Axiomatic and numerical conjoint measurement: An evaluation of diagnostic efficacy

摘要: Synthetic data are used to examine how well axiomatic and numerical conjoint measurement methods, individually comparatively, recover simple polynomial generators in three dimensions. The study illustrates extensions of (NCM) identify model distributive dual-distributive, addition the usual additive, structures. It was found that while minimum STRESS criterion fit, another statistic, predictive capability, provided a better diagnosis known generating model. That NCM methods were able models conflicts with Krantz Tversky's assertion that, general, direct axiom tests provide more powerful diagnostic test between alternative composition rules than does evaluation correspondence. For all dual-distributive most difficult recover, consistent past studies, additive is robust fitted models.