The Ordering ofnItems Assigned tokRank Categories by Votes ofmIndividuals

Abstract

A method is proposed for arranging in order of acceptability a set of n items (research projects, candidates for positions, articles used, viewed, or smelled) on the basis of judgments of m judges. Unlike previous methods, this permits each judge independently to group in any manner those items tied for his favor, and to omit voting on any items concerning which he cannot decide. The sampling probabilities for estimating accuracy are not here considered.