Solved
Randomization without replacement
I have 4 vignettes that I would like each participant to see, in a random order, and then answer one set of questions (from 4 possible sets) after seeing each vignette. I have put all 4 vignettes into a single block and used the randomizer to specify that 1 vignette should be displayed randomly. Then I have the 4 sets of questions in 4 blocks and have used the randomizer to specify that 1 block should be shown. That all works fine, but the problem arises when I go to the next step, with is showing another vignette and another block of questions. My intention is for each of the 4 vignettes and each of the 4 blocks of questions to be shown once, in a random order (for 16 possible combinations) and they should not be repeated once they have been shown. However, once I get to the subsequent sections, it is still just showing any 1 of the 4, including possibly the one they have already seen. Essentially I need randomization without replacement, where each item can only be used once in the survey. Is this possible?
Best answer by JenCX
@hwilliams Sorry about that! Okay, so you can do branches within the randomizer to group items of tasks together. In this case, you'll use a branch with any statement that will always be true (I used RecipientEmail not empty) to group your block with your image together with an embedded data field to capture that you used that image:
!
Then for the remaining 3 sets of randomizers, you'll just double-check that your Embedded Data variable wasn't used for each option, and be sure to set the tag at the end of the branch again:
!
Then for the remaining 3 sets of randomizers, you'll just double-check that your Embedded Data variable wasn't used for each option, and be sure to set the tag at the end of the branch again:
!
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.

Here's how I have it set up right now, but I can of course change it if it won't work this way. Participants need to see one of the primes, then one of the question blocks, then it repeats 3 times such that they have seen all 4 primes and all 4 question blocks with no repeats. So I have this same thing repeating 3 more times, but obviously right now the randomizer can pick one of the same ones that a previous rendomizer picked.
For the branching bit, you'll do a randomizer but include branches that check for the use of that Embedded data bit:
!