Metric methods for analyzing partially ranked data (Q1074273)

Kendall's tau and Spearman's rho are two well known measures of the closeness of two rankings of n distinct items. Here the ranking is full. This monograph presents general methods for extending metrics to partially ranked data. Partially ranked data refers, for instance, to the situation in which each judge specifies only the first through K-th choices \((K<n)\). The problem is one of extending metrics on the permutation group to metrics on a coset space of the permutation group. The material presented in this monograph is divided into six chapters, including an introductory chapter. Chapters II, III and IV deal with metrics on fully ranked and partially ranked data. Chapter V gives the distributional properties of the metrics. Chapter VI is an important part of the monograph, concerning data analytic applications of the metric. The presentation of the material is fairly lucid.