Abstract
The problem of pooling customer preference ratings within a conjoint analysis experiment has been addressed. A method based on the nonparametric combination of rankings has been proposed to compete with the usual method based on the arithmetic mean. This method is nonparametric with respect to the underlying dependence structure and so no dependence model must be assumed. The two methods have been compared using Spearman’s rank correlation coefficient and related test. Moreover, a further nonparametric testing method has been considered and proposed; this method takes both correlation and distance between ranks into account. By means of a simulation study it has been shown that the NPC Ranking method performs better than the arithmetic mean.
DOI
10.22237/jmasm/1130804340
Included in
Applied Statistics Commons, Social and Behavioral Sciences Commons, Statistical Theory Commons
