scispace - formally typeset
Search or ask a question

Showing papers on "Scalability published in 1963"


Journal ArticleDOI
TL;DR: This model postulates the presence of a unidimensional attribute underlying both a set of items and subjects’ responses to this set, such that a subject will respond to (or pass) all items falling below his own position on this assumed dimension, and no items located above this point.
Abstract: CONSIDERABLE attention has been given in recent years to the quantitative treatment of scalogram data, based on Guttman’s cumulative model of scalogram analysis (cf. White & Saltz, 1957). This model postulates the presence of a unidimensional attribute underlying both a set of items and subjects’ responses to this set, such that a subject will respond to (or pass) all items falling below his own position on this assumed dimension, and no items located above this point. As Torgerson (1958) has pointed out, this model has a logical counterpart in the non-cumulative, or point model, in which a similar unidimensional attribute underlying both stimuli and responses is postulated, but v i th the subject responding only to those items lying a t or close to his own position on the scale. A good example of a scale fitting this model is an adaptation of a Thurstone attitude scale, in which, from among a set of items presumably sampling different points along a single dimension, a subject checks those with which he agrees most closely. I n spite of its direct relevance to the measurement of attitudes, however, this model has not been subjected to mathematical treatment, with the exception of an unpublished paper by hlosteller (1949). I n particular, the problem of the measurement of scalability

15 citations