Interval-valued scaling of successive categories

Title: Interval-valued scaling of successive categories
Creator: Miyano, Hisao; Beh, Eric J.
Relation: Advanced Studies in Behaviormetrics and Data Science: Essays in Honor of Akinori Okada p. 197-209
Relation: Behaviormetrics: Quantitative Approaches to Human Behavior 5
Publisher Link: http://dx.doi.org/10.1007/978-981-15-2700-5
Publisher: Springer
Resource Type: book chapter
Date: 2020
Description: Correspondence analysis represents the row and column categories of a contingency table as points in a low dimensional space, irrespective of whether the categories are successive or not. In this paper, a scaling method is considered for successive categories that are regarded as a series of boxes (intervals) or numbers defined on a line scale. By using this method, each category is represented as a region not as a point under the assumptions that (1) each category is represented by an interval for which the end points lie on the boundary of its adjacent category, and (2) the scale values of the category are uniformly distributed over the interval. It is shown that the proposed method has simple links to correspondence analysis and multiple correspondence analysis. The effectiveness of the method is confirmed by considering some examples.
Subject: contingency table; low dimensional space; scaling method; analysis
Identifier: http://hdl.handle.net/1959.13/1438380
Identifier: uon:40587
Identifier: ISBN:9789811526992
Language: eng

		Thumbnail	File	Description	Size	Format