Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Multiple CBC's in the same questionnaire

We have a study in which we have 3 different CBC's, each with 10 versions. We find in our testing that if "version x" is picked up by the program for the first CBC, the same "version x" is picked up for the remaining CBC's. Is there and way of ensuring that different versions are picked for the 3 CBC's for the same respondent?
related to an answer for: Rotation of versions for CBC
asked Sep 26, 2012 by Asset (340 points)

1 Answer

+1 vote
If the designs have the same number of attributes and levels, and same number of concepts per task and #tasks, and the same DESIGN SEED, then indeed each design would be identical within a subject.

Remember, you set the design seed when you generate the experimental plan within our software.  By default, it is set to a seed of 1.  The simple solution is to generate the second CBC study with a different design seed (a different integer), and the third CBC study with yet another unique integer.  That will create unique designs for each respondent, rather than having the designs be precisely the same due to the exact same setup (including seed) across the three conjoint modules.
answered Sep 26, 2012 by Bryan Orme Platinum Sawtooth Software, Inc. (164,715 points)
...