Congruous and Incongruous Processes in Attitude Evaluation: Response Mode and Behavioral Intention

ABSTRACT - Millar and Tesser (1986) propose that when attitude evaluation (i.e., cognitive vs. affective) and behavioral intention (i.e., instrumental vs. consumatory) are matched the attitude-behavior link is enhanced, whereas when mismatched the relationship suffers. The current study conceptually extends that model to attitude change, and uses a methodological technique that allows for an objective assessment of evaluation processes. Subjects either judged automobile alternatives (i.e., cognitive evaluation) or chose one (i.e., affective evaluation), and participated in either an instrumental or a consumatory behavioral condition. As anticipated, more attitude change occurred under conditions of a mismatch than a match. Implications for how consumer attitudes toward product alternatives change as a function of evaluative conditions are discussed.

Citation:

James R. Bailey (1995) ,"Congruous and Incongruous Processes in Attitude Evaluation: Response Mode and Behavioral Intention", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 217-221.

Advances in Consumer Research Volume 22, 1995      Pages217-221

CONGRUOUS AND INCONGRUOUS PROCESSES IN ATTITUDE EVALUATION: RESPONSE MODE AND BEHAVIORAL INTENTION

James R. Bailey, Rutgers University

ABSTRACT -

Millar and Tesser (1986) propose that when attitude evaluation (i.e., cognitive vs. affective) and behavioral intention (i.e., instrumental vs. consumatory) are matched the attitude-behavior link is enhanced, whereas when mismatched the relationship suffers. The current study conceptually extends that model to attitude change, and uses a methodological technique that allows for an objective assessment of evaluation processes. Subjects either judged automobile alternatives (i.e., cognitive evaluation) or chose one (i.e., affective evaluation), and participated in either an instrumental or a consumatory behavioral condition. As anticipated, more attitude change occurred under conditions of a mismatch than a match. Implications for how consumer attitudes toward product alternatives change as a function of evaluative conditions are discussed.

ATTITUDE CHANGE AND THE MATCH-MISMATCH HYPOTHESIS

Millar and Tesser (1986) have tested a model in which attitude evaluation and behavioral intention interact to determine attitude-behavior correspondence (AB). They reasoned that behaviors performed for instrumental purposes (e.g., the development of a skill or knowledge) are driven by the cognitive attitude component, and behaviors engaged in for consumatory purposes (e.g., preferences) are driven by the affective attitude component. Hence, when attitudes are cognitively evaluated and behaviors are instrumental, or when attitudes are affectively evaluated and behaviors are consumatory, the AB relationship is enhanced because the component emphasized and the type of behavior are commensurate. Conversely, when cognitive evaluations are paired with consumatory tasks or affective evaluations with instrumental tasks, the AB relationship suffers because the component emphasized and behavior type are incommensurate. The former two conditions represent a match between evaluation and behavioral intention, whereas the latter two represent a mismatch.

Although Millar and Tesser's (1986) model provides a compelling account of the relation of attitude evaluation to subsequent behavior and effectively integrates several contradictory studies, they did not test the most likely operative causal mechanism: attitude change. Wilson (e.g., Wilson, Dunn, Kraft and Lisle, 1989) has maintained that cognitive evaluation disrupts the AB link by changing attitudes. Using a repeated measures design, Wilson, Kraft and Dunn (1989) compared analyzing reasons for an attitudeCa cognitive focusCto a control group, where knowledge about the attitude object was included as a moderating variable. Results indicated that analytic evaluation did change attitudes, but only when subjects were unfamiliar with the attitude object.

