The problem of different pairwise comparisons scales in the AHP framework

Show simple item record

dc.contributor.author Mazurek, Jiří
dc.date.accessioned 2019-09-30T11:23:36Z
dc.date.available 2019-09-30T11:23:36Z
dc.date.issued 2019
dc.identifier.issn 1211-555X (Print)
dc.identifier.issn 1804-8048 (Online)
dc.identifier.uri https://hdl.handle.net/10195/74244
dc.description.abstract In pairwise comparisons, several scales are used to compare objects. Perhaps the most known is Saaty’s fundamental scale for the AHP/ANP from 1 to 9 (with reciprocals), but other scales with 3, 5 or 10 items are also used in practice. Since the AHP/ANP is scale invariant, the following problem arises: for example, a preference of one object over other object expressed by the value 2 means something else for the scale from 1 to 3 (it expresses the medium preference) and the scale from 1 to 100 (in this case the preference is almost negligible). Therefore, the need of a normalization for pairwise comparison scales arises. The aim of the article is to propose a suitable transformation of a general linear scale for pairwise comparisons to a unit real interval that preserves several natural and desirable properties. en
dc.format p. 124 - 133
dc.language.iso en
dc.publisher Univerzita Pardubice cze
dc.relation.ispartof Scientific papers of the University of Pardubice. Series D, Faculty of Economics and Administration. 46/2019 en
dc.rights open access en
dc.subject AHP en
dc.subject normalization en
dc.subject pairwise comparisons en
dc.subject comparison scale en
dc.title The problem of different pairwise comparisons scales in the AHP framework en
dc.type Article en
dc.peerreviewed yes en
dc.publicationstatus published en
dc.subject.jel D81
dc.subject.jel D83


This item appears in the following Collection(s)

Show simple item record

Search DSpace


Advanced Search

Browse

My Account