Millisecond Forums

Assigning 3 groups of participants across 6 conditions?

https://www.millisecond.com/forums/Topic15550.aspx

By socialpsych - 2/17/2015

I'm running a study where participants first complete a Qualtrics survey, and then complete an Inquisit experiment. In the Qualtrics survey, participants are randomly assigned to one of 3 conditions, and participants in different conditions see different versions of tasks when they arrive at Inquisit.

I've set up URL parameters so that Inquisit receives their subject number from Qualtrics (for purposes of matching together the two datasets), and also receives the condition they were assigned to from Qualtrics (1, 2 or 3), which is used as their group number. I've then used groupnumber to assign people to see different sequences of blocks:

<expt version1>
/subjects = (1 of 3)
/groupassignment = groupnumber

This is all working fine... but!!! (there is obviously a 'but' coming somewhere or else I would not be here.) I've just realised that within one of the tasks all my participants complete during the Inquisit session, I also need to counterbalance the order of the blocks across participants.

Is there any way to use groupnumber in this way, but then within these groups, randomly assign people to two block orders?

It seems like I could possibly do this by splitting my tasks into separate files and making two <batch> orders, but I'm wondering if there is an easier way to do this just using my existing single script. I only have one Inquisit license, which rules out the solution of redirecting participants to three separate experiments.

Many thanks to anyone who can offer me any advice!
By Dave - 2/18/2015

If you need to *counterbalance* block order, you need more conditions (six, as you yourself mention). You might set up *six* conditions in Qualtrics, where conditions 1 & 2 are identical in terms of the Qualtrics portion, 3 & 4 are identical, 5 & 6 are identical. Then forward to Inquisit with the condition parameter and assign to *six* <expt> elements, where <expt version1> and <expt version2> are identical *except* for the reversed block order and so forth.
By socialpsych - 2/22/2015

Dave - that seems really simple now that you put it like that. Thanks for the advice!