Wilson, Kraft and Dunn's (1989) study represents the first direct test of the attitude change hypothesis, but it leaves a number of important questions unanswered. Their results support, at least provisionally, the idea that attitudes change as a function of evaluation, and are consistent with an emerging perspective in consumer research that mental representations are constructed by, among other things, task factors of the decision context (see Payne, Bettman and Johnson 1992 for a review; Upmeyer 1989). But because Wilson, Kraft and Dunn compared a cognitive focus only against a control condition, their study does not test how nonanalytic evaluation (i.e., affective focus) or behavioral intention (i.e., instrumental vs. consumatory) alters attitudes. The purpose of the current study, then, is to test the attitude change explanation within the complete theoretical framework of Millar and Tesser (1986). Attitude change should follow the same patterns as AB consistency: When attitude evaluation and behavioral intention are matched (cognitive-instrumental and affective-consumatory) there should be less attitude change than when mismatched (cognitive-consumatory and affective-instrumental).

Integrating Decision Research

One shortcoming of this line of research, though, has been a presumption that manipulations affect the nature and degree of evaluation differentially. Yet evidence for this claim is indirect, as these differences have only been inferred through outcome measures like coded written responses, thought listing or recall. One technique from the decision literature designed expressly to gauge deliberative process is the Information Display Board (IDB), which allows the careful charting of patterns of information acquisition, and shifts analytic focus from the residue of evaluative process (e.g., written responses) to process itself. Accordingly, the IDB appears optimal for representing attitude objects and observing the depth of evaluation directed toward them, and provides a methodological link between the decision making and attitude literatures (Upmeyer 1989).

Another beneficial meeting of attitude and decision research involves the manipulations themselves. Bailey and Billings (1994) and Billings and Scherer (1988) have evoked evaluative strategies similar to those evoked in the AB studies through response mode. Here, subjects who either chose one product alternative from many or judged each alternative represented on an IDB engaged in drastically different decision processes because of the different type of evaluation each induces. The choice mode is driven by the activation of preferences and a subsequent acquisition strategy geared toward identifying the alternative most harmonious with those preferences. In comparison, the judgment mode is compensatory in that features of each alternative are weighed and combined in relation to other alternatives as part of developing coherent representations. As a result, judgment is more analytical and thorough than choice.

The attitude focus manipulations employed by Millar and Tesser (1986) and the response mode manipulations employed by Bailey and Billings (1994) and Billings and Scherer (1988) are conceptually similar in important respects. Specifically, observing attitudes (an affective focus) reminds subjects of their attitudes, just as choice cues preferences, presumably driven by affective concerns (Zajonc and Markus 1982). These manipulations are similar in that subjects consider what they feel, not why they feel it, and result in an affective type of evaluation. Further, analyzing reasons for an attitude (a cognitive focus) and judgment both trigger a cognitive type of evaluation; the reasons focus by requesting an analysis of why, and the judgment mode by requiring subjects to render evaluative statements for each alternative.

Summary and Hypotheses

In summary, this study extends the research of Millar and Tesser (1986) and Wilson, Kraft and Dunn (1989) in three important ways. First, it directly examines the interactive effects of evaluation type and behavioral intention by measuring attitudes before and after manipulations. Second, it employs the IDB methodology to verify the evaluation elicited by these manipulations. Third, it proposes that the judgment and choice distinction operates in a parallel manner to the cognitive and affective distinction. Formally stated, the hypothesis was: Under conditions where evaluation type and behavioral intention are matched (i.e., judgment-instrumental and choice-consumatory) there will be less attitude change than under conditions of mismatch (i.e., judgment-consumatory and choice-instrumental).

As a manipulation check based on previous research using response mode and the IDB methodology (Bailey and Billings 1994; Billings and Scherer 1988), we predict that the judgment conditions should result in more thorough evaluation (i.e., amount of information uncovered) than the choice conditions.

METHODS

Subjects and Overview

Subjects were 43 female and 42 male undergraduates at Rutgers University who were given course extra-credit for participating in a consumer study on automobiles, and completed a questionnaire regarding automobile characteristics (i.e., the pre-test attitude measure). Upon arrival for testing subjects were presented with the target IDB randomly assigned to one of four conditions defined by a 2 x 2 (Response Mode: Judgment vs. Choice x Behavioral Intention: Instrumental vs. Consumatory) factorial design. Once the task was completed, manipulation checks and the post-test attitude questionnaire were administered.

