Hi there,
I am creating a survey for a 4x4x2 within-subjects design where
- Each question is a different vignette.
- There are 32 vignettes total, since there are 32 possible combinations of the levels of the three variables.
- Only one sentence in each vignette is systematically varied to display one of the 32 combinations. The rest of the vignette is regular text (this text is also different in each vignette, since each vignette is different, but it is not systematically varied or being looked at as a variable).
- The vignettes themselves are not presented in random order. The systematically varied sentence in each vignette is displayed in random order.
- Each participant needs to see each of the 32 possible combinations, and not see any of the same combinations twice (non-repetition).
I have been scouring the forum and have found similar posts but not exactly right for what I need. So far I have used embedded data and the survey randomizer for this design. Right now the vignettes are being generated perfectly (i.e., the embedded data is working) but the combinations are not being evenly presented across questions. I.e., across vignettes, the participant is seeing the same combination more than once and not seeing all 32 combinations overall. I imagine the solution involves a different way to use the randomizer and/or display logic? Here is what my survey flow currently looks like:
Each vignette is its own block. The randomizer is first in the flow for each block, with the 32 combinations included under embedded data. Right now the randomizer is set to present 1 of the 32 elements.
Any help is much appreciated.
