Reports
Consequences of discrepant explicit and implicit attitudes: Cognitive dissonance and increased information processing

https://doi.org/10.1016/j.jesp.2008.07.006Get rights and content

Abstract

Attitudes research has shown that evaluations assessed directly (explicit attitudes) and indirectly (implicit attitudes) can diverge for many reasons. However, only recently has work begun to examine the phenomenology of experiencing discrepant explicit and implicit attitudes, and a number of important questions remain unanswered. What are the consequences of explicit–implicit attitude discrepancies on information processing? What psychological states accompany these discrepancies, and can they account for behavior? In two experiments, the current work examined whether dissonance-related discomfort results from discrepant explicit and implicit attitudes and considered its role in directing subsequent information processing. Dissonance and additional information processing were observed in experimental conditions where explicit and implicit attitudes diverged (and increased dissonance-related discomfort accounted for greater information processing; Experiment 1), but they were eliminated by a manipulation that reduced dissonance (i.e., self-affirmation; Experiment 2). The role of cognitive dissonance in explicit–implicit attitude inconsistencies and information processing is discussed.

Introduction

At times our feelings about people or objects are conflicted. That is, our explicit attitudes (evaluations that people can report and for which expression can be controlled) and implicit attitudes (evaluations for which people may not initially have conscious access and for which activation cannot be controlled) seem discrepant. For instance, one might dislike a co-worker despite the complete inability to articulate anything other than positive details about the person or be drawn to junk food despite its expense and fat content.

Explicit and implicit attitudes can diverge for a number of reasons, including self-presentational concerns (Olson, Fazio, & Hermann, 2007), quick implicit (Karpinski & Hilton, 2001) or explicit (Rydell & McConnell, 2006) attitude change, conflicting evaluations of individuated behaviors and social group memberships (McConnell, Rydell, Strain, & Mackie, 2008), extra-personal associations (Han, Olson, & Fazio, 2006), or exposure to inconsistently valenced subliminal primes and behavioral information (Rydell, McConnell, Mackie, & Strain, 2006). Although now documented extensively, little research has examined the consequences of explicit–implicit attitude discrepancies in terms of their phenomenology and their impact on information processing.

The only research examining the psychological consequences of divergent explicit–implicit attitudes or beliefs has shown that increased discrepancies lead to greater implicit ambivalence (a stronger association between the attitude object and doubt in memory) and increased information processing of attitude relevant information (Petty, Tormala, Briñol, & Jarvis, 2006). This research showed that once attitudes formed, they were not completely replaced when attitudes changed because increased implicit ambivalence accompanied attitude change. Briñol, Petty, and Wheeler (2006) showed that the greater the discrepancy between standardized measures of explicit and implicit self-beliefs (e.g., one’s own shyness), the more extensive processing of persuasive messages related to the domain of discrepancy (e.g., arguments favoring shyness). As explicit–implicit discrepancies increased, people were motivated to carefully consider subsequently presented, relevant information. Yet, why does this outcome occur? What phenomenology is driving this increased information processing?

Although there are no data directly addressing this question, Petty, Briñol, and colleagues suggested that increased explicit–implicit attitude discrepancies lead to implicit ambivalence, which people attempt to reduce by elaborating on subsequent information about the attitude object. Indeed, research on explicit and implicit attitude ambivalence has shown that when attitudes are ambivalent, people engage in more detailed processing of subsequently presented attitude-relevant information (e.g., Bell and Esses, 2002, Jonas et al., 1997, Petty et al., 2006).

In the current work, we examined whether holding discrepant explicit and implicit evaluations produces cognitive dissonance, which might in turn affect social information processing. It is clear that when people hold inconsistent cognitions, these discrepancies elicit feelings of psychological tension or discomfort (e.g., Aronson, 1992, Festinger, 1957). And, in response to dissonance-induced discomfort, people may attempt to reduce these feelings with responses ranging from justifying their beliefs (Aronson, 1997, Cooper and Fazio, 1984) to engaging in self-affirmation (Steele, 1988). Because explicit–implicit attitude discrepancies represent valence inconsistent cognitions (i.e., evaluations) about an attitude object, we explored whether they would lead to feelings of dissonance-induced discomfort (Olson & Fazio, 2007). Thus, as explicit–implicit attitude discrepancies increase, greater dissonance should be aroused and, as an attempt to reduce dissonance, increased information processing of attitude relevant information observed.

