Abstract
This study tests and contrasts the ability of multidimen sional scaling (MDS) and nonlinear mapping (NLM) in re covering complex data structures in attribute space, and aiding researchers and practitioners in making neighbor hood interpretations. The relative merits of both MDS and NLM for product positioning are explored and discussed. A formal comparison of the performance of NLM versus MDS is presented using both simulated and actual data. The re sults of this study provide direction as to the conditions under which a nonlinear mapping algorithm is preferable over MDS.
Get full access to this article
View all access options for this article.
