AN EMPIRICAL STUDY OF ATTITUDE-COMPONENT THEORY.

Sociologist Louis Guttman has conceived of two methods for placing individuals into relative positions along an attitude continuum. One method, that which is conventionally used, makes use of the observed marginals of selected items from an attitude universe to obtain cutting points or scale ranks. The second method and the one with which this study deals, is that of utilizing bending points of the regressions of higher principal components of scalable attitudes to obtain the various cutting points. Of the two methods, the latter, utilizing measures of principal components, is thought to be superior because of its quality of "invariance," that is, the cutting points are not dependent on the sample of items used. This paper reports the findings of a study which proposed to re-examine Guttman's "closure" as an empirical third component of scalable attitudes, and to introduce another psychological variable, suggested by Talcott Parsons, the sensitivity of the human personality to the attitudes of others, and test its empirical fit to the mathematical model of the third component.