View article

Throughout its history in social psychology, the attitude construct has been defined in myriad ways. Core to most definitions has been that attitudes reflect evaluations of objects on a dimension ranging from positive to negative. Thus, researchers have characterized attitudes in terms of their valence and extremity. In practice, attitudes have been routinely represented by a single numerical index reflecting the position of an attitude object on an evaluative continuum. However, social scientists have long recognized that characterizing attitudes solely in terms of valence and extremity is insufficient to fully capture all relevant properties of an attitude. For example, in his seminal article on attitude measurement, Thurstone (1928) noted that attitudes are multifaceted and that attempting to describe them with a single numerical index is analogous to attempting to describe an object like a kitchen table with a single numerical index. Other early attitude researchers also noted a variety of relevant attitudinal properties. For example, early advocates of the tripartite perspective proposed that evaluative responses could be classified into the categories of affect, behavior, and cognition (eg, Katz & Stotland, 1959; Rosenberg & Hovland, 1960; Smith, 1947). Other scholars distinguished among the underlying functions a global evaluation might serve (eg, Katz, 1960; Katz & Stotland, 1959; Smith, Bruner, & White, 1956). And still others noted that evaluations might vary in the amount of information on which they were based (eg, Rosenberg & Abelson, 1960) and the extent to which they were linked to other attitudes (eg, Converse, 1964). Thus, social scientists have …