Randomise stimuli according to 2 criteria???

[Leisha Wharfield]

Dear Anna,

I've struggled with this question, too, in an experiment where we showed
smoking & food-related pictures and words. We didn't want more than two
stimuli in the same category (i.e., smoking-related image followed by
smoking-related image), yet the selection of the stimuli was random. Now
I think I would try to write a bit of script that would analyze the last
two stimuli and avoid choosing the same type three times in a row, but I
think that might result in a bunch at the end that would all be the
same, which may cause an error.

I'd be interested in other programmers' thoughts on this one.

Leisha Wharfield
Decision Research
Eugene, Oregon, USA

Anna Holmes wrote:

> Dear fellow E-Primers,
>
> As an E-Prime rookie (with no scripting experience), I am hoping that
> someone may be able to help me.
>
> I am trying to create a lexical decision priming task with 2 blocks of
> the following stimuli:
>
> Number of trials
>
> 	
>
> Condition Number
>
> 	
>
> Condition Name
>
> 	
>
> Correct response
>
> (ResponseType)
>
> 12
>
> 	
>
> 1
>
> 	
>
> Expected-related
>
> 	
>
> Yes (1)
>
> 12
>
> 	
>
> 2
>
> 	
>
> Expected-unrelated
>
> 	
>
> Yes (1)
>
> 4
>
> 	
>
> 3
>
> 	
>
> Unexpected-unrelated
>
> 	
>
> Yes (1)
>
> 4
>
> 	
>
> 4
>
> 	
>
> Unexpected-related
>
> 	
>
> Yes (1)
>
> 12
>
> 	
>
> 5
>
> 	
>
> Neutral
>
> 	
>
> Yes (1)
>
> 44
>
> 	
>
> 6
>
> 	
>
> Nonword
>
> 	
>
> No (2)
>
>
> I understand that script can be written (e.g., samples - No Repeats;
> String Array to Populate List) such that the same type of stimulus
> condition is not presented on consecutive trials (e.g., another
> related word pair will not directly follow a related word pair).
>
> However, I am trying to work out if I can randomise my stimuli
> according to 2 criteria:
>
>    1. That no condition occurs more than 3 times consecutively
>    2. That no single response type occurs more than 3 times
>       consecutively (no more than 3 yes (or no) responses in a row)
>
> I have chosen these two criteria because if I only randomise according
> to condition, it will likely result in strings of stimuli that require
> a 'yes' response as 5 of the 6 conditions call for the participant to
> respond 'yes'.
>
> Hope my question is clear although reasonably complicated.
>
> Look forward to any suggestions.
>
> Thanks
> Anna
> -----------------------------------------------------------------------
> Anna D Holmes. BSpPath(Hons)
> Division of Speech Pathology
> School of Health and Rehabilitation Sciences
> University of Queensland
> QLD 4072 Australia
> Phone: +61 7 3346 9726
> Fax: +61 7 3365 1877
> a.holmes at uq.edu.au <mailto:a.holmes at uq.edu.au>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/eprime/attachments/20040916/69fa4b29/attachment.htm>