Special Session Summary Judgment Correction: Antecedents, Consequences, and Explanations

Prashant Malaviya, INSEAD
[ to cite ]:
Prashant Malaviya (2001) ,"Special Session Summary Judgment Correction: Antecedents, Consequences, and Explanations", in NA - Advances in Consumer Research Volume 28, eds. Mary C. Gilly and Joan Meyers-Levy, Valdosta, GA : Association for Consumer Research, Pages: 375-377.

Advances in Consumer Research Volume 28, 2001     Pages 375-377



Prashant Malaviya, INSEAD

Judgment correction refers to the process by which people change prior judgments to incorporate new information. The success of several marketing actions, such as re-positioning and erasing effects of negative publicity, hinges on consumers correcting prior judgments. Consumers could also initiate judgment correction of their own accord because they became aware of a bias in prior judgments, or they might engage in correction when confronted with new, may be inconsistent, information. While a mountain of research has examined how judgments are formed, less attention has been paid to judgment correction (Johar and Simmons, 2000). Thus, for both practical and theoretical reasons, the issue of judgment correction should be of interest to consumer researchers.

The first presentation, by Wegener, briefly reviewed the bias correction literature and offered evidence in support of the Flexible-Correction Model (FCM; Wegener and Petty 1997). A key feature of this model is that people correct prior judgments by invoking their "nanve theories" about how their judgments became biased (in the consumer psychology context, this could be the notion of a "schemer schema"; Friestad and Wright, 1994). Another feature of the FCM is that, unlike other accounts of correction, which hold that correction attempts lead to contrast effects, the FCM predicts that correction could result in both assimilation and contrast outcomes. Another key postulate of the FCM is that correction is a resource demanding task, which is dependent on the respondent’s motivation and ability to engage in bias correction.

The second presentation, by Johar and Sengupta, examined correction processes when consumers encounter new information that is incnsistent with their prior attitudes. This research examined conditions in which inconsistent information would lead to attitude correction and the effect of such correction on attitude strength, which is an important predictor of subsequent behavior. Findings suggested that correction could result in attitudes that are either stronger or weaker than prior attitudes, depending on whether the correction process (i.e., the resolution of inconsistent information) was successful or not.

Building on theorizing presented by Gilbert (1991), in the third presentation, Jung, Malaviya and Sternthal examined how negated product attributes (such as "not difficult to use") are processed. The findings suggest that the processing of negated information involves two steps: first, people elaborate on the affirmative message ("difficult to use") and then correct it to incorporate the negation ("not"). This process is resource demanding and produces the interesting outcome that when adequate cognitive resources are available product judgments reflect the negation, but when only moderate resources are available judgments are the opposite of what is implied by the message, presumably due to a failure to incorporate the negation.

The discussion leader of the session, Brendl, raised important issues about judgment correction that need additional research. One observation was that while the FCM assumes awareness of the bias to be a pre-condition for correction to occur, it might be possible that under some conditions correction occurs without such awareness (e.g., Schwarz and Clore 1983). Another issue was raised about the outcome of judgment correction, where it was suggested that correction could either result in the "overwriting" of the prior judgment, or lead to the formation of a second, new judgment (e.g., Wilson, Lindsey and Schooler 2000).


Friestad, M and P. Wright (1994), "The Persuasion Knowledge Model: How People Cope with Persuasion Attempts," Journal of Consumer Research, 21 (1), 1-31.

Gilbert, D. T. (1991), "How Mental Systems Believe," American Psychologist, 46(2), 107-119.

Johar, G. V. and C. J. Simmons (2000), "The Use of Concurrent Disclosures to Correct Invalid Inferences," Journal of Consumer Research, 26(4), 307-322.

Schwarz, N., and Clore, G. L. (1983), "Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states," Journal of Personality and Social Psychology, 45(3), 513-523.

Wegener, D. M. and R. E. Petty (1997), "The flexible correction model: The role of naive theories of bias in bias correction," in M. P. Zanna (Ed.), Advances In Experimental Social Psychology, 29, 141-208, Mahwah, NJ: Erlbaum.

Wilson, T. D., Lindsey, S., and Schooler, T. Y. (2000), "A model of dual attitudes," Psychological Review, 107(1), 101-126.




Duane T. Wegener, Purdue University

Please refer to full text included in this publication.



