Attitudes and Behavior 317situational demands. Kardes, Sanbonmatsu, Voss, and Fazio
(1986) found that the attitudes of low self-monitors were
more accessible than were the attitudes of high self-monitors,
perhaps because low self-monitors think about their attitudes
more often than do high self-monitors. As noted earlier, atti-
tudes high in accessibility predict behavior better than do
attitudes low in accessibility (e.g., Fazio & Williams, 1986).
Thus, there are probably multiple determinants of the differ-
ences in attitude-behavior consistency between low and high
self-monitors.
Private self-consciousness (Fenigstein, Scheier, & Buss,
1975) is another variable that has been examined in the
attitude-behavior consistency literature. Private self-con-
sciousness reflects the extent to which individuals are aware of
their internal states (e.g., moods, values, and attitudes); it
corresponds to stable individual differences in the state of
objective self-awareness or self-focused attention (Duval &
Wicklund, 1972). If awareness of one’s attitudes increases the
likelihood of attitude-consistent behavior (which seems
likely), then individuals who are high on this dimension
should exhibit stronger attitude-behavior correlations than
should those who are low in private self-consciousness. This
result has in fact been obtained (Scheier, Buss, & Buss, 1978;
Wicklund, 1982).
Composite Model of Attitude-Behavior Consistency
As the preceding sections have indicated, social psycholo-
gists have made significant advances in the understanding of
when and how attitudes predict behavior. Distinctions be-
tween types of attitudes, types of behaviors, and personality
subgroups have all helped to clarify the relation between
measures of attitudes and measures of behavior.
Numerous models have been proposed to account for
attitude-behavior consistency, such as the theory of rea-
soned action(Fishbein & Ajzen, 1975), the theory of planned
behavior(Ajzen, 1985), and the MODE (motivation and
opportunity as determinants of how attitudes influence be-
havior) model(Fazio, 1990). Based on a comprehensive
review of the literature and building on these previously
proposed models, Eagly and Chaiken (1993, 1998) devel-
oped a composite model of attitude-behavior consistency.
The model parallels the theory of reasoned action by propos-
ing that attitudes toward behaviors predict intentions, which
in turn predict behaviors. Working back from attitudes
toward behaviors, however, the model identifies five factors
that influence attitudes toward behaviors: habits (past behav-
ior), attitudes toward targets (especially the target of the be-
havior), utilitarian outcomes (rewards and punishments that
are expected to follow from engaging in the behavior),
normative outcomes (approval and disapproval from others
that are expected to follow from engaging in the behavior),
and self-identity outcomes (implications of engaging in the
behavior for the self-concept). Some of these factors are also
hypothesized to influence either intentions or behavior di-
rectly; for example, habits are proposed to affect behavior
directly (i.e., not via attitudes toward the behavior).
Eagly and Chaiken’s model is unique in its inclusion of
both attitudes toward behaviors and attitudes toward targets
as predictors of specific actions. The inclusion of habits is
also noteworthy; many researchers have found that past
behavior predicts future behavior even when attitudes and
norms are held constant (see Ouellette & Wood, 1998). Fi-
nally, the categorization of expected outcomes into utilitarian,
normative, and self-identity classes extends previous models.
Although this model has not yet been exposed to direct em-
pirical tests, it provides a heuristically useful framework for
future research.Applications to Social BehaviorWe have reviewed various theories of attitude-behavior con-
sistency and outlined the conditions under which strong rela-
tions between attitudes and behavior can be expected. The
title of this chapter is “Attitudes in Social Behavior,” so we
close with the consideration of some of the important social
behaviors to which the concept of attitude can be applied. In
each case, data support the hypothesis that attitudes facilitate
attitude-consistent behaviors.
For example, this Handbook contains several chapters that
rely heavily on attitudes to understand social behavior. The
chapter on prejudice, racism, and discrimination (see the
chapter by Dion in this volume) explores interpersonal and
intergroup settings in which negative attitudes toward an out-
group (prejudice) can cause conflict and violence. Prejudice
is one of the oldest topics in social psychology; it continues
to be a vibrant research area today—recent attention has ex-
panded to incorporate the consequences of being a target of
prejudice (e.g., impaired performance on intellectual tasks
caused by fear of confirming a stereotype; Spencer, Steele, &
Quinn, 1999; Steele & Aronson, 1995). Altruism (see the
chapter by Batson & Powell in this volume) is another do-
main in which attitudes are important. Positive evaluations of
potential recipients of assistance serve to motivate prosocial
actions (e.g., Goodstadt, 1971). Similarly, attitudes influence
individuals’ responses to situations involving justice consid-
erations (see the chapter by Montada in this volume). For
example, individuals are more likely to tolerate a situation in
which distributive or procedural justice was violated when
they have positive attitudes toward the responsible authority