randomly drawing stimuli from a bank

[mpatton]

Hello,

We would like to present participants with a series of approximately 25 scenarios randomly drawn from one of two banks (each of which contain 50 scenarios). We are using common dependent measures for each scenario, but we need to keep track of which scenario was drawn to match the dependent measures with the scenarios.

I know this can be mostly accomplished by writing unique dependent variables for all 100 scenarios and using Randomize-W, but this will require dividing the stimuli bank into chucks of 25 conditions and randomly assigning people to one of those conditions. (We’d also like to save the time of creating 100 unique variables manually if possible!)

Thought we’d check. Thanks!

[jarvis24]

You might be able to do it like this:

Create two versions of the same questionnaire--one for each bank. Use identical variable names in each. Use just one condition. In your experiment file, for this condition, assign each of the two questionnaires a WGR value of 1. That will randomize which of the two .que files is administered first. Then, at the end of each questionnaire, have a final item which unconditionally skips to some variable which is contained in neither of the two versions. That variable should be the first variable of the next querstionnaire you would like them to complete.

The gist is that when they finish whichever version they get first, MediaLab will skip all the items in the next version and will resume when it finds the skipto variable located in some other followup questionnaire.

Does that make sense? Let me know if you have any trouble with it!

See also: www.empirisoft.com/support/showthread.php?t=462