Gita Venkataramani Johar and Jaideep Sengupta

The entire set of responses that is associated with an attitude object is referred to as its internal structure. Previous research has found that inconsistency in structural elements (e.g., evaluative-cognitive inconsistency) leads to lower attitude strength, as manifested in lowered attitude accessibility (Bargh et al. 1992) and attitude-behavior link (e.g., Chaiken, Pomerantz, and Giner-Sorolla 1995; Norman 1975; Rosenberg 1960). Notably, these findings all relate the holding of inconsistent attitudes to attitude strength. However, research on the immediate consequences of incorporating inconsistencies is sparse. The present research aims to fill this gap in the literature by examining how the incorporation of inconsistencies affects the strength of the corrected, or updated, attitude.

Intuitively, it appears that incorporating inconsistencies should yield strength results similar to those produced by holding inconsistenciesBi.e., a lowering of attitude strengthBbecause of the response competition between inconsistent attitudinal elements. On the other hand, the introduction of inconsistencies may sometimes result in activation and elaboration of existing links, increasing the strength of the corrected attitude. Thus, it seems possible that the incorporation of inconsistent information in an existing attitude structure can lead to both strengthening and weakening effects. The current research seeks to isolate both of these effects. Because these two opposing consequences can potentially mask each other, we examine a context which is conducive to documenting a pure strengthening effect (i.e., one where rehearsal of existing links occurs, but not creation of a competing link), and also a context which enables us to isolate a pure weakening effect (i.e., a context in which a competing link is set up, but no activation of existing links occurs).

One particularly interesting context that is conducive to studying the strengthening effect is that of attitude dissimulationBi.e., lying about one’s true attitude. People often lie about their attitudes in their daily lives (e.g., white lies, lying for social desirability reasons).

Lying about existing attitudes represents a particularly suitable context for studying the strengthening effects of inconsistency incorporation because, while such dissimulation involves the rehearsal of existing links, it may not result in the creation of an actual link between the attitude object and the "false" attitude. Based on this reasoning, the first two studies of the current research use the dissimulation context (lying about product attitudes) to seek to provide process evidence for the hypothesis that inconsistency incorporation produces strengthening via activation of existing links. Such evidence is provided in two different ways. First, we examine the moderating effects of initial attitude strength on the strengthening impact of repeated dissimulation about product attitudes. For both strongly held and weakly held product attitudes, participants in our experiments were repeatedly asked to either express the opposite of their true attitudes (dissimulation) or their true attitudes (truthful expression). The effects of these manipulations were compared to a control, wherein participants did not express any attitudes. Each successive dissimulation trial should involve activation and elaboration of the true attitude, which is highly accessible on exposure to the attitude object, for strong attitudes. Such repeated activation and elaboration is less likely for weak attitudes. Thus, repeated dissimulation and repeated truthful expression (compared to the control) should result in equivalent amounts of attitude strengthening for strong attitudes but not weak attitudes (cf. Maio and Olson 1995). Support for this hypothesisBwhich relies on the premise that dissimulation leads to strengthening only to the extent to which it involves actvation of existing linksBis obtained in both our experiments, using two related measures of attitude strengthBattitude accessibility and the attitude-behavior link. Further support for our hypothesized process comes from our finding that the difference between strong and weak attitudes disappears if participants are forced to activate the true attitude on each dissimulation trial.

Experiment 3 investigates a context that allows for a demonstration of the weakening effects of inconsistency incorporation. While evaluating a new product, people may first be exposed to only positive information about the product (e.g., through advertising). Later, they may also be exposed to negative product information, from sources such as word-of-mouth, independent consumer agencies, etc. Such a scenario is likely to create attitudinal ambivalence, and result in a weakening effect (as compared to a condition in which the subsequent evaluative information is also positively valenced) because of response competition between the inconsistent elements in the attitude structure. Apart from proposing that inconsistent information in such a context should lead to greater ambivalence (and weakening) than consistent information, we also investigate a boundary condition for this effectBinformation accessibility. We hypothesize that exposure to inconsistent information is likely to result in a process of elaborative reconciliation when earlier information is more accessible (vs. less accessible). Such elaboration will prevent a weakening effect, and can in fact produce a strengthening effect (as compared to a condition where only consistent information is provided). However, if earlier information is not accessible at the time of exposure to the inconsistent information, elaboration is less likely to occur. In such a case, a weakening effect (as compared to the consistent information condition) should be observed because of heightened attitude ambivalence. Results from a 2 (Accessibility of initial information: high/low) * 2 (Valence of later information: consistent/inconsistent) study, with the attitude-behavior link as the indicator of attitude strength, are strongly supportive of our predictions.