Stimulus Material

The IDB's consisted of eight used automobiles listed vertically and six dimensions listed horizontally, with values for each dimension recorded on cards contained in envelopes on the corresponding locations on the board. The dimensions and values were as follows: (a) make: Chrysler, Ford, GM, Volkswagen, Toyota, Honda, Nissan, Subaru; (b) type: sports, compact, mid-sized, full-sized; (c) mileage: 1,000-15,000; 15,000-30,000; 30,000-45,000; 45,000-60,000 (d) price: $2,500-5,000; $5,000-7,500; $7,500-10,000; $10,000-above; (e) features: stereo system, digital display instruments, air conditioning, none; and (f) color: red, blue, black, brown, green, white, orange, silver (the values under mileage, price, and features were each presented twice to cover all eight alternatives). This matrix yielded 48 pieces of information, 6 per alternative. To obtain information, subjects removed the cards from the envelopes and placed the cards back on the board so that the information was displayed. The configuration was randomized after each session to control for order effects.

Attitude Questionnaire

Studies that concentrated on the AB link (e.g., Millar and Tesser 1986) used only one item to measure attitudes, as did Wilson Kraft and Dunn (1989, Study 2) in their pre- and post-test design. Herein a more complete approach was adopted, using 38 items and assessing two dimensions. The importance of each dimension (e.g., make, mileage) was rated on a scale ranging from extremely unimportant (1) and extremely important (7), and the desirability of specific characteristics (e.g., Ford, Toyota) was rated on a scale ranging from extremely undesirable (1) to extremely desirable (7).

Prior to statistical analysis, component and composite measures of attitude change were calculated from the pre- and post-test questionnaires. The component measure was computed as the average absolute amount of change in the importance ratings added to the average absolute amount of change in the desirability rating (both assessed from pre- to post-manipulation). This variable could assume values from 0 to 8.

The composite measure was computed as the difference between pre- and post-weighted attitude scores. These were calculated as the weighted average of the desirability ratings (D_i) for the pieces of information uncovered about an alternative during the task, where the weights were the importance ratings (I_i): SI_i D_i/I_i. For example, if four pieces of information about alternative A were uncovered, the score would be the weighted average of the four product scores, each product being an importance rating multiplied by a desirability rating. Separate scores were computed using the pre- and post-questionnaires, and the composite measure was computed as the average absolute difference between these weighted scores for each alternative, and could assume values from 0 to 25.

The rationale behind computing absolute amount of change in the component and composite measures was the absence of persuasive intent. Accordingly, change in either direction was considered. This was also the approach used by Wilson, Kraft and Dunn (1989).

Procedure

Once participants arrived at the laboratory (approximately one week later), the experimenter reiterated that this was a consumer study on automobiles, and in order to enhance experimental and mundane realism stressed the importance of consumer research and asked subjects to vividly image an automobile purchase. The IDB was then thoroughly explained, and participants were asked to practice on a sample IDB until comfortable with the task. When ready and after any questions were answered, participants were introduced to the test IDB, reminded of the nature of the study, and provided with one of two standardized sets of instructions to manipulate attitude focus and behavioral intention.

Choice versus judgment manipulation. Choice was manipulated by instructing subjects to choose one of the eight automobiles alternatives. Judgment was manipulated by instructing subjects to evaluate each alternative on a 7-point scale ranging from very high quality (1) to very low quality (7). These are the exact manipulations used in previous research on response mode (Bailey and Billings 1994; Billings and Scherer 1988).

Instrumental versus consumatory manipulation. The instrumental condition was manipulated by informing subjects that after completing the task, they would be given a test to measure their ability to discern automobile quality. Subjects in the consumatory condition were told that they would be given an unrelated test on social sensitivity following the task. These are the exact manipulations used by Millar and Tesser (1986).

