Individual preference scales within a multidimensional "similarities" space.

Steinheiser, Frederick H.

doi:10.1037/h0029942

Cited by 8 publications

(3 citation statements)

References 9 publications

(5 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The judgment is distinct from the judgment of preference. Preference can be defined as a forced choice between at least two entities (Slater, 1960;Steinheiser, 1970) rather than judgment of the desirable qualities of each entity independently. In the latter case it is possible for the individual to like all objects within a set; all can be given therefore a positive response rather than only one.…”

Section: Instrumentsmentioning

confidence: 99%

Liking and Disliking Household Objects: An Empirical Study of Value

Boyd

Allen²

1981

Home Economics Research Journal

View full text Add to dashboard Cite

The judgment of like/dislike may be an initial judgment in a hierarchy of criteria used to assign value to everyday objects. The finding implies that this subjective judgment, in turn, constrains succeeding value judgments, such as the perception of an object's utility and stylish ness. Forty‐one subjects responded to two instruments: pair comparisons of 16 objects on a nine‐point scale of similarity of value, and judgment of like/dislike of the same objects. The matrix of similarities was subjected to cluster analysis, producing a dendrogram for each indi vidual. Each dendrogram was compared to all others using an objective criterion. The data matrix of comparisons which resulted was subjected to principal component analysis. Like/dislike decisions were compared with both the dendrograms of each individual and the scatter diagrams summarizing the group. The judgment of like/dislike is discussed in relation to a conceptual framework based on the concepts of inherent and instrumental value.

show abstract

Section: Instrumentsmentioning

confidence: 99%

Liking and Disliking Household Objects: An Empirical Study of Value

Boyd

Allen²

1981

Home Economics Research Journal

View full text Add to dashboard Cite

show abstract

“…Garner and Clement reported this same reversal in goodness ratings for these two patterns; thus, the only departure from prediction on the basis of equivalence set size is, in fact, predictable on the basis of their data. Steinheiser's (1970) results suggested the possibility that a composite pattern space might accurately reflect pattern preferences, even though the dimensions of preference varied between subjects. Since he used abstract patterns and an undirected criterion of preference, that finding was not alarming, but in an experiment designed to investigate pattern goodness preference rather than simply pattern preference, it is imperative that patterns in the "good" regions of the MDS configuration actually be those most preferred by individual subjects and not just those preferred in the pooled data.…”

Section: Resultsmentioning

confidence: 92%

Pattern goodness and redundancy revisited: Multidimensional scaling and hierarchical clustering analyses

Glushko

1975

Perception & Psychophysics

View full text Add to dashboard Cite

“…The most common study of this type related some sort of liking or prefer ence measure to the dimensions. Most studies have found that the degree of liking increases the further one goes in some particular direction in the space, but some (35,189) find cases more consistent with there being a preferred locus in the space, with liking decreasing as a function of distance from this place. One study (80) of liking for television shows failed to find close relations, however.…”

Section: Validity Of Mdsmentioning

confidence: 99%

Scaling

Cliff¹

1973

Annu. Rev. Psychol.

View full text Add to dashboard Cite

The premise of the following review is that scaling or measurement is not simply the assignment of numbers to events according to rule. On the contrary, it is rather that meaningful scaling only takes place when several ingredients exist. The fi rst of these is a set of data relations which are at least fairly consistent and moderately numerous. A second requirement is the existence of a mathematical model whose variables can be matched with the empirical conditions and the relations among them. This, together with solution procedures, provides a basis for solving for numerical values of the variables in the model, and it is these numerical values attached to empirical conditions which are the "scales." More over, unless the scales have a use external to the immediate set of data from which they have been derived, the scales serve no real function except perhaps as one form of verifi cation of the model. Correspondingly, this review concentrates on contributions to (a) describing the conditions which data relations must satisfy in order to provide scales (measurement theory); (b) models which may be used for scaling; (c) methods for the solution of models; (d) the demonstration of external validity for the scales so derived.Instances of the application of scaling methodology in a straightforward man ner are not noted as such. Testing and factor analysis are not included here be cause they are covered separately, although in their more sophisticated applica tions they fi t well into our paradigm.The scope of the survey on which this review is based was intended to be rather wide, covering most of the major English language sources (the amount of litera ture in other languages is signifi cant and growing rapidly), but it cannot be claimed to be complete. Space restrictions led to several compromises with full coverage. With a few exceptions, a decision was made to limit, the literature cited in the review to materials already published in regular journals and books. Some references are noted only tersely, sometimes at grave injustice to their importance, and in some cases multiple references by the same author on related topics are omitted. The reviewer regrets all these inadvertant and deliberate omissions, and hopes that the authors involved will be tolerant of his choices among injustices.Much important literature exists only in technical report or other mimeographed form, an undesirable state of affairs.

show abstract

Individual preference scales within a multidimensional "similarities" space.

Abstract: Grateful thanks are expressed to M. G. Grecnberg and Win. N. Dcmber for helpful comments and suggestions, and to the Procter and Gamble Company for use of its computing facilities. 2 Requests for reprints should be sent to Frederick 11.

Cited by 8 publications

References 9 publications

Liking and Disliking Household Objects: An Empirical Study of Value

Liking and Disliking Household Objects: An Empirical Study of Value

Pattern goodness and redundancy revisited: Multidimensional scaling and hierarchical clustering analyses

Scaling

Contact Info

Product

Resources

About