Technical Papers
Title: A Second Test of Adaptive Choice-Based Conjoint Analysis (The Surprising Robustness of Standard CBC Designs) Description: This article describes a second test and a refinement of an adaptive choice-based conjoint methodology by Johnson and co-authors. In contrast to the first experiment, the authors did not include a self-explicated importance question in the "priors" used to generate the adaptive designs. Whereas the self-explicated importance question had not seemed to be of much use in a prior ACBC experiment, the lack of this information appeared to derail the success of this test. ACBC is shown to be equally effective for both hit rates and share predictions as standard CBC in this case, but no better.Importantly, the ACBC algorithm is shown to produce more efficient designs with respect to the part worth information it is given. However, the authors suggest that the initial self-explicated data (less the importance ratings) provided a sub-standard starting point for the ACBC algorithm to perform well in this test. Johnson and colleagues point to future developments and strategy that may lead to a successful application of ACBC. Download: |
 Applied methodologiesTechnical papers |
#01
International Market Research Agency
#01 "I thought it was an excellent session - just the right mix of statistics and real life application
info."
|
Download this technical paper