Your data analysis solution
Name is required.
Email address is required.
Invalid email address
Answer is required.
Exceeding max length of 5KB

Conjoint analysis - different results in hierarchical bayes estimation and normal choice model

By Aidan Mulloy | Feb 03, 2017 02:13PM CET


I carried out choice based conjoint analysis using the normal conditional logit model and Hierarchical Bayes regression. I understand that HB analysis gives extra insight as it gives individual level utilities for the various characteristics, however in the HB analysis, when all the mean utilities were aggregated into overall mean utilities, the relative importance of the various attributes changed in order of importance, when compared to the the analysis run through the conditional logit model.

With this in mind, obviously both forms of analysis gives substantially different results when in terms of the overall sample. Which form of analysis would give more reliable results? Is there any reason to think that one form of analysis is more effective than the other in giving accurate results?

Thanking you,




By Sébastien | Feb 03, 2017 03:36PM CET | XLSTAT Agent

Dear Aidan,

Both methods may indeed lead to different results. The HB method is more robust as it borrows more effectively information from the population (means and covariances) describing the preferences of other respondents in the same dataset.
One thing you should be sure of when running CBC-HB is the convergence quality, before interpreting any result.

Jean Paul

Post Your Public Answer

Contact Us

Email Sales

Contact our technical support team:
seconds ago
a minute ago
minutes ago
an hour ago
hours ago
a day ago
days ago
Invalid characters found