clarification on counterbalancing question

[Leisha Wharfield]

Hi, Prof. LoCasto,

I would accomplish your task with one list that has three other lists 
nested in it.

The master list would have 6 items in it and one column, perhaps named 
ABC. Each of the 6 items would have its own nested ABC list, perhaps 
named ABC1 through ABC6. Each of those nested lists would have three 
items in it: the a, b, & c version of that number's data set. Master 
list would choose the six items randomly; nested lists would choose 1 
item randomly.

You want to make sure that each person gets 2 a's 2 b's 2 c's, & 
Eprime's built-in counterbalancing does not work? What if you keep a 
string, a character string, as an attribute that gets your List Response 
added to it as you go, then check its contents with an in-line object? 
You could write an If that would check to see whether the object 
contains two a's, two b's, or two c's. If so, you could blank out the 
items from the upcoming nested lists that you don't want--that is, set 
their weight to 0 so that they can't be selected.

Total control. You want something easy, though. Maybe something easier 
than writing in-line code? Alas, when you want to do something a bit 
different, something that is not out of the box, you have to write a 
little in-line code. And it's very satisfying when you get the results 
you want.

Maybe someone else has an easier way?

Leisha Wharfield
Decision Research
Eugene, Oregon, USA


LoCasto, Paul C. Prof. wrote:

> Hello all-
>
>  
>
> I see I might have been unclear in my original post- I apologize. I am 
> aware of being able to create a number of lists- and for each list 
> counterbalance stimuli manually i.e:
>
>  
>
> List 1= 1a, 2b, 3c, 4a, 5b, 6c,
>
> List 2= 1b, 2c, 3a, 4b, 5c, 6a
>
> List 3= 1c, 2a, 3b, 4c, 5a, 6b
>
>  
>
> and then at the Block level have E-prime choose to present one of 
> these lists using the counterbalance display option. However, unless I 
> manufacture each possible list sequence this is really pseudorandom in 
> that anyone who sees stimulus 1a, will also see stimulus 4a and 5b etc.
>
>  
>
> Would someone be so kind as to tell me how I might list my stimuli in 
> a way that has E-prime *randomly* select whether a Ss sees version 
> 'a', 'b', 'c' of a given stimulus and across the whole experiment 
> would see an equal number of stimuli across the three conditions (a, 
> b, c) where no stimulus (e.g. stimulus 1)  was seen in more than one 
> version (e.g. 1a *or* 1b *or* 1c).
>
>  
>
> I would really appreciate any insight you can afford. Thanks.
>
>  
>
>  
>
> Paul
>
>  
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/eprime/attachments/20061031/afe90e2e/attachment.htm>