On the self-definitional nature of attitudes: An affective-motivational view of attitude resistance

Attitudes can sometimes be extremely resistant to change even in the face of information that disconfirms beliefs relevant to the attitude. A recent debate has emerged regarding the nature of this attitude resistance. Whereas some have favored a purely cognitive model of attitude resistance, others have argued that the attitude is maintained by affective-motivational mechanisms. The current research tested two hypotheses derived from an affective-motivational account of attitude resistance whereby the attitude becomes closely linked to the self-concept and is thus defended in the same biased manner that people defend against unfavorable feedback about the self. The current research focused on attitudes toward homosexuality and the stereotypes associated with homosexuals. The first hypothesis (affirmational resources hypothesis) is that increased self-affirmational resources would result in less motivation to defend the attitude and less bias in evaluating attitude-relevant information, while decreased self-affirmational resources would result in more motivation to defend the attitude and more bias in evaluating the attitude-relevant information. Study 1 tested this hypothesis by manipulating state self-esteem through favorable or unfavorable personality trait feedback before presenting mixed attitude-relevant information. Study 2 tested the hypothesis by simply measuring trait self-esteem before presenting the attitude-relevant information. Study 3 tested the hypothesis using a manipulation designed to reaffirm the self-image after reading attitude-disconfirming information. The second hypothesis (compensatory inflation of unrelated beliefs hypothesis) is that, when presented with information disconfirming one specific attitude-relevant belief, people may defend their overarching attitudes by inflating the strength of other attitude-relevant beliefs. Study 3 tested this hypothesis by measuring other attitude-relevant beliefs both before and after presentation of belief-disconfirming information. Neither hypothesis was supported. Study 1 was plagued by a manipulation that did not have the intended effect. Studies 2 and 3 did not support the affirmational resources hypothesis, although an interesting asymmetry between those high and those low in prejudice toward homosexuals tended to emerge. Study 3 found very limited support for the compensatory inflation of unrelated beliefs hypothesis. The discussion focuses on how the results inform the current controversy regarding the nature of attitude resistance.