The ambivalence and dissonance accounts make similar predictions for attitude discrepancies and information processing. However, it is clear that dissonance and ambivalence are not isomorphic constructs (Maio, Esses, & Bell, 2000). For the current concerns, discomfort is not a necessary aspect of ambivalence (Newby-Clark, McGregor, & Zanna, 2002), whereas dissonance is always uncomfortable (Cooper & Fazio, 1984). Because discomfort is not necessary for ambivalence, to the extent that discomfort is necessary for explaining the relation between increased explicit–implicit discrepancies and increased information processing, then a dissonance account is given relatively more credence than an ambivalence account. Moreover, if the introduction of a manipulation known to undercut dissonance affects both dissonance-induced discomfort and additional information processing but has no impact on ambivalence, a dissonance account would be further supported.

Therefore, we examined the mechanisms by which greater explicit–implicit attitude discrepancies produce increased information processing. We suggest that explicit–implicit attitude discrepancies induce dissonance (discomfort) and, when subsequent information is available about the attitude object, people will attend to and elaborate on this information to reduce dissonance arousal. Also, manipulations that neutralize dissonance (e.g., self-affirmation; Steele, 1988) should reduce information processing about an attitude object for which discrepant explicit and implicit evaluations are accessible because they should eliminate the dissonance driving information processing.

In the current work, we adopted a technique developed in our lab to establish conflicting implicit and explicit attitudes toward the same object (Rydell et al., 2006). We demonstrated that explicit attitudes can form in response to consciously available information whereas implicit attitudes can form in response to the valence of subliminally-presented primes. When participants were presented with a series of trials in which a target person (“Bob”) was preceded by a subliminal prime (either positive or negative in valence) and who was described in a sentence as having performed a particular behavior (the valence of which was always opposite of the subliminal prime), implicit attitudes toward Bob reflected the valence of the subliminal primes whereas explicit attitudes corresponded to the valence of the behaviors presented.

Section snippets

Experiment 1

We borrowed this technique in Experiment 1, crossing the valence of the subliminal primes (positive vs. negative) with the valence of the behaviors (positive vs. negative) to produce conditions where implicit and explicit attitudes toward Bob either were, or were not, inconsistent with each other. We predicted that when primes and behaviors were valence inconsistent (as opposed to consistent), explicit–implicit attitude discrepancies would increase as would feelings of discomfort (i.e.,

Experiment 2

The results of Experiment 1 provide initial support for a dissonance account in explaining the relation between increased explicit–implicit attitude discrepancies and increased information processing. Specifically, explicit–implicit discrepancies resulted in dissonance-based discomfort, which in turn uniquely accounted for additional elaboration that produced greater attitude change. However, some lingering questions remain regarding a dissonance interpretation. First, Experiment 1 relied on a

General discussion

These experiments showed that when explicit–implicit attitude discrepancies were greater, dissonance was aroused. This dissonance, in turn, induced greater information processing of attitude object relevant information. The results are consistent with the idea that implicit–explicit attitude discrepancies can create dissonance that people attempt to reduce through learning more about the attitude object. Although this finding is novel, future research should examine the implications of

Conclusions

Understanding discrepancies between explicit and implicit attitudes has been at the forefront of contemporary social cognition research (Gawronski & Bodenhausen, 2006), and speaks to larger debates regarding: (a) the processes though which automatic and controlled cognition diverge (Chaiken & Trope, 1999), (b) the consequences of their divergence (Briñol et al., 2006), and (c) exactly how to conceptualize attitude representation (Fazio, 2007, Schwarz, 2007). From a functional perspective, it

Acknowledgments

This research was supported by NIMH Grant MH068279, NSF Grant BCS 0601148, and PHS Grant MH 63762.

References (29)

  • A.J. Elliot et al.

    On the motivational nature of cognitive dissonance: Dissonance as psychological discomfort

    Journal of Personality and Social Psychology

    (1994)
  • R.H. Fazio

    Attitudes as object-evaluation associations of varying strength

    Social Cognition

    (2007)
  • L. Festinger

    A theory of cognitive dissonance

    (1957)
  • B. Gawronski et al.

    Associative and propositional processes in evaluation: An integrative review of implicit and explicit attitude change

    Psychological Bulletin

    (2006)
  • Cited by (85)

    • A Systematic Study of Ambivalence and Well-Being in Romantic Relationships

      2024, Social Psychological and Personality Science
    View all citing articles on Scopus
    View full text