Can questionorder.xls be referenced by responses.xls formula during a session?

[I][edited by moderator][/I]
I want to use a formula on the [I]responses.xls[/I] sheet to reference the [I]question_order.csv[/I] sheet in the middle of an experiment. Or at least, that's the only way I can think of to randomize the presentation of 4 stimuli, but then later, show these stimuli again, but in the same order that they were shown the first time within the experiment.

Is this possible? Can you think of other ways if you can't reference questionorder.xls with responses.xls? Again, I want to show, for instance, subject a and b the same four stimuli, 1, 2, 3, and 4. I randomize the presentation, and subject a sees 1,3,4,2 and b sees 2,1,4,3. However, at a later point within their respective sessions, I want to show each subject the stimuli again, and I want a to see 1,3,4,2 and I want b to see 2,1,4,3. It's crucial that each subject sees the stimuli in the same order both times they see the stimuli in the experiment.

thanks!

There is no built-in function to do this in MediaLab (at least not as of v2008.1.33.

I can imagine this being possible but not by accessing question_order.csv file (which I agree would be a nice feature!!). It would be some tricky conditional logic coding in Excel, but I imagine you could create variables that would record the order of the 4 items. These "order" variables could check to see how many of the items had values. When only one of the four items has a value in the Excel file, then that item must have come first. When two items have values then the second one must be the one that has a value but that is not the one defined as the 1st. And so on.

Do you see where I am going with this?

Also just curious--does this thread also cover your question from this thread (see link below)? The current thread seemed to be a simpler version of this earlier one:

[URL]http://www.empirisoft.com/support/showthread.php?t=2382[/URL]

Also, I should mention that there is a function in DirectRT that does exactly this--i.e., allow you to present stimuli later in a repeated random order.Is that a possibility for you? Drawback is that the current version (v2008) does not display HTML files--they would need to be converted to images or text. Just a thought.

Thanks for the help. I had figured the conditional logic stuff might be the only way to go. And yes, the post to which you responded was a simplified version of the first I posted.

Not sure if the lab has access to DirectRT, but I will check it out.

Thanks,

Sam
Judgment and Decision Lab
University of Michigan