Dear SoSci Survey Support Team,
we are currently trying to implement the randomization for a 3×2×2 within-subjects design in our study.
The three factors are:
• Text type: human vs. AI-like vs. neutral (3 levels)
• Perceived origin: German vs. Arabic name (2 levels)
• Gender: male vs. female name (2 levels)
This results in 12 experimental conditions. Each participant should see all 12 conditions exactly once. At the same time, the assignment of the specific essays to the conditions should be rotated across participants, and the order of the 12 trials should be randomized.
We have already created 60 variables for trial 01–12 (essay, name, origin, gender, and text type) and are currently trying to implement the randomization using PHP and put().
Our previous approach used separate randomizers for essays and names, but this resulted in an unbalanced assignment of the experimental conditions across participants - this was an problem in the data analysis...
Could you please advise us on the recommended way to implement this type of complete 3×2×2 randomization in SoSci Survey? Is there a built-in randomization function or a simpler approach that would ensure that every participant receives all 12 conditions exactly once while also rotating the essay assignment and randomizing the trial order?
Thank you very much for your help!
Best regards,
Amalia