Value-Percept Disparity: an Alternative to the Disconfirmation of Expectations Theory of Consumer Satisfaction

Robert A. Westbrook, University of Arizona
Michael D. Reilly, University of Arizona
ABSTRACT - The disconfirmation of expectations paradigm has dominated research on consumer satisfaction/dissatisfaction. While support has been found for this theory of satisfaction, an alternative theoretical approach may be more appropriate from a conceptual standpoint as well as more parsimonious. Termed the value-percept disparity model, this theory of satisfaction is presented in this paper and examined empirically directly against the disconfirmation of expectations theory. Data from a preliminary study were analyzed using structural equations modeling with allowance for measurement error. The results suggest the need for continued efforts to improve the measurement of constructs theorized to determine consumer satisfaction.
[ to cite ]:
Robert A. Westbrook and Michael D. Reilly (1983) ,"Value-Percept Disparity: an Alternative to the Disconfirmation of Expectations Theory of Consumer Satisfaction", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 256-261.

Advances in Consumer Research Volume 10, 1983      Pages 256-261

VALUE-PERCEPT DISPARITY: AN ALTERNATIVE TO THE DISCONFIRMATION OF EXPECTATIONS THEORY OF CONSUMER SATISFACTION

Robert A. Westbrook, University of Arizona

Michael D. Reilly, University of Arizona

ABSTRACT -

The disconfirmation of expectations paradigm has dominated research on consumer satisfaction/dissatisfaction. While support has been found for this theory of satisfaction, an alternative theoretical approach may be more appropriate from a conceptual standpoint as well as more parsimonious. Termed the value-percept disparity model, this theory of satisfaction is presented in this paper and examined empirically directly against the disconfirmation of expectations theory. Data from a preliminary study were analyzed using structural equations modeling with allowance for measurement error. The results suggest the need for continued efforts to improve the measurement of constructs theorized to determine consumer satisfaction.

INTRODUCTION

Recent years have witnessed a growing interest in understanding consumer satisfaction/dissatisfaction (Day 1977; Oliver 1980; Westbrook 1980; Swan and Martin 1981). The dominant conceptual paradigm reflected in these efforts is essentially a cognitive consistency model, in which consumers compare pre-purchase beliefs about a product to post-purchase beliefs formed during consumption of the product. The extent to which post-purchase beliefs disconfirm their pre-purchase counterparts is theorized to be the principal determinant of their satisfaction/dissatisfaction. While this approach has received some empirical support, there exists an alternative model of the process determining consumer satisfaction which is more parsimonious and perhaps more appropriate from -a conceptual standpoint. This paper develops this alternative model of the satisfaction/ dissatisfaction process, known as the value-percept disparity model, and compares its explanatory ability to the disconfirmation of expectations model.

Conceptualization of Satisfaction

One of the major obstacles to the development of theory in consumer satisfaction/dissatisfaction has been a lack or attention to the specification of the construct. While there have been some efforts to clarify the construct, these attempts do not distinguish the phenomenon from the process which precedes and determines it. For example, Hunt (197, ) summarized the First Annual Consumer Satisfaction Conference by suggesting that satisfaction is "an evaluation rendered that the (product ownership and usage) experience was at least as good as it was supposed to be." Engel and Blackwell (1982, p. 501) conceive satisfaction as "an evaluation that the chosen alternative is consistent with prior beliefs with respect to that alternative." Howard and Sheth (1969, p. 145) define satisfaction as "the buyer's cognitive state of being adequately or inadequately rewarded for the sacrifices he has undergone." While these views suggest something about the evaluative process which may lead to states of satisfaction and dissatisfaction, they tell us little as to the nature of satisfaction per se. Some writers (e.g., LaTour and Peat 1979) have viewed satisfaction as a general evaluative response to a product, perhaps not discernibly different than the well-studied concept of attitude.

This formulation, however, fails to recognize the unique nature of satisfaction and its positioning in the sequence of post-purchase cognitive affective processes. While it is true satisfaction and attitude are both evaluative responses, satisfaction is best conceived as an evaluation of the consumer's own purchase and consumption experiences, while attitude is construed as an evaluation of the product per se. Oliver's recent empirical work (1980) supports this distinction in that satisfaction wa-s found to precede and influence post-purchase attitude in a path-analytic causal model.

Oliver (1981) has also proposed that consumer satisfaction is the summary psychological state created by the coupling of two sources of affect: the surprise of finding one's pre-purchase product-specific expectations disconfirmed by post-purchase experiences with the product, and one's pre-purchase feelings toward product-related experiences. Oliver describes satisfaction 25 a dynamic phenomenon. arising after expectancy disconfirmation and decaying over time into overall attitude toward the product. Oliver's conceptualization explicitly recognizes the affective character of satisfaction, as well as the focus of the concept on consumer experiences rather than the product per se. However, the major shortcoming of this formulation is its dependence on a particular theory of the process determining consumer satisfaction, notably the disconfirmation of expectations paradigm.

A simpler but more general conceptualization of consumer satisfaction may be advanced. Locke's (1967) seminal analysis of job satisfaction suggests that satisfaction is a general psychological phenomenon, describing the emotional state resulting from an evaluation of one's experiences in connection with an object, action, or condition. Consumer satisfaction, then, is an emotional response to the experiences provided by, or associated with, particular products or services purchased, retail outlets, or even molar patterns of behavior such as shopping and buyer behavior, as well as the overall marketplace.

Further explication of this conceptualization of satisfaction requires discussion of the relationships between cognition, evaluation, and emotion. The faculty of cognition (sensation, perception, and conception), concerns the identification of existents in the psychological field, i.e., things, objects, actions, etc. Evaluation concerns the estimation of the beneficial relationship of perceived existents to the individual, enabling choice among alternative actions to fulfill the individual's needs.

The most basic evaluations are simply physical sensations of pleasure and pain, which inform the organism whether its action is conducive to or jeopardizing of life. The higher mental faculties enable identification of requirements through reasoning and conscious foresight. To this end, we acquire a complex set or code of values. Values are simply "that which one acts to gain and/or keep" and represent what one regards as conducive to one's welfare (Rand 1964; Branden 1966). The process of evaluation, then, consists of estimating the relationship of an object, action or condition to an individual's values. Essentially, evaluation addresses the question of whether the object, action, or condition enhances or threatens the individual's values. Values may be conscious or subconscious. That is, the full set of human values may not simultaneously be held in focal awareness but rather, like knowledge. may be brought to consciousness more or less on demand.

When a value judgment is reached, emotions result. While numerous types of emotions have been discerned (Izard 1977; Plutchik 1980), the most basic differentiation concerns pleasure/joy versus displeasure/ suffering. Pleasure is the result of perceived attainment or one's values, while displeasure derives from the perceived negation or destruction of one's values.

Applied to consumer behavior, this analysis suggests that consumer satisfaction is the pleasurable emotional state resulting from the appraisal of a product, service, retail outlet or consumer action (a complaint, shopping trip, purchase decision, etc.) as leading to or achieving one's values. Conversely, consumer dissatisfaction refers to the unpleasurable emotional state resulting from an appraisal that an object or action or condition in the consumer domain blocks or denies the achievement of one's values or attains one's disvalues.

This conceptualization of consumer satisfaction/dissatisfaction is not only more precise in its differentiation of the construct from its determination, but also more general. Expectancy disconfirmation is really a specific type of value judgment. Not only is the conceptualization advanced here consistent with earlier views, but it is also able to deal with instances of satisfaction which may involve value judgments other than whether prior product expectations were confirmed or disconfirmed by post-purchase beliefs arising from experiences with the product.

The Disconfirmation of Expectations Model

The disconfirmation or expectations model holds that satisfaction/dissatisfaction responses arise from a cognitive evaluation process in which pre-purchase "expectations" or prior beliefs about the likelihood of product-related experiences or outcomes are retrieved from memory and compared to cognitions about the product-related experiences or outcomes actually realized in the consumption of the product. The result of this comparison is expectancy disconfirmation, which ranges .rom negative (expectations exceed realized outcomes) through zero (expectations just equal realized outcomes) to positive (realized outcomes exceed expectations). Various studies have empirically confirmed the direct effect of disconfirmation or post-choice product evaluation (Cardozo 1968; Cohen and Goldberg 1970; Olson and Dover 1976) and satisfaction responses (Swan 1977; )liver 1980; Westbrook 1980).

Oliver (1980) has argued, however, that disconfirmation alone is insufficient to explain satisfaction responses. Rather, he has proposed that satisfaction is also directly related to pre-choice expectations, independently of the effect of disconfirmation. Pre-choice expectations act as a frame of reference for post-choice disconfirmation. Outcomes poorer than expected are evaluated below the reference point, while those better than expected are evaluated above this base. Theoretical support for this proposition derives from Helson (1959), whose seminal work in perception suggested that stimuli are perceived only in relation to an adapted standard which sustains perceptions in the general vicinity of the standard. Oliver's formulation, then, proposes that satisfaction results from an additive combination or the expectation level and subsequent disconfirmation. Empirical support for these propositions has been reported by Swan (1977), Linda and Oliver (1979), and Oliver (1980).

Despite its widespread adoption in the consumer satisfaction literature, the disconfirmation of expectations model may be challenged. First, its explanation of satisfaction in some circumstances is less than convincing. As suggested by LaTour and Peat (1979), consider the situation in which a consumer has no choice but purchase a product or brand which is liked, but only second- or third-best, as might be the case when the preferred brands were not available. In this case, despite moderately high expectations and positive disconfirmation through actual product usage, the consumer may still be less than fully satisfied since the preferred brand was not consumed Proponents or the disconfirmation of expectations model would argue, that in such cases, the lower level of satisfaction due to the lower "anchor" established by the initial expectation level for the second or third ranking brand and the inability of even a positive disconfirmation effect to overcome these lower expectations. While plausible, this explanation may be unnecessarily complex. Moreover, it fails to offer an explanation when the expectation effect will be greater than, equal to, or lesser than the disconfirmation effect in producing satisfaction judgments.

Again using an example developed by LaTour and Peat (1979), consider the case in which a new brand is introduced whose attributes the consumer evaluates more favorably than those of the other brands presently on the market. Even if its marketer had created expectations that were too high, such that when the product was consumed expectations were negatively disconfirmed, would satisfaction result? The expectancy disconfirmation model makes no clear prediction. Rather, the occurrence of satisfaction or dissatisfaction depends on the relative strength of the expectation and disconfirmation effects, whose variation over consumers or purchase situations is not explained.

These examples suggest that a major problem with the disconfirmation of expectations model is that it does not provide sufficient differentiation between cognitive and evaluative notions. Expectations, in the true sense, refer to beliefs about what will occur in the future. What is expected in a product, however, may or may not correspond to what is wanted or desired in that product. Conversely, that which is valued may or may not correspond to what is expected. Product breakdowns, improper function, or unattractive appearance produce dissatisfaction regardless of whether they are expected. In practice. values and expectations often coincide because consumers choose purposefully to achieve their goals. When values and expectations have been separated experimentally. values rather than expectations are seen to determine satisfaction (Locke 1967). In other words, success in relation to aspirations -- not to expectations -- seems primarily responsible for the positive affective response of satisfaction.

When product outcomes create positive disconfirmations, satisfaction results if the disconfirmation is in the direction of what one values or desires. If the disconfirmation is negative, and in the direction of what one disvalues, dissatisfaction will result. The disconfirmation effect may be an artifact of the association between (a) positive disconfirmation and achievement of one's values and (b) negative disconfirmation and blockage or one's values (Locke 1967).

A further problem with the disconfirmation or expectations model is the assumption that the evaluation or product usage and consumption experiences is limited only to those beliefs for which expectations had been formulated prior to purchase. Consumers often appear to report satisfaction (or dissatisfaction) based on aspects of the product which may not have materialized until after purchase and consumption, i.e., for which expectations never existed.

The Value-Percept Disparity Model

Formulated originally by Locke (1967, 1969), this model asserts that satisfaction/dissatisfaction is an emotional response triggered by a cognitive-evaluative process in which the perceptions of (or beliefs about) an object, action, or condition are compared to one's values (or needs, wants, desires). The smaller the disparity between percepts of the object, action, or condition, and one's values. the more favorable the evaluation, and the greater the generation of positive affect associated with goal attainment, i.e., satisfaction. Conversely, the greater the value-percept disparity, the less favorable the evaluation, the less the generation of positive affect, and the greater the generation of negative affect associated with goal frustration, i.e., dissatisfaction.

The approach requires only three elements: (1) one or more perceptions of the aspects of a product or institution or marketplace behavior; (2) one or more value standards held by the consumer; and (3) a conscious or subconscious judgment of the relationship between one's perception(s) and one's value(s).

Since it is the attainment of values which consumers seek, rather than the confirmation of their expectations, it is posited that perceptions of products, institutions, or marketplace actions are simply tested against the extent to which then meet the consumer's values. While products may provide more of an attribute or outcome than what is desired, this has no impact on satisfaction unless this aspect causes a blockage in the attainment of another value.

Test of Satisfaction Models

The conceptual advantages and theoretical parsimony of the value-percept disparity model suggest further scrutiny of the theory in an empirical setting. In particular, does the value-percept disparity model succeed in explaining consumer satisfaction/dissatisfaction better than the commonly used disconfirmation of expectations model? Addressing this question requires an empirical study to provide observations of all constructs in both theories, followed by a strong inference test of which competing model provides the better explanation of observed satisfaction responses.

METHODOLOGY

Source of Data

A convenience sample of 72 undergraduate business students was surveyed using two separate instruments, administered approximately two weeks apart and portrayed as different research investigations. Each instrument was a self-administered questionnaire, completed during class session. Participation was voluntary; while no subjects refused to complete either instrument, there were six individuals excluded since data was not obtained on both instruments, reducing the effective sample size to 66. To provide a setting for the investigation of the competing models appropriate to the student subject population, the two surveys concerned respondents' experiences with their automobiles (cars, trucks, or off-the-road vehicles).

Measures

Multiple measures of the various constructs comprising each theory were made. Satisfaction with automobiles was measured with three indicators: (a) the 7-point "Delighted-Terrible" scale (Westbrook 1980), (b) a "Very Satisfied" (7) to "Not at all Satisfied" (1) itemized rating scale, and (c) a "Very Dissatisfied" (7) to "Not at all Dissatisfied" (1) itemized rating scale. These three criterion measures were administered two weeks prior to all other measures and presented as a separate study to minimize method variance in common with the other pencil and paper measures of explanatory constructs.

Level of expectations relative to one's automobile was measured by a two-step procedure: First, respondents were asked to recall in writing the specific characteristics of the automobile which they expected it to have at the time it was purchased. Next, they were asked to provide an overall rating of the level of these expectations, on an 11-point semantic differential-type scale ranging from "A Great Deal" to "Very Little." Disconfirmation of expectations was measured with three items as suggested by Oliver (1980), each requiring respondents to provide a rating of the extent to which the automobile had turned out as expected. One item dealt with the advantages, benefits, and good points of the car, another with the disadvantages, problems, and bad points, and the third-with all aspects of the car considered at once. Ratings were made on an itemized, fully anchored 7-point rating scale whose extremes ranged from "Much Better than Expected" (7) to "Much Worse than Expected (1) with "About as Expected" (4) at the midpoint.

While both disconfirmation and expectation were measured at a time later than satisfaction measurement, to minimize common method variance, the fact that these assessments did not coincide with the actual order of their occurrence is not problematic. Consumers are fully capable of retrieving semantic information from long term memory, such as a level of global expectations or whether their expectations were disconfirmed, regardless of the order in which they may be questioned about these data. While some memory loss may occur relative to detailed outcome expectations, it is not likely to present a problem for the recollection or global expectation and disconfirmation beliefs presumed salient by the theory.

Value-percept disparity was operationalized as the extent to which the respondent's automobile provided the features and performance characteristics needed or desired. To assure an appropriate frame of reference for this evaluation, respondents were required to reflect on their-"needs" in an automobile, as defined above, and to list as many of these as they could in a free-recall task. Then they were asked to rate the extent to which the automobile met these needs, on a interval semantic differential scale anchored with "Provides Far Less than my Needs" (7) and "Provides Exactly What I Need" (1).

Descriptive statistics for each measure and simple correlations between all pairs of measures are shown in Table 1.

Analytic Procedures

Since the purpose of the research was to test competing theories that can be specified as subsets of a more general model, the LISREL IV analysis methodology is appropriate. The applications of LISREL to theory construction have been well explicated (Bagozzi 1980). Advantages include the ability to handle multiple indicators of unmeasured latent constructs, simultaneous estimation of measurement error for indicators, and a > 9 estimate which describes how well a given theory fits observed data. LISREL requires the input of a covariance or correlation matrix which is then decomposed into two measurement models and a structural model. To determine which of the competing models better fits the data, the variable correlation matrix in Table 1 was submitted to two iterations of LISREL IV. The results of this analysis for each of the postulated models of satisfaction is shown in Figure 1.

TABLE 1

DESCRIPTIVE STATISTICS AND SIMPLE CORRELATIONS

RESULTS AND DISCUSSION

The results of the LISREL IV analysis indicate that the disconrirmation of expectations model fits the data better than the value-percept disparity model, as evidenced by their respective observed values of X2 relative to their respective degrees of freedom. However, it is also apparent that neither of the tested models fits the data well by absolute standards. Thus. it could not be concluded that either theory adequateLy explained the data.

FIGURE 1

LISREL IV ESTIMATES FOR COMPETING SATISFACTION MODELS

Analysis of the stress values, the residuals, and the first derivatives indicated that the model fit could be improved by relaxing the constraints such that all three of the exogenous constructs could causally load on the satisfaction construct. Accordingly, a hybrid model was postulated, and another LISREL IV analysis was undertaken in which all three structural parameters (gamma values) were estimated. The results or this effort are diagrammed in Figure 2.

As can be seen from Figure 2, the hybrid model fits substantially better than the value-percept disparity model. Since X2 is asymptotically additive we can approximate the X2 value associated with the difference between the hybrid model and the value-percept disparity model as 81.17-38.66 = 42.51, with 20-18 = 2 degrees of freedom (p < .0000). Similarly, the hybrid model represents a marginal improvement over the disconfirmation of expectations model: X2 = 41.78-38.66 = 3.12, with 19-18 = 1 degree of freedom (p = .07). However, even the hybrid model still does not meet the usual criterion of acceptable fit. These findings must be viewed as surprising, inasmuch as the disconfirmation of expectations model has received empirical support in the literature and represents the cominant conceptual paradigm for studying consumer satisfaction. Moreover, it is surprising that the value-percept disparity model alone, despite its conceptual advantages over the disconfirmation model, does not succeed in fitting the data well either. The marginal improvements offered by the hybrid model incorporating disconfirmation, expectation, and value-percept disparity constructs suggest that the processes determining satisfaction may be more complex than typically assumed, perhaps requiring an integration of the theoretical streams underlying each separate model evaluated in this paper.

