Perspectives on attitude processes – the heuristic-systematic model, the elaboration likelihood model, the MODE model, the IAT, embodied attitudes, and neural attitudes – integrate traditional attitude research and new insights from social cognition and social neuroscience. As such, they represent at least a “second generation” approach (S. J. Sherman, 1987). They insist that attitude formation, change, and operation are not entirely rational, in contrast to traditional approaches that presupposed recipients who necessarily learned and considered the message arguments, if persuasion was successful, and who necessarily consciously considered their attitudes when influenced by them. This older rational view failed in at least three ways. First, people do not have to learn and recall a message to be persuaded by it; they may instead react to it online, resulting in an attitude based on their own responses but not on the message arguments as given. Second, people can be persuaded by more cognitively economical methods, whether persuasion heuristics or other peripheral routes. Third, people access their attitudes in relatively automatic ways, indicated by cognitive (response latency), physical (movement), and neural (activation) data.
More recent attitude theories have drawn heavily on social cognition theories and methods. Some new theories examine different types of attitude processing; some relatively thoughtful and analytic, others relatively rapid and automatic. The heuristic-systematic model posits that attitudes are often changed by cognitive shortcuts or heuristics in the form of simple persuasion rules that avert the need to process message content. Attitudes can also change by more systematic processing of the message arguments. Considerable research supports these ideas.
The elaboration likelihood model also proposes that attitudes can change by two routes, but the more automatic peripheral route includes a variety of superficial strategies, all of which share the feature of being relatively inattentive to message quality. In contrast, the central route to persuasion, typical of more motivated recipients, involves thorough consideration of the merits of the arguments given. In response to message arguments, people engage in more or less cognitive elaboration; that is, idiosyncratic responses pro or con. These, as measured by cognitive response analysis, predict attitude change via the central route. Traditional variables, such as characteristics of the communicator (credibility, expertise, and attractiveness), the message (its quality, repetition, difficulty, and length), and the audience (its outcome involvement, need for cognition, uncertainty orientation, and need to evaluate), all contribute to the degree and direction of cognitive elaboration as well as the resulting attitude change. This approach has yielded quantities of data illuminating old problems with new sophistication.
The MODE approach concentrates on the automatic activation of attitudes, based on the mere encounter with the attitude object. Attitudes activate more easily when recently or frequently activated in the past or when one has just reviewed attitude-relevant behavior. Low self-monitors, who orient to their attitudes, also seem to have more accessible attitudes. Easily activated attitudes more dramatically influence judgments about attitude-relevant information, resist contradiction, endure longer, and affect behavior more directly. The process seems to be relatively automatic. Clearly, some attitudes can activate immediately upon perception of the pertinent attitude object.
Implicit associations share many features of attitudes, predicting behavior. Associations between categories of an attitude object and positive or negative words show reliability and validity. Implicit associations best predict attitudes in contested or sensitive domains, whereas more traditional self-reports may best predict attitudes in more ordinary domains.
Other novel approaches to relatively spontaneous or implicit attitudes include embodied attitudes, correlated with arm flexion or extension, head nodding or shaking, and facial muscle manipulation. And neural patterns of activation implicate the amygdala in attitude intensity and other areas in valence.