RESULTS

Manipulation Checks

We used the amount of information searched (i.e., the total number of cards examined) as a meaningful variable sensitive to the thoroughness or depth of evaluation evoked by experimental conditions. Consistent with Bailey and Billings (1994) and Billings and Scherer (1988), subjects in the judgment condition searched more information (M=32.42) than subjects in the choice condition (M=22.02), F(1, 81)=28.05, p<.001, indicating the response mode manipulations were effective.

To test the effect of behavioral intention manipulations, responses on the post-task questionnaire indicated that subjects in the instrumental version expected to be tested on their ability to discern automobile quality to a greater degree (M=3.50) than did subjects in the consumatory version (M=2.60), F(1, 81)=5.84., p<.001. These results testify that the behavioral intention manipulations were effective.

FIGURE 1

COMPONENT ATTITUDE CHANGE AS A FUNCTION OF RESPONSE MODE AND BEHAVIORAL INTENTION

FIGURE 2

COMPOSITE ATTITUDE CHANGE AS A FUNCTION OF RESPONSE MODE AND BEHAVIORAL INTENTION

Attitude Change

Response mode and behavioral intention should interact to determine attitude change in an antagonistic pattern. Analysis of the component measure of attitude change indicated a significant interaction between response mode and behavioral intention, supporting this hypothesis, F(1, 81)=26.38, p<.001 (see Figure 1).

Analysis of the composite measure of attitude change yielded precisely the same pattern of results, F (1,81)=6.4, p<.05 (see Figure 2).

Additional Analysis

The possibility that the change reported above was due to polarization was tested by coding any change in extremity of evaluation (e.g., undesirable to extremely undesirable) as polarization. Consistent with the results of Wilson, Kraft and Dunn (1989), there were no effects of the condition on polarization (all Fs<1.2; ps<.05).

Analyses were also conducted to address the possibility of artifactual explanations. Wilson, Kraft and Dunn (1989) demonstrated that knowledge about an attitude object moderates the effects of a cognitive focus: Those unfamiliar with attitude objects were more affected by cognitive deliberations than those familiar with the objects. Hence, the post-experimental questionnaire queried subjects about experience assessing automobiles and their general level of interest about automobiles. Analyses indicated no significant main effects or interactions involving familiarity or interest (all Fs<1.4; ps<.05).

Finally, it is also possible that the amount of information uncovered on the IDB differentially affects attitudes and accounts for the interaction between response mode and behavioral intention. Hierarchical regression analyses were conducted on the component and composite measures of attitude change with amount of information searched (a continuous variable) entered as an additional independent variable along with response mode and behavioral intention. The interactions described previously were unchanged by this analysis, indicating that experimental conditions affected attitudes independent of amount of information searched.

DISCUSSION

This study demonstrated that the manner of evaluation and behavioral intention act in concert to influence attitudes. Although the framework put forth by Millar and Tesser (1986) was designed to explain attitude-behavior consistency, it appears to generalize to attitude change, and the current study is a more comprehensive test of the match-mismatch hypothesis than has been offered previously.

In those conditions where little attitude change occurred response mode elicited attitude evaluation that was congruous with the behavioral intention. The judgment-instrumental condition required an incorporation and organization of information into coherent wholes so that comparative judgments could be rendered about each alternative, and the expectation of a test of ability to discern automobile quality evokes a parallel process. Similarly, the requirements of the choice-consumatory condition are consonant; the individual preferences and desires cued by the instruction to choose among alternatives can be considered without concern for a test of discriminatory ability. In both cases, no attitude change occurs because the mental activity of the attitude focus and the behavior type are compatible.

