Rankings as ordinal scale measurement results

Rankings as ordinal scale measurement results

Bibliographic Details
Parent link:Metrology and Measurement Systems
Vol. 13, iss. 1.— 2007.— P. 9-24
Main Author: Muravyov (Murav’ev) S. V. Sergey Vasilyevich
Summary:Title screen
Rankings (or preference relations, or weak orders) are sometimes considered to be non-empirical, nonobjective, low-informative and, in principle, are not worthy to be titled measurements. A purpose of the paper is to demonstrate that the measurement result on the ordinal scale should be an entire (consensus) ranking of n objects ranked by m properties (or experts, or voters) in order of preference and the ranking is one of points of the weak orders space. The consensus relation that would give an integrative characterization of the initial rankings is one of strict (linear) order relations, which, in some sense, is nearest to every of the initial rankings. A recursive branch and bound measurement procedure for finding the consensus relation is described. An approach to consensus relation uncertainty assessment is discussed
Published: 2007
Subjects:
труды учёных ТПУ
электронный ресурс
ordinal scale
weak order
consensus relation
recursive algorithm
порядковые шкалы
рекурсивные алгоритмы
Online Access:http://metrology.pg.gda.pl/full/2007/M&MS_2007_009.pdf
Format: Electronic Book Chapter
KOHA link:https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=668513

Similar Items

Similar Items