We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
WHAT WEIGHT SHOULD WEIGHTS HAVE IN INDIVIDUAL DIFFERENCES SCALING?
- Authors
Borg, Ingwer; Lingoes, James C.
- Abstract
The article focuses on the methods of individual differences scaling. Numerous models and algorithms have been proposed but relatively little evidence as to the validity of the produced representations is available so far. Here the authors reanalyze some data collected by Procrustrean INdividual DIfference Scaling (PINDIS). The results are compared with those produced by INDSCAL which is presently the most popular procedure. The validity and interpretability of the dimension weights must be evaluated against the amount of variance explained by admissible transformations. Clearly, if no substantial fit improvement can be achieved by differential weights, then one should rather assume unit weights. Only a small fraction of the potentially useful information provided by PINDIS has been reported. Not only is the average fit value interesting but also the individual coefficients upon which it is based. Although authors have not addressed ourselves to substantive issues in this article, the PINDIS analysis is clearly suggestive that more could be said about the data than is possible by a dimensional or cluster approach.
- Subjects
INDIVIDUAL differences; SCALING (Social sciences); SOCIAL science methodology; MATHEMATICAL models; DATA analysis; MODEL validation
- Publication
Quality & Quantity, 1978, Vol 12, Issue 3, p223
- ISSN
0033-5177
- Publication type
Article
- DOI
10.1007/BF00144137