A number of theoretical approaches have been utilized to explain the relationship between disconfirmation and satisfaction (Oliver, 1980b;

Anderson, 1973). These approaches can be viewed as variations of consistency theories and focus mainly on the nature of the consumer's post-usage comparison process (Oliver, 1980b). Consistency theories suggest that when expectations and the actual product performance do not match the consumer will feel some degree of tension. In order to relieve this tension the consumer will make adjustments either in expectations or in the perceptions of the product's actual performance. Four theoretical approaches have been advanced under the umbrella of consistency theory: 1. assimilation theory, 2. contrast theory, 3. assimilation-contrast theory, and 4. negativity theory. 1. Assimilation theory. Festinger's (1957) dissonance theory forms the basis of assimilation theory. Dissonance theory posits that consumers make some kind of cognitive comparison between expectations about the product and the perceived product performance. If there is a discrepancy between expectations and perceived product performance then dissonance arises. This view of the consumer post-usage evaluation was introduced into the satisfaction literature in the form of assimilation theory (Anderson, 1973). According to Anderson (1973), consumers seek to avoid dissonance by adjusting perceptions about a given product to bring it more in line with expectations. Consumers can also reduce the tension resulting from a discrepancy between expectations and product performance either by distorting expectations so that they coincide with perceived product performance or by raising the level of satisfaction by minimizing the relative importance of the disconfirmation experienced (Olson & Dover, 1979). Assimilation theory has a number of shortcomings. First, the approach assumes that there is a relationship between expectation and satisfaction but does not specify how disconfirmation of an expectation leads to either satisfaction or dissatisfaction. Second, the theory also assumes thatAllied Academies International Conference page 43

Proceedings of the Academy of Organizational Culture, Communications and Conflict, 7(2) Las Vegas, 2003 consumers are motivated enough to adjust either their expectations or their perceptions about the performance of the product (Forman, 1986). If the consumer adjusts either expectations or perceptions about product performance then dissatisfaction would not be an outcome of the post-usage evaluation process. A number of researchers have found that controlling for actual product performance can lead to a positive relationship between expectation and satisfaction (Olson & Dover, 1979; Anderson, 1973). Therefore, it would appear that dissatisfaction could never occur unless the evaluative process were to begin with negative consumer expectations (Bitner, 1987). 2. Contrast theory. Contrast theory, first introduced by Hovland, Harvey and Sherif (1957), presents an alternative view of the consumer post-usage evaluation process than was presented in assimilation theory in that post-usage evaluations lead to results in opposite predictions for the effects of expectations on satisfaction (Cardozo, 1965). The approach holds that when consumers experience disconfirmation they seek to minimize the discrepancy between prior expectations and actual product performance by shifting their evaluations away from the expectations. Dawes, Singer and Lemons (1972) define contrast theory as the tendency to magnify the discrepancy between "one's own attitudes and the attitudes represented by opinion statements" (p. 281) endorsed by persons with opposing views. While assimilation theory posits that consumers will seek to minimize the discrepancy between expectation and performance, contrast theory holds that a surprise effect occurs leading to the discrepancy being magnified or exaggerated. 3. Assimilation-contrast theory. The assimilation-contrast theory has been proposed as yet another way to explain the relationships among the variables in the disconfirmation model (Hovland, Harvey & Sherif, 1957). A combination of both the assimilation and the contrast theories, this paradigm posits that satisfaction is a function of the magnitude of the discrepancy between expected

and perceived performance. Generally, consumers have zones or latitudes of acceptance or rejection with respect to their perceptions. As with assimilation theory, the consumers will tend to assimilate or adjust differences in perceptions about product performance to bring it in line with prior expectations, but only if the discrepancy is relatively small. When there is a large discrepancy between expectations and perceived performance, contrast effects occur and the consumer tends to magnify the perceived difference. Whether assimilation or contrast occurs depends upon the perceived disparity between expectations and actual product performance. Arguing that Cardozo's (1965) attempt at reconciling the two earlier theories was methodologically flawed, Anderson (1973) presented his adaptation of an earlier work. He asserted that consumers possess a "noticeable difference" disconfirmation threshold. Assimilation-contrast theory attempts illustrate that both the assimilation and the contrast theory paradigms have applicability in the study of consumer satisfaction. The approach makes it possible to " hypothesize variables other than the magnitude of the discrepancy that might also influence whether the assimilation effect or the contrast effect would be observed when product performance is difficult to judge, expectations may dominate and assimilation effects will be observed contrast effects would result in high involvement circumstances. The strength of the expectations may also affect whether assimilation or contrast effects are observed" (Bitner, 1987, p. 13). Researchers attempting to empirically test this theory have met with mixed results. For example, Olson and Dover (1979) and Anderson (1973) found some evidence to support the assimilation theory approach. In discussing both of these studies, however, Oliver (1980a) arguespage 44 Allied Academies International Conference Las Vegas, 2003 Proceedings of the Academy of Organizational Culture, Communications and Conflict, 7(2) that they only measured expectations and assumed that there were perceptual differences between disconfirmation or satisfaction. This criticism is of some significance because researchers do not actually measure satisfaction or dissatisfaction. Instead, researchers generally believed that it is the perception of disconfirmation that leads to satisfaction or dissatisfaction (Forman, 1986).

In contradiction to the findings supporting the assimilation perspective, Cadotte, Woodruff and Jenkins (1983) reported negative correlation between expectation and disconfirmation. They therefore concluded that satisfaction is really an additive function of the two concepts. Further ambiguity is created by results of studies which found no relationship between expectation and disconfirmation (Oliver, 1977a, 1977b, 1979). 4. Negativity theory. Like the three previous theories, negativity theory has its foundations in the disconfirmation process. Introduced into the consumer satisfaction literature by Anderson (1973), negativity theory posits that when expectations are strongly held, consumers will respond negatively to any disconfirmation. Accordingly, dissatisfaction will occur if perceived performance is less than expectations or if perceived performance exceeds expectations (Carlsmith & Aronson, 1963; Anderson, 1973). CONCLUSION The satisfaction construct has been defined in a number of different ways. Researchers have defined satisfaction in terms of need fulfillment, pleasure/displeasure, cognitive state, attribute or benefit evaluation, and subjective evaluation of experience. Currently, satisfaction is viewed as an emotional response to a product experience. However, while researchers have used a number of different definitions for satisfaction, they generally agreeALTERNATIVE DEFINITIONS OF SATISFACTION: IDENTIFYING A CONSENSUS A three-step approach, loosely based on grounded theory (Glaser and Strauss 1967; Strauss and Corbin 1