sampling on their own, without resetting? I've set "No Repeat After Reset" to "Yes," but it's still happening anyway.

I've attached my experiment to show what I mean. If anyone has any advice I'd be grateful--I can't figure this one out.

Best,

Ceci

[David McFarlane]

Ceci,

A couple of little-known facts about using nested Lists:

1) Whenever a row of List calls a nested List, it uses up one sample
from that nested List, even if the Procedure does not use the sample.

2) When you use colon syntax (Attrib:n) to get extra samples from a
nested List, it uses up n+1 samples from that List, even if your
Procedure does not use all of them.


So, e.g., the first row of your Block1ListCoffey calls upon all four
nested Lists, even though it needs samples from only two of those nested
Lists. But that also uses up samples from the extra nested Lists. So
your program goes through your nested Lists faster than you think it does!

But don't take my word for that, test it for yourself -- remove *all*
randomization from *all* Lists so that they run in a knowable sequence,
and see what happens. You should find that your trials skip over some
stimuli in the way that I described. (General development tip -- Set
all Lists to Sequential during development, and leave the randomization
only for the final stage of development!)

I know that it is convenient to just repeat the same list of nested
Lists for each row of your List here, but that is a mistake. To fix
this, for each row of your List try putting only the nested Lists
required for that row.

---------------
David McFarlane
E-Prime training online:
http://psychology.msu.edu/Workshops_Courses/eprime.aspx
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster)