Developing Negatives: Expectancy Assimilation and Contrast in Product Judgments

ABSTRACT - The effect of expectations on attribute weighting and product evaluations was examined. It was found that positive and negative expectations were assimilated into performance judgments and overall evaluations of a camera that subjects were asked to consider purchasing. A contrast effect also occurred: positive expectations led to greater weighting of product flaws than did negative expectations. This negativity effect, however, did not damage overall evaluations, because of the assimilation of the positive expectancy. This research thus supports the elevation of product expectations in order to boost product perceptions.

Citation:

Jill Gabrielle Klein (1999) ,"Developing Negatives: Expectancy Assimilation and Contrast in Product Judgments", in NA - Advances in Consumer Research Volume 26, eds. Eric J. Arnould and Linda M. Scott, Provo, UT : Association for Consumer Research, Pages: 463-468.

Advances in Consumer Research Volume 26, 1999      Pages 463-468

DEVELOPING NEGATIVES: EXPECTANCY ASSIMILATION AND CONTRAST IN PRODUCT JUDGMENTS

Jill Gabrielle Klein, INSEAD

[The author would like to thank Bridgett Braig, Bobby Calder, Richard Ettenson, Andrew John, Ann McGill, Eyal Maoz, and Brian Sternthal for their helpfil comments.]

ABSTRACT -

The effect of expectations on attribute weighting and product evaluations was examined. It was found that positive and negative expectations were assimilated into performance judgments and overall evaluations of a camera that subjects were asked to consider purchasing. A contrast effect also occurred: positive expectations led to greater weighting of product flaws than did negative expectations. This negativity effect, however, did not damage overall evaluations, because of the assimilation of the positive expectancy. This research thus supports the elevation of product expectations in order to boost product perceptions.

Marketers are in somewhat of a quandary when it comes to managing expectations, even though the effect of expectations on product judgments has received considerable attention from consumer researchers (e.g., Cardozo 1968; Chuchill and Surprenant 1982). It is clear that expectations act as a frame of reference for judging product or service performance (Oliver 1980; Oliver and Bearden 1985; Parasuraman, Zeithaml, and Berry 1985; 1988). But there are conflicting findings concerning whether expectations are assimilated into product performance judgments (Boulding, Kalra, Staelin and Zeithaml 1993; Olshavsky and Miller 1972; Olson and Dover 1979), or whether performance is instead contrasted against expectations, perhaps resulting in expectancy disconfirmation (for a review see Yi 1990).

A great deal of research from both psychology and marketing suggests that expectations can act as powerful interpreters of the world around us. In particular, we often perceive performance as consistent with our expectations (e.g., Anderson 1973; Fiske and Neuberg 1989; Hoch and Ha 1986). Evidence for assimilation comes from studies that show that expectations can influence final judgments in the direction of the expectation, even when performance disconfirms the prior beliefs. For example, Olson and Dover (1979) found that consumers who were given a positive expectation about the taste of bitter ground coffee rated the product more positively than did a control group of consumers who simply tasted the bitter coffee without any prior expectation.

Instead of being assimilated into overall product judgments, however, positive expectations can be disconfirmed. The consumer may notice that the product does not live up to expectations, which then may lead to disappointment and dissatisfaction (Oliver 1993). Thus, it is unclear whether marketers benefit or are harmed by generating positive expectations about their product or service. The present research approaches this problem by examining a specific type of expectancy disconfirmationCthat which occurs when product flaws (i.e., poor performance) are contrasted against a positive expectation, resulting in the disproportionate weighting of these flaws.

Evidence on the weighting of attributes in product evaluation indicates that product flaws have a more powerful effect than product strengths on both product evaluations and purchase intention (see Weinberger, Allen and Dillon 1980, for a review). Social psychologists label this tendency the negativity effect (see Kanouse and Hanson 1972, and Skowronski and Carlston 1989, for a review). A dominant social-psychological explanation for negativity in the perception of other people is the expectancy-contrast explanation. This explanation is predicated on the empirically-supported assumption that we generally have positive expectations of others (Kaplan 1976; Sears 1983): we tend to judge others positively and tend to expect others to act in a positive fashion (Jones and Davis, 1965). Negative behavior is distinctive because it contrasts with our expectations, and so it receives greater attention and greater weight in impression formation (Fiske 1980; Skowronski and Carlston 1989).

The expectancy-contrast explanation for negativity suggests that if one has a positive expectation about a product, but the product performs poorly on a particular attribute, this negative attribute should be weighted more heavily in overall evaluations (Kanouse and Hanson 1972; Skowronski and Carlston 1989). Thus, negativity might be a process by which expectations are disconfirmed, and satisfaction is reduced: positive expectations create a perceptual anchor against which negative information is contrasted and weighted more heavily.

Assimilation and Contrast in Product Judgments

The evidence that expectations are assimilated into overall evaluations in product judgments comes from studies that used overall evaluations or measures of overall satisfaction as their primary dependent measures. These studies did not assess attribute weighting. Therefore, they are silent on whether or not information that contrasts with expectations is weighted more heavily. But given that researchers have found support both for expectancy assimilation and for expectancy-contrast, it is natural to ask: can assimilation and contrast operate together when consumers form product judgments?

The present research investigates whether or not a negative product attribute that contrasts with a positive expectation receives disproportionate weight in judgments, even while overall evaluations remain consistent with expectations. This coexistence of assimilation and contrast could occur for two reasons. First, flaws may be perceived and interpreted as less detrimental when a positive expectation is assimilated into judgments. Second, other, more positive aspects of the product will be judged more favorably when the positive expectation is assimilated.

The current study examines the effect of expectations on overall product evaluations and on the weighting of product attributes. Two variables were manipulated: expectation (positive or negative) and actual performance (good or poor). Subjects were exposed to one of the four treatment conditions and were asked to consider the purchase of a camera. After receiving the expectation, subjects were given photographs taken by the camera and asked to judge the camera’s performance. The following hypotheses were proposed:

H₁:  Subjects in the positive expectation conditions will judge the performance of the camera more positively than will subjects in the negative expectation conditions. Ratings of the focus and non-focus attributes, and of overall evaluations, will be more positive in the positive expectation conditions than in the negative expectation conditions, regardless of the camera’s performance.

H₂:  Subjects in the positive expectation/poor performance condition will weight the negative attributes (focus) of the camera most heavily, while subjects in both negative expectation conditions and subjects in the positive expectation/good performance conditions will not show differential weighting of the attributes.

METHODS

Subjects

Subjects were 59 master of management students at a Midwestern university who volunteered to participate. Participants were run in groups of 5 to 12.

Design

A 2 X 2 X 2 mixed-subjects design was employed in which the two between-subjects factors were expectation (negative expectation versus positive expectation) and performance (good performance versus poor performance) and the within-subject factor was attribute (focus versus non-focus attributes).

Stimuli and procedure

Subjects read instructions that asked them to imagine that they were planning to purchase a new 35mm camera. They were told that in this study they would evaluate one of the cameras that they might consider purchasing. To encourage involvement in the task, subjects were asked to imagine that they were purchasing the camera because they would be taking pictures at a friend’s wedding.

Expectation Manipulation. All subjects read that they would be given some information about the camera and a set of photographs taken with the camera. Subjects then received information that conveyed the expectation manipulation. In the negative expectation condition this information read:

"The photographs in the enclosed envelope were taken with a camera made by a company that has been given cnsistently low ratings by photography experts. According to a national survey the cameras made by this company are viewed as below standard by professional photographers and casual photographers alike. Many photo-journalists say they will not use this brand, and it is also viewed negatively among non-professionalsCpeople who simply use their cameras for vacations, holidays, etc. Models made by this company often receive among the lowest ratings from Consumer Reports."

Subjects in the positive expectation condition read:

"The photographs in the enclosed envelope were taken with a camera made by a company that has been given consistently high ratings by photography experts. According to a national survey the cameras made by this company are highly respected by professional photographers and casual photographers alike. Many photo-journalists use this brand, and it is also a favorite among non-professionalsCpeople who simply use their cameras for vacations, holidays, etc. Models made by this company often receive among the highest ratings from Consumer Reports."

A pretest of the expectation manipulation based on the ratings of an additional 13 master of management students showed that subjects were significantly more favorable towards the camera after reading the positive paragraph (M=5.71, on a 1-to-7 scale) than were subjects who read the negative paragraph (M=1.33, t(11)=4.71, p<.001).

Performance Manipulation. Performance was manipulated with the goal of creating two types of camera profiles: a profile in which all attributes were positive, and a profile in which some attributes were positive and some were negative. Thus, in the poor performance condition, attributes related to the focusing ability of the camera were manipulated to be negative. This was achieved by a manipulation of the quality of the photographs presented to subjects. Following the expectation manipulation, all subjects read:

"The camera is fully automatic. The shutter speed, focus, and aperture are all set automatically as each picture is taken. A group of 20 people who are members of the Chicago Area Photography Club volunteered to shoot one 24 exposure, 200 ASA role of film. From the nearly 500 pictures taken, 10 were randomly selected and placed in your envelope. Thus, the pictures you have should accurately represent the quality of pictures taken by the camera across a number of club members taking the pictures. The 20 people who took the pictures were not professional photographers, but are active photography hobbyists and were shown how to use the camera. They were given the opportunity to practice with the camera before shooting the test roll. The film used in the study was a commonly used, high-quality film."

This information ensured that subjects would view their packet of photos as a representative sample of pictures taken with the camera, independent of the abilities of any one photographer.

Subjects were asked to examine the photos in their envelope and were given as much time as they desired to do so. In the poor performance condition, five of the ten pictures were out of focus. In the good performance condition, all ten pictures were in focus. [Only the focus of the photographs was manipulated. Five photographs were identical across the performance conditions and were in clear focus. The other five photographs were identical except for the focus manipulation. Thus, subjects saw the same ten photographs except for the focus differences. In the good-performance condition, the five manipulated photographs were in clear focus. In the poor performance conditions, three photographs were slightly out of focus, one was moderately out of focus, and one was extremely out of focus (such that it was difficult to tell the subject of the photo). The packet of ten photos included pictures of people, close objects, and distant objects. All photographs were developed with the same equipment in order to control for color and lighting.] After subjects examined the photos they completed a questionnaire rating the camera on focus (e.g., "takes photos in which objects are clear"), and non-focus attributes such as color (e.g., "takes colorful photos") and lighting (e.g., "is adaptive to different levels of lighting"). Subjects indicated how well each of the phrases described the camera on a one-to-seven semantic differential scale with "not well at all" and "extremely well" as anchors. Subjects then indicated how favorable or unfavorable they felt towards the camera overall on a seven-point scale ranging from"very unfavorable" to "very favorable". They also indicated how likely they would be to purchase the camera on a seven-point scale ranging from "not at all likely" to "very likely".

After they completed the questionnaire, subjects were debriefed and thanked for their participation.

TABLE 1

MEAN CAMERA RATINGS

RESULTS

Attribute Factors

A principal components analysis (with varimax rotation) revealed that the ten attributes loaded on two components. The first component (Eigenvalue 6.26), which will be referred to as "focus", included the items "takes sharp photos", "focuses well", "takes photos in which objects are clear", and "takes photos in which shapes and lines are clearly defined". The second component (Eigenvalue 1.34) , which will be referred to as "non-focus", included the items "takes colorful photos", "takes photos that show accurate colors", "is capable of capturing the full range of colors", "is adaptive to different levels of lighting", "takes photos well in darker conditions", and "provides accurate light exposure". Two indices were created by averaging the variables within each component. Cronbach’s a’s for both the focus (M=4.1) and non-focus (M=4.5) indices were .92. [The same two factors emerged when the analyses were conducted separately for the poor and good performance conditions.]

Manipulation Checks

To examine the effects of the performance manipulation on perceptions of performance, a 2 X 2 mixed-model ANOVA with performance (good or poor) as the between-subjects factor and type of attribute (focus or non-focus) as a within-subject factor was conducted. As intended, in the poor performance conditions the performance manipulation led to significantly lower ratings of the focus attributes (M=3.84) than the non-focus attributes (M=4.51, F(1,57)= 11.94, p<01). In the good performance conditions the focus (M=4.36) and the non-focus attributes (M=4.48) received similar ratings, (F(1,57)<1). This attribute X performance interaction was significant (F (1,57)=4.08, p<.05).

Evaluations of the Camera

As Table 1 shows, attribute ratings were affected by the expectation and performance manipulations. A 2 (positive or negative expectation) X 2 (good or poor performance) X 2 (focus or non-focus attributes) mixed-model MANOVA showed a main effect for attribute, with non-focus attributes rated significantly higher than the focus attributes (F(1,55)=8.47, p<.01). There was also a significant main effect for expectation (F(1,55)=20.43, p<.001). All other effects were non-significant (except for the attribute X performance interaction reported in the manipulation checks sections, above).

Overall evaluations of the camera were submitted to a 2 (positive versus negative expectation) X 2 (good or poor performance) ANOVA. [Nearly identical results were obtained using either overall favorableness toward the camera or purchase intention as separate dependent measures. The two items are correlated at .90. Thus, an average of the two items was used as a measure of overall evaluation.] Overall ratings were significantly higher for those in the positive expectation conditions (M=4.15) than in the negative expectation conditions (M=2.97), regardless of how well the camera performed (F(1,55)=11.85, p<.01). The main effect for performance was non-significant (F(1,55)<1), as was the performance X expectation interaction (F(1,55)<1).

Thus, as H₁ predicts, the expectation manipulation had an impact on attribute judgments and overall evaluations regardless of the camera’s performance.

Weighting of Attributes

The expectancy-contrast explanation for negativity predicts that indications of poor camera performance will have the most powerful impact on overall evaluatins when expectations are positive. To test H₂, multiple regressions were run with overall evaluation as the dependent variable and the indices for the focus and non-focus attributes as the two independent variables. Analyses were run separately for each of the four conditions. Table 2 displays the regression slope coefficients, standard errors, and significance levels.

When the camera performed well there were no significant differences in the predictive power of the focus and non-focus attributes in either the negative expectation (t(31)=-.80, n.s.), or the positive expectation conditions (t(29)=-.80, n.s.). When the camera performed poorly, there were no significant differences in the predictive power of the two types of attributes when the expectation was negative (t(25)=-1.13, n.s.), but, as the expectancy-contrast theory predicts, focus ratings were significantly more predictive of overall evaluations than were non-focus ratings when expectations were positive (t(29)=4.12, p<.01).

Further, in the negative expectation conditions, focus ratings had equal weight in the good and poor performance conditions (t(28)=.96, n.s.), but in the positive expectation conditions, the focus ratings were significantly higher in the poor performance than in the good performance conditions (t(29)=2.47, p<.05). [There was no relationship between the standard deviation of the attribute ratings and the slope coefficient.]

TABLE 2

REGRESSION UNSTANDARDIZED COEFFICIENTS, STANDARD ERRORS, AND SIGNIFICANCE LEVELS

DISCUSSION

The results indicate that a positive expectation led to two outcomes: (1) relatively high overall evaluations of the camera and its attributes, even in the face of evidence that the camera performed poorly; and (2) greater weighting of the focus attributes when the camera performed poorly on those attributes and expectations were high. Thus, evidence was found both for the assimilation of expectations and for the expectancy-contrast explanation for negativity.

The focus characteristics of the camera were clearly noted by the high expectation/poor performance subjects; these attributes had a disproportionate impact on overall judgments relative to the other camera attributes (and the focus attributes in the other three conditions). It might have been expected that the overall evaluations of the camera would suffer. This was not the case, however. The focus attributes did contrast with expectations and therefore were weighted heavily. But the positive expectation was assimilated to such an extent that the focus attributes were not judged as very negative. Further, in the negative expectation/poor performance condition, the focus attributes were judged quite negatively (below the median of the scale). These attributes did not contrast with expectations, however, and therefore did not receive disproportionate weight.

The current research has implications both for theory development and for marketing management. From a theoretical viewpoint, the study contributes to our understanding of the mechanisms underlying negativity. Skowronski and Carlston (1989) argue against the expectancy-contrast theory as an explanation for negativity on the basis of Kaplan’s (1973, 1976) studies that found assimilation of expectation. They suggest that if negativity results from the contrast of negative information against positive expectation, we would not find assimilation. The present research suggests that both contrast and assimilation effects can occur in tandem in the same judgment. Expectancy-contrast may well underlie negativity, but the effect of this disproportionate weighting on final judgments may sometimes be obscured by the assimilation of the positive expectation.

This study may also help us understand the process by which the disconfirmation of a positive expectation occurs. As mentioned above, expectancy disconfirmation has been found in some studies to reduce satisfaction. The results of the present study suggest that disconfirmation may negatively affect satisfaction because product or service flaws ar contrasted against expectations and weighted heavily in consumers’ overall evaluations.

The key question concerning the setting of expectations is whether or not a positive expectation will be powerful enough to be assimilated into judgments of performance, even in the face of poor performance. In answering this question, three issues must be examined: the strength of the expectation, the clarity of the product’s performance, and the extent of negativity produced by the contrast with positive expectations.

In the present study, the expectation manipulation was both powerful and convincing. Professionals, amateurs, and a reputable consumer magazine all touted the superiority of the camera in the positive expectation condition. The assimilation of a weaker expectation would probably not have had such a strong influence on product evaluations. But, a weaker expectation might not have created enough contrast for negativity to occur either. Future research should examine the effects of weak, moderate, and extreme expectations on assimilation and negativity effects.

There is some evidence that the clarity of a product’s performance moderates the assimilation of expectations. In their examination of the effects of performance ambiguity on expectancy assimilation, Hoch and Ha (1986) found that performance judgments were more susceptible to expectation-based interpretations if product performance was difficult to judge. In the present study, the camera’s performance was not difficult to judge, but performance ambiguity was not manipulated. It would be interesting to examine the presence of assimilation and negativity under conditions in which performance is either ambiguous or clear-cut.

The degree of negativity produced by a positive expectation may depend on a number of factors. It is unlikely that only one mechanismCexpectancy-contrastCunderlies negativity. Concerns about potential negative outcomes (cost-orientation) or perceiving negative information to be indicative of overall performance (diagnosticity) have both been found to increase negativity (Peeters and Czapinski 1990; Skowronski and Carlston 1987, 1989). Thus, the presence of these conditions may intensify negativity. An investigation of the joint and interactive effects of expectation, cost-orientation and diagnosticity might provide another fruitful avenue for future research.

The fact that no positivity effects were found in the present study is very likely due to the nature of the performance manipulation. In the good performance condition, the camera performed very well, but there was nothing about the photographs that would be particularly outstanding in a positive direction. Further, while the good performance should have been unexpected in the negative expectation condition, past research suggests that to get a positivity effect, positive information must be extreme (Fiske 1980). As mentioned above, other factors, such as concerns about negative outcomes, may predispose us more toward negativity than to positivity. In sum, given that negative expectations can be assimilated into product judgments, and that extremely positive performance is required for positivity to result, at the very least it seems clear that marketers should avoid the development of negative expectations (Boulding et al. 1993; Spreng, Mackenzie, and Olshavsky 1996).

The marketing manager is still faced with a dilemma. Decisions about the management of expectations are complex, and consumer research to date has not produced clear-cut guidelines or simple rules for managers to follow. Should expectations be raised in hope of assimilation, or should they be lowered in fear of the negativity effect? Even the most recent research on consumer expectations fails to answers this question decisively. Spreng, MacKenzie, and Olshavsky (1996, p. 27) state:

"Further research is needed to establish more clearly the appropriate managerial actions to take in certain situations. The best or safest route still may be to increase performance. Thus, managing expectations may be more difficult than was oriinally thought."

Although the present research does not fully resolve the dilemma facing marketers who must manage consumer expectations, several approaches to the design of marketing research studies can be recommended. Managers should measure consumer expectations, either those that consumers already hold, or those produced by specific promotional strategies to which consumers are exposed during testing. Research participants can then be exposed to the actual product and given the opportunity to evaluate performance. The weighting of product liabilities can then be assessed either though a regression approach where attribute judgments are used to predict overall ratings, conjoint analysis, or self-explicated weights (or some combination of these). The marketing decision-maker will then be able to assess whether expectations will be assimilated to such an extent that product flaws are not viewed very negatively, andCeven if allotted disproportionate weightCdo not damage product evaluations.

REFERENCES

Anderson, R.E. (1973). Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance, Journal of Marketing Research, 10, 38-44.

Boulding, W., Kalra, A., Staelin, R. & Zeithaml, V.A. (1993). A Dynamic Process Model of Service Quality: From Expectations to Behavioral Intentions. Journal of Marketing Research, 30, 7-27.

Cardozo, R. N. (1968). An Experimental Study of Consumer Effort, Expectation, and Satisfaction," Journal of Marketing Research, 2, 244-9.

Churchill, G.A. Jr. & Surprenant, C. (1982). An Investigation into the Determinants of Customer Satisfaction, Journal of Marketing Research, 19, 491-504.

Fiske, S.T. (1980). Attention and Weight in Person Perception: The Impact of Negative and Extreme Behavior, Journal of Personality and Social Psychology, 38, 889-906.

Fiske, S.T. & Neuberg, S.L. (1990). A Continuum Model of Impression Formation from Category-Based to Individuating Processes: Influence of Motivation on Attention and Interpretation. In M.P. Zanna (Ed.), Advances in Experimental Social Psychology, Vol. 23, pp. 1-74. New York: Academic Press.

Hoch, S.J. & Ha, Y.W. (1986). Consumer Learning: Advertising and the Ambiguity of Product Experience, Journal of Consumer Research, 13, 221-33.

Jones, E. E. & Davis, K. E. (1965). From Acts to Dispositions: The Attribution Process in Person Perception, In L. Berkowitz (Ed.), Advances in Experimental Social Psychology, Vol. 2, New York: Academic Press.

Kanouse, D.E. & Hanson, L.R. (1972). Negativity in Evaluations, in E.E. Jones et al., eds., Attribution: Perceiving the Causes of Behavior, Morristown, N.J.: General Learning Press.

Kaplan, M. F. (1973). Stimulus Inconsistency and Response Dispositions in Forming Judgments of Other Persons. Journal of Personality and Social Psychology, 25, 58-64.

Kaplan, M. F. (1976). Measurement and Generality of Response Dispositions in Person Perception, Journal of Personality, 44, 179-194.

Klein, J. G. (1991). Negativity Effects in Impression Formation: A Test in the Political Arena, Personality and Social Psychology Bulletin, 17, 412-418.

Oliver, R.L. (1980). A Cognitive Model of the Antecedents and Consequences of Satisfaction Decisions, Journal of Marketing Research, 17, 460-469.

Oliver, R.L. & Bearden, W.O. (1985). Disconfirmation Processes and Consumer Evaluations in Product Usage, Journal of Business Researc, 13, 235-246.

Oliver, R.L. (1993). Cognitive, Affective, and Attribute Bases of the Satisfaction Response, Journal of Consumer Research, 20, 418-430.

Olshavsky, R.W. & Miller J.A. (1972). Consumer Expectation, Product Performance, and Perceived Product Quality, Journal of Marketing Research, 9, 19-21.

Olson, J.C. & P.A. Dover (1979). Disconfirmation of Consumer Expectations through Product Trial, Journal for Applied Psychology, 64, 179-189.

Parasuraman, A. Valarie A. Zeithaml, and Leonard L. Berry (1985), "A Conceptual Model of Service Quality and Its Implications for Future Research," Journal of Marketing, 49, 41-50.

Parasuraman, A. Valarie A. Zeithaml, and Leonard L. Berry (1985), "SERVQUAL: A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality," Journal of Retailing, 64, 12-40.

Peeters, G. & Czapinski, J. (1990). Positive-Negative Asymmetry in Evaluations: The Distinction between Affective and Informational Effects, In W. Stroebe and M. Hewstone, eds., European Review of Social Psychology (Vol. 1), Chichester: Wiley.

Sears, D.O. (1983). The Person-productivity Bias, Journal of Personality and Social Psychology, 44, 233-249.

Skowronski, J.J. & D.E. Carlston (1987). Social Judgement and Social Memory: The Role of Cue Diagnosticity in Negativity, Positivity, and Extremity Biases, Journal of Personality and Social Psychology, 52, 689-699.

Skowronski, J.J. & Carlston, D.E. (1989). Negativity and Extremity Biases in Impression Formation: A Review of Explanations, Psychological Bulletin, 105, 131-142.

Spreng, R. A., MacKenzie, S. B., & Olshavsky, R.W. (1996). A Reexamination of the Determinants of Consumer Satisfaction. Journal of Marketing, 60, 15-32.

Weinberger, M.G., Allen, C.T. & Dillon, M.R. (1980). Negative Information: Perspectives and Research Directions, in J. Olson, ed., Proceedings of the Association for Consumer Research, Ann Arbor, MI.: Association for Consumer Research, 398-404.

Yi, Youjae (1990). A Critical Review of Consumer Satisfaction. In Review of Marketing 1990, V. A. Zeithaml (Ed.). Chicago: American Marketing Association, 68-123.

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