FIGURE 2

LISREL IV ESTIMATES FOR A HYBRID SATISFACTION MODEL

These observations must be tempered, however, by the limitations of the present study regarding construct measurement. The measure of expectations was different than Oliver's (1980) in several important respects. First, it-involved retrospective estimation by respondents or their pre-choice expectation level, instead of a pre-choice assessment of the likelihood of obtaining particular product outcomes weighted by their evaluations by respondents. Second, it measured the global level of expectations rather than separate measurements .or each product outcome. Measurement at the global level rather than individual outcome level assumes, of course, that consumers are capable of performing the cognitive algebra of information integration readily and with accuracy. While it would have been desirable to measure expectations prior to purchase for each salient outcome, the measurement strategies employed here were dictated by the cross-sectional nature of the research. In any event, the expectation measure employed in this study should be validated against the measure employed by Oliver (1980). If the results of similar validation efforts for corresponding measurements of the construct of disconfirmation are any guide (Swan and Trawick 1980), a reasonable correspondence between the expectation measure used here and Oliver's would be expected. Perhaps the most convincing support for the expectation measure used in this study, however, is its nomological validity. Despite the differences noted above, the measure of expectations behaved according to theory, as evidenced by its independence from the disconfirmation measures (simple correlations averaging .08, n.s.).

The differences between Oliver's measurement expectations and the approach followed here provide some interesting insight into the operation of the disconfirmation Of expectations model. Expectations, in the true sense or the concept, refer to the subjectively perceived likelihood of obtaining one or more particular outcomes. In Oliver's (1980, 1981) formulation. however, the concept of expectation is also defined to include an evaluation of the outcome(s) and, as such, may be viewed, as the cognitive aspect of attitude in the multiattribute formulation. The fact that this study did not support the disconfirmation of expectations model when a "pure" expectation measure was used as an indicator of the construct suggests that the evaluative component of the measure employed by Oliver (1980) was responsible for his results which did support the model. Thus, the expectation effect in the model may be more a product of what was wanted or desired than the likelihood of getting those things. If this is true, one must question whether the concept of expectation as a prediction of future outcomes is really necessary in modeling consumer satisfaction.

Measurement error was also detected in the LISREL analysis relative to--the disconfirmation construct. The convergent validity (l) coefficients for two of the indicators are model, indicating that they possess a considerable amount of unique error variance. The simple correlations among the three measures are comparable to the results reported in previous satisfaction studies.

Perhaps the most serious of potential measurement problems pertains to the construct of value-percept disparity. While the measurement approach taken in this research for the product under study (automobiles) appears to have some content validity, its actual measurement error is unknown. The LISREL analysis assumed this indicator was error-free, for lack or parallel indicators of the construct. It is conceivable that the sole indicator of value-percept disparity did not in fact adequately represent the construct, thereby biasing downward the test of the theory. Obviously, a priority for future efforts in this area is the development of validated multiple measures of the construct.

These considerations point squarely to the need for further work on the measurement of constructs theorized to determine consumer satisfaction. Some progress has already been made in improving the quality of satisfaction measures (Westbrook 1980; Westbrook and Oliver 1980), and it is clear that similar efforts would be helpful for the constructs of value-percept disparity, level of expectations, and disconfirmation. No doubt the advent of improved measures will enhance the ability of the two competing models to explain the observed variations in satisfaction. That they did not tare well in this research, even notwithstanding measurement limitations. is disturbing. On the basis of these results, it might be expected that rigorous future efforts may confirm the need suggested here for amplifying and extending current theories of the processes responsible for consumer satisfaction/dissatisfaction.

REFERENCES

Bagozzi, Richard P. (1980), Causal Models in Marketing, New York: John Wiley & Sons.

Branden, N. (1966), "Emotions and Values," The Objectivist, 5 (5), 1-11.

Cardozo, Richard N. (1968), "An Experimental Study of Consumer Effort, Expectation, and Satisfaction," Journal of Marketing Research, 2 (August), 244-249.

