This article aims to alert researchers interested in the attitudes vs. action debate to a long-standing discontinuity between research and theory. Our review of 28 recent reports of research on the attitude-behavior relationship indicates a consistent neglect of theoretical formulations specifying the relevance of object-centrality, extremity, and intensity for understanding attitude-behavior consistency. Yet, bits and pieces of relevant data gleaned from these reports plus the results of our secondary analysis of data gathered in 1969-70 from some 19,000 members of the 1966 cohort of U.S. college freshmen point clearly to the conclusion that these variable properties of merit most careful consideration in attitude-behavior research. Indeed, they suggest that the failure of past empirical investigations to spport the assumption, central to the work of many scholars and practitioners, that attitudes are important keys to understanding behavior may be due, at least in part, to this gap between research and theory., That attitudes vary not only in direction (positive vs. negative) but also in extremity (degree of favorableness or unfavorableness) and intensity (strength of feeling), and that their objects may be more or less important, central, or egoinvolving to the individual are common themes in conceptual analyses of the construct (Scott). These variable properties of figure prominently in the theoretical literature dealing with the attitude-behavior relationship. (See D. Campbell; Converse; Himmelstrand; Inkso and Schopler; Katz; Katz and Stotland; Newcomb et al.; Rosenberg.) This being the case it is surprising that so little empirical research on the attitude-behavior relationship has given systematic attention to them.1 The aim of this article is to alert researchers interested in the attitudes vs.action debate to this long-standing discontinuity between theory and research and to point out how consideration of these variable properties of attitudes and their objects might help us understand why research has long indicated . . . that the person's verbal report of his has a rather low correlation with his actual behavior toward the object of the attitude (McGuire, 156). For good reason, such findings have been a matter of concern for some *Research for this article was supported in part by the American Council on Education. We are indebted to Alexander W. Astin, Director, Office of Research, ACE, and his staff for making the survey data analyzed herein available to us. We should also like to express our appreciation to the institutions and students whose cooperation made the survey possible, and to the anonymous reviewers whose careful critiques were both helpful and challenging. The conclusions, opinions, and other statements expressed are those of the writers and are not necessarily those of the persons and organizations named here.
Read full abstract
Mar 1, 1992
Dharma Kumar 