In sum, our research investigates the corrective impact of incorporating inconsistent links, in terms of its effects on attitude strength. Apart from providing some interesting insights into the specific contexts studied (e.g., the effects of dissimulation), these findings further our understanding of the relationship between attitude structure and attitude strength. In particular, we provide support for the processes that underlie opposing effects of inconsistency incorporation. If such incorporation leads to a rehearsal of existing links, a strengthening effect is likely to be obtained; however, in the absence of such rehearsal, the creation of an inconsistent link in the attitude structure should yield a weakening effect.


Bargh, John A., Shelly Chaiken, Rajen Govender and Felicia Pratto (1992), "The Generality of the Automatic Attitude Activation Effect," Journal of Personality and Social Psychology, 62, 893-912.

Chaiken, Shelly, Eva M. Pomerantz, and Roger Giner-Sorolla (1995), "Structural Consistency and Attitude Strength," in Richard E. Petty and Jon A. Krosnick (eds.), Attitude Strength: Antecedents and Consequences, 387-412, Mahwah, NJ: Lawrence Erlbaum.

Maio, Gregory R. and James M. Olson (1995), "The Effect of Attitude Dissimulation on Attitude Accessibility," Social Cognition, 13(2), 127-144.

Norman, R. (1975), "Affective-Cognitive Consistency, Attitudes, Conformity and Behavior," Journal of Personality and Social Psychology, 32, 83-91.

Rosenberg, Milton J. (1960), "Cognitive Reorganization in Response to the Hypnotic Reversal of Attitudinal Affect," Journal of Personality, 28, 39-63.



Susan Jung, Prashant Malaviya and Brian Sternthal

In this paper we examine how consumers process negated product attribute information. For example, the attribute "not difficult to use" is a negated form of presenting the affirmative "easy to use". While on the surface both forms of presentation appear to convey the same information, research in semiotics suggests that affirmation and negation communicate distinct meanings (Pinson 1998). In psychology, research indicates that when people are offered possible gains versus non-losses, they invoke different goals (Higgins 1997), experience different affective reactions (Brendl, Higgins and Lemm 1995), differ in their recall and judgment (Tykocinski, Higgins and Chaiken 1994), and form different preferences and choices (Tversky and Kahneman 1981). These findings highlight that negated and affirmative presentation of information differ in how people process them and integrate them into judgment.

Although most advertising copy employs the affirmative form of presentation, advertisers sometimes present negated information about the brand. Negated attributes are included in so called "two-sided" advertising messages, when advertisers acknowledge some apparent product weakness, which of course is overshadowed by the positive features of the product that are the focus of the advertising message. Negated attributes are also included in some forms of comparative advertising. Finally, negated attributes are sometimes included in advertising copy to establish a departure from the past to communicate a point of difference (e.g., "not your father’s Oldsmobile"). Thus, advertising messages often speak about something that the product will "not" deliver, presumably because the benefit from the product is obtained only when the particular attribute is absent. How do message recipients process negated product information and form judgments of the product?

Theorizing proposed by Gilbert (1991) seems appropriate for understanding how people process a negated message. According to this thesis, processing of negated information ("not difficult to use") follows a two-step process. Message recipients first encode the affirmation ("difficult") and then correct this initial encoding to incorporate the negation ("not"). In contrast, information that is presented with affirmation ("easy to use") would require only one step to process.

An implication of a two-step process is that under conditions of resource constraint, such as when there are distractions to message processing, elaboration of a negated ad message could breakdown at either of the two stages of processing. Under some conditions resources might be so limited that the threshold for processing even the affirmative portion of the information is not crossed and consequently, no aspect of the negated product attribute would be processed. In this situation, the negated product attribute should have no impact on the product’s judgment.

In other conditions where more moderate levels of resources become available (e.g., fewer distractions are present during message processing), message recipients might be able to process the affirmative portion of the message, although the available resources might not be adequate for the next step of correcting for and incorporating the negation. Such processing would produce different judgment outcomes. Specifically, if the available resources are sufficient to process affirmation ("difficult") but not for correcting negation ("not"), judgments should be the opposite of what is implied by the message content.