By contrast, conditions resulting in a larger degree of attitude change represent a mismatch of evaluation and behavioral intention. In the choice-instrumental condition, subjects access their feelings regarding automobiles in order to choose, but their feelings are tempered by the simultaneous concern over an ability test. These incongruous purposes lead subjects to scrutinize feelings activated by the choice response mode in order to make them consistent with the concern over evaluation. The judgment-consumatory condition is more difficult to explain, but because the consumatory manipulation specifies no purpose, perhaps individual preferences were activated and subsequently reevaluated in the service of judgment. Attitude change occurs when evaluation and intention are mismatched because the mental activity involved with each is contradictory.

Two recent studies have tested hypotheses similar to those tested here. Millar and Millar (1990) and Edwards (1990) investigated attitude change as a function of attitude type (i.e., cognitive vs. affective) and means of persuasion (i.e., cognitive vs. affective). Like the results reported here, Millar and Millar find more attitude change when attitude type and means of persuasion were mismatched (e.g., cognitive and affective) than when they were matched (e.g., cognitive and cognitive). Edwards' results did not follow a mismatch-match pattern, but this may be due to subtle variations in procedure. There are differences among these studies, but all distinguish attitudes on a cognitive-affective dimension, and all suggest the utility of a matching principle for understanding attitude change.

Despite the similarity of our results to those obtained by Millar and Tesser (1986), it could be argued that response mode is mapping onto something entirely different than attitude evaluation. The reasoning against this criticism is strong. First, the results reported here support predictions generated from Millar and Tesser's (1986) model and Wilson's (e.g., Wilson, Kraft and Dunn 1989) attitude change hypothesis. Second, these results are precisely what would have been expected if manipulations like those of the above researchers had been employed. Finally, the processes are logically similar. The judgment mode, like analyzing reasons (a cognitive focus), requires an evaluative statement for each alternative, and results in a more complete and thorough search (Bailey and Billings 1994; Billings and Scherer 1987). By way of contrast, the choice mode, like observing attitudes (an affective focus), is guided by individual preferences and desires resulting in a less thorough information search. The focus here is clearly on one's feeling; what is wanted, not why it is wanted. Because of the truncated search, information that might call the prevailing attitude into question is never encountered. Similarly, when individuals focus on feelings, they do not necessarily scrutinize these feelings. Although Johnson and Russo (1984) have argued that choice requires and optimal decision and is therefore more analytic than judgment, their point is that choice calls for the use of decision rules that promote a more detailed alignment with preferences.

Implications and Future Research

Marketing efforts have been largely concerned with changing consumer attitudes or preferences through persuasively styled messages. The presumption is that once changed these attitudes will extrapolate to consumer decisions. However, the current data suggest that attitudes, no matter how they may have been influenced by persuasive attempts, change as a function of the manner of evaluation engaged during the actual multi-alternative decision episode. Specifically, pre-decisional preferences may predict decisions under conditions where the underlying attitudes remain stable (e.g., judgment-instrumental and choice-consumatory), but the relationship may diminish when conditions promote reevaluation (e.g., judgment-consumatory and choice-instrumental). In general, any circumstances that influence the purpose of decision making may also influence attitude structure.

This effort is not without limitations. Perhaps the most prominent of these is that the instrumental behavioral intention manipulationCby posing the possibility of a test on analytic abilityCseems removed from everyday consumer decision making. Nevertheless, to the extent that the instrumental condition furnished a type of accountability manipulation, it does have relevance for purchasing agents or other organizational operatives that are responsible for product evaluations. Second, the exclusive use of amount of information as a measure of evaluation is clearly a narrow index. Enlisting convergent, qualitative techniques such as thought-listing or verbal protocol techniques for corroborative evidence would improve confidence in this conjecture. Third, several small procedural variations could have significantly strengthened the study's design. Including a control group to gauge natural attitudinal fluctuations would have provided a sounder base for comparison. Also, collecting a measure of behavioral intention (as opposed to only manipulating it) would have allowed testing whether pre- or post-test attitudes better predict behaviors, and provided an additional test of Millar and Tesser's (1986) AB hypothesis.

