PREDISPOSITIONS AND ATTITUDE CONSTRUCTION PROCESSES

先前资助的研究调查了内省的影响

Previously-funded research has investigated the effects of introspection (specifically, analyzing the reasons for one's attitudes) on attitude change, decisions, self-predictions, and attitude-behavior consistency. Considerable progress has been made in understanding why analyzing reasons causes attitude change, and why this change can have negative consequences for people. The goals of the present research are threefold: (l) To continue investigating introspections about attitudes, to understand further how analyzing reasons causes attitude change. The way in which the strength of an initial evaluation moderates the effects of analyzing reasons is of particular interest. (2) To broaden the research considerably, placing it in the context of a recent debate over a fundamental question: To what extent are attitudes a function of evaluations stored in memory, versus constructions based on information that is currently accessible? A model is presented specifying the conditions under which prior evaluations versus current constructions determine attitudes, which incorporates three main moderator variables: The strength of an initial attitude; the availability and accessibility of other information about the attitude object; and people's cognitive capacity and motivation to process this information. Several studies are proposed that explore the implications of this model for the effects of analyzing reasons on behavior and other attitude construction processes. (3) To test novel predictions generated by the model, specifically concerning the conditions under which prior evaluations are overridden versus changed. We predict that when people with moderately-accessible initial attitudes construct a new attitude (e.g., after analyzing reasons), the newly-constructed attitude "overrides" the predisposition, such that people have a dual representation of the attitude object: A predisposition that is expressed automatically on spontaneous measures of attitudes, and an alternative, constructed attitude that is expressed on more deliberative, controlled measures. Several of the proposed studies will investigate the causes and consequences of such dual attitudes. This research will move considerably beyond our prior work on the effects of analyzing reasons, in that the predictions apply to any situation that causes people with moderately-accessible attitudes to construct new attitudes, including changes in the social context, persuasive communications, and self-perception processes. These basic issues have implications for a number of applied areas in which it is difficult to change people's behavior due to the intransigence of their initial attitudes (e.g., getting people to engage in safer sex or to act in less prejudiced ways). The research will also have significant methodological implications for applied research, by showing that attempts to change attitudes may succeed on deliberative but not spontaneous measures of attitudes.