Finally, in yet other conditions the available resources could be adequate to process the entire message, including the negation of the product attribute. In such conditions, which may arise under minimal distractions to message processing, the available resources might be sufficient for processing the affirmative information, as well as for correcting for negation. In these circumstances, product judgments should follow the implication of the message content ("not difficult to use").

The two-step model could be contrasted to a one-step model of processing negated information. Such a model would be based on the assumption that negation is processed simultaneously with affirmation, such that the part-phrase "not difficult" is processed together, rather than as two separate words or meanings (McKoon and Ratcliff 1989). This one-step model would also predict that negation requires more cognitive resources to process than affirmation, because compared to the phrase "easy to use", "not difficult to use" requires that we elaborate on more words and meanings.

Although both one-step and two-step models would predict that negation requires more resources to process than affirmation, depending on which model operates, different outcomes would be anticipated. A one-step model would be implicated if under conditions of minimal resource constraint the negated product attribute has an impact on product judgments, but under more severe constraints it does not have an effect. In contrast, a two-step model would predict that depending on the severity of resource constraint, three outcomes are possible: no effect of negated attribute, an effect that is opposite to the message content, or an effect that is consistent with the message claims.

We test these ideas in a study in which respondents were required to read an advertisement for toothpaste and later answer some questions related to the product. One attribute included in the ad was presented in a negated form ("not difficult to use dispenser"). All other attributes (ten in all) were kept constant across conditions. The focal dependent variable was product evaluation. Respondents were exposed to the target ad under conditions that varied in the extent of resource constraint for processing the information presented in the target advertisement. This was done so that predictions pertaining to the one- and two-factor account of the processing of negated information could be tested. These resource constraints were either naturally occurring individual differences ("need for cognition"), or were induced by experimental manipulations ("distraction" by misleading respondents about whether they would be asked to judge the target product or a competing brand; and "involvement" by asking respondents to assume either that they were evaluating the product to buy it for themselves or for someone else).

Consistent with Gilbert’s theorizing, results of this study offer support for the notion that processing of negated information follows a two-step process. Specifically, product judgments revealed three effects. In some conditions, the negated attribute had no effect on judgments, suggesting that respondents did not elaborate on this information. This outcome was observed for respondents who indicated low need for cognition. In other conditions, product judgments were just the opposite of what the message content implied, suggesting that respondents had failed to incorporate the negation in the message. This outcome was observed for respondents classified as high in need for cognition, but who were distracted and did not have high motivation to elaborate on the message. Finally, in some conditions, product judgments followed the intent of the message, suggesting that respondents had successfully processed the negated attribute. This outcome was observed for respondents high in need for cognition, and who were either not distracted or had a high degree of motivation to process the message. This pattern of outcomes is consistent ith the notion that negated product attributes are processed in two steps: elaboration of affirmation, followed by correction for negation.


Brendl, C.M., E. T. Higgins, and K. M. Lemm (1995), "Sensitivity to Varying Gains and Losses: The Role of Self-Discrepancy and Event Framing," Journal of Personality and Social Psychology, 69(6), 1028-1051.

Gilbert, D. T. (1991), "How Mental Systems Believe," American Psychologist, 46(2), 107-119.

Higgins, E. T. (1997), "Beyond Pleasure and Pain," American Psychologist, 52(12), 1280-1300.

McKoon, G. and R. Ratcliff (1989), "Assessing the Occurrence of Elaborative Inferences with Recognition: Compatibility Checking versus Compound Cue Theory," Journal of Memory and Language, 28, 547-563.

Pinson, C. (1998), "Semiotics of Marketing," Concise Encyclopedia of Pragmatics, ed. J. L. Mey, London: Pergamon Press, 538-544.

Tversky, A. and Kahneman, D. (1981), "The Framing of Decisions and the Psychology of Choice," Science, 211, 453-458.

Tykocinski, O., E. T. Higgins and S. Chaiken (1994), "Message Framing, Self-Discrepancies, and Yielding to Persuasive Messages: The Motivational Significance of Psychological Situations," Personality and Social Psychology Bulletin, 20, 107-115.



C. Miguel Brendl, INSEAD