Cohen, Joel B. and Goldberg, Marvin E. (1970), "The Dissonance Model in Post-Decision Product Evaluation," Journal of Marketing Research, 7 (August), 315-321.

Day, Ralph L. (1977), "Extending the Concept of Consumer Satisfaction," in Advances in Consumer Research, Vol. 4, William D. Perreault, Jr., ed., Ann Arbor: Association for Consumer Research, 149-154.

Engel, James F. and Blackwell, Roger D. (1982), Consumer Behavior, New York: Hole, Rinehard and Winston.

Helson, Harry (1959), "Adaptation Level Theory," in Psychology: A Study of a Science, Vol. 1, Sigmund Koch, ed., New York: McGraw Hill Co.

Howard, John A. and Sheth, Jagdish N. (1969), The Theory of Buyer Behavior, New York: John Wiley & Sons.

Hunt, H. Keith (1977), "CS/D -- Overview and Future Research Directions," in Conceptualization and Measurement or Consumer Satisfaction and Dissatisfaction, H. Keith Hunt, ed., Cambridge, Mass.: Marketing Science Institute.

Izard, Carroll E. (1977), Human Emotions, New York: Plenum.

LaTour, Steven A. and Peat, Nancy C. (1979), "Conceptual and Methodological Issues in Consumer Satisfaction Research," in Advances in Consumer Research, W. L. Wilkie, ed., Ann Arbor: Association for Consumer Research, 431-440.

Linda, Gerald and Oliver, Richard L. (1979), "Multiple Brand Analysis of Expectation and Disconfirmation Effects on Satisfaction," paper presented at the 87th Annual Convention of the American Psychological Association.

Locke, Edwin A. (1967), "Relationship of Success and Expectation to Affect on Goal Seeking Tasks," Journal of Personality and Social Psychology, 7 (2), 125-134.

Locke, Edwin A. (1969), "What is Job Satisfaction?," Organizational Behavior and Human Performance, 4, 309-336.

Plutchik, Robert (1980), "A General Psychoevolutional Theory of Emotion," in Emotion: Theory, Research and Experience, Vol. 1, Robert Plutchik and Henry Kellerman, eds ., New York: Academic Press.

Oliver, Richard L. (1977), "Effect of Expectations and Disconfirmation on Post Exposure Product Evaluations: An Alternative Interpretation," Journal of Applied Psychology, 62 (August), 480-486.

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

Oliver, Richard L. (1981), "Measurement and Evaluation of Satisfaction Processes in Retail Settings," Journal of Retailing, 57 (Fall).

Olson, Jerry C. and Dover, Philip (1976). "Effects of Expectation Creation and Disconfirmation on Belief Elements of Cognitive Structure," in Advances in Consumer Research, Vol. 3, Beverlee B. Anderson, ed., Chicago: Association for Consumer Research.

Swan, John (1977), "Consumer Satisfaction with a Retail Store Related to the Fulfillment of Expectations on an Initial Shopping Trip," in Consumer Satisfaction, Dissatisfaction and Complaining Behavior, Ralph L. Day, ed., Bloomington: School of Business, Indiana University, 10-17.

Swan, John E. and Martin, Warren S. (1981), "Testing Comparison Level and Predictive Expectations Models or Satisfaction," in Advances in Consumer Research, Vol. 8. Kent B. Monroe, ed., Ann Arbor: Association for Consumer Research, 77-82.

Westbrook, Robert A. (1980), "Intrapersonal Affective Influences upon Consumer Satisfaction," Journal of Consumer Research, 7 (June), 49-54.

Westbrook, Robert A. (1980), "A Rating Scale for Measuring Product/Service Satisfaction," Journal of Marketing.

Westbrook, Robert A. and Oliver, Richard L. (1980), "Developing Better Measures of Consumer Satisfaction: Some Preliminary Results," in Advances in Consumer Research, Vol. 1, Jerry C. Olson, ed., Ann Arbor: Association for Consumer Research.

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