Finally, although the congruency explanation evoked here is intuitively appealing, this study was not designed to test which model of attitude change applies. The results rule out polarization, but other models may apply. One interpretation pertains specifically to the type of evaluation elicited by manipulations. Judgment entails a careful and deliberative type of mental activity which directs attention to the merits of an attitude structure. On the other hand, choice is a preferential-based process where the concentration is on sentiment without concern for justification. These contrasting processesCjudgment versus choiceCbear significant resemblance to the dual process model of persuasion described by Petty and Cacioppo (e.g., 1986) and Chaiken, Liberman and Eagly (1989). Although there are significant differences between the approach employed herein and these two paradigms, the similarities warrant further exploration.

CONCLUSION

The current research was unique in that subjects evaluated their attitudes towards numerous objects simultaneously. Hence, in addition to the ostensible purpose of verifying differences in process, the IDB allowed for the presentation and consideration of attitude objects in a more realistic fashion. This research also extended inquiry on attitude evaluation to the decision making arena, because in most instances evaluation of product alternatives is in the service of making a consumer decision. There is mounting evidence that the underlying constellation of values or preferences that guide decisions are not stable, but rather are constructed by, among other things, the circumstances which require their use (Payne, Bettman and Johnson 1991; Upmeyer 1989). If different types of evaluation lead to differential changes in attitudes, this knowledge should facilitate speculation into what decisions consumers make, and more importantly, why they are made. A meeting of these two literatures could prove mutual profitable by furnishing an intelligible springboard for relating attitude change to actual purchasing decisions.

REFERENCES

Bailey, James. R. and Robert S. Billings (1994). "Sequential phases of judgment and the value representation of product alternatives," Advances in Consumer Research, 21, 437-441.

Billings, Robert S. and Lisa L. Scherer (1988), "The effects of response mode and importance on decision making strategies: Judgment versus choice," Organizational Behavior and Human Decision Processes, 41, 1-19.

Chaiken, Shelly, A. Liberman and Alice H. Eagly (1989). "Heuristic and systematic information processing within and beyond the persuasion context," in Unintended thought, ed. J. S. Uleman and J.A. Bargh, New York: Guilford Press.

Edwards, Karen. (1990). "The interplay of affect and cognition in attitude formation and change," Journal of Personality and Social Psychology, 59, 202-216.

Johnson, Eric, J. and J. Edward Russo (1984). "Product familiarity and learning new information," Journal of Consumer Research, 11, 542-550.

Millar, Murray. G. and Karen U. Millar (1990). "Attitude change as a function of attitude type and argument type," Journal of Personality and Social Psychology, 59, 217-228.

Millar, Murray. G. and Abraham Tesser (1986). "Effects of affective and cognitive focus on the attitude-behavior relation," Journal of Personality and Social Psychology, 51, 270-276.

Payne, John. W., James R. Bettman and Eric J. Johnson (1992), "Behavioral decision research: A constructive processing perspective," Annual Review of Psychology, 43, 87-131.

Petty, Richard. E. and John T. Cacioppo (1986). "The elaboration likelihood model of persuasion," in Advances in experimental social psychology, Vol. 19, ed. L. Berkowitz, New York: Academic Press.

Upmeyer, A. (Ed.) (1989). Attitudes and behavioral decisions. New York: Springer-Verlag.

Wilson, Timothy D., Dana. S. Dunn, Dolores Kraft and Douglas. J. Lisle (1989), "Introspection, attitude change, and attitude-behavior consistency: The disruptive effects of explaining why we feel the way we do," in Advances in Experimental Social Psychology, Vol. 22, ed. L. Berkowitz, New York: Academic Press.

Wilson, Timothy. D., Dolores Kraft, and Dana S. Dunn (1989). "The disruptive effects of explaining attitudes: The moderating effect of knowledge about the attitude object," Journal of Experimental Social Psychology, 25, 379-400.

Zajonc, Robert. B. and Hazel Markus (1982). "Affective and cognitive factors in preferences," Journal of Consumer Research, 9, 123-131.

----------------------------------------