Have an idea?

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

CBC/HB Analysis of zero-centered diffs


I've conducted a CBC survey. The report of my HB report states that the average utilities are zero-centered diffs. I'm not completely sure what that means?
Can I compare the utilities to each other or do I have to adjust/equalize the utilities somehow, so that I can compare them directly?
Further the average importances - do I have to apply any method or are these my results of my study?

Thank you very much!
asked Aug 26, 2015 by anonymous

1 Answer

0 votes
Zero-centered diffs are just a way to put each respondent on the same scale so that each respondent gets essentially equal weight when averaging the scores for groups or for the sample.

Even after this normalization, the cardinal rules of interpreting conjoint results apply: you cannot directly compare one level of one attribute to one level of a different attribute.  You can compare levels within the same attribute.

Regarding importances, this is just a way to summarize the impact each attribute can have (in terms of going from worst to best levels you included in the study) and making them sum to 100%
answered Aug 26, 2015 by Bryan Orme Platinum Sawtooth Software, Inc. (173,090 points)
How to get this summarized report?
Importance Scores/Zero-centered diffs?