Hi, I was hoping to get some help with the survey flow for my study.

I am conducted a repeated-measures study where all participants complete 18 trials of a task (labeled T1 to T18). Each trial has 6 potential versions (e.g., for T1, there is T1.A, T1.B, T1.C, T1.D, T1.E, T1.F). I want to design the survey flow so that for each trial, participants are randomly assigned to one version of each trial. However, they should only get three of each "version" (i.e., if they get T1.A, T2.A, and T3.A, they should not get any more version A's for the rest of the study). For instance, once they reach 3 "A" versions, the following trials should be randomly assigned to versions B-F. The outcome should be that each participant will get eighteen trials total - three of each version. Any ideas on how to program the survey flow to do this?

Note: I am also randomizing the presentation order of the trials!