I would like to run a partial profile conjoint analysis on 6 attributes. I prefer that trials be limited to 3 attributes each (so not to overwhelm the survey taker). Assuming the sample was big enough (each respondent gets 4 trials) and I had many responses of 3-attribute responses, can JMP 13 provide the results that I need to understand the desirability of the 6 attributes?
Take a look at the online documentation for Choice Models, specifically the design generation. If you don't have JMP you could get the trial version and walk through the examples to see how it fits your needs.
Hi, I'm new at this so if you have advice please provide:
From the Design Generation module, my inputs would be:
3 of 6 attributes can change
2 profiles per choice set
4 choice sets (trials) per survey
# surveys ..... this is what I need help with; how many pre-determined choice sets do I need to create
survey population: ~100 responses of a total survey population of 300
Right now, I've scripted choice sets to be created at random. However, I'm concerned that because there are so many possible random combinations, the data I receive might be too shallow. So, do I need to predetermine X number of choice sets and only deploy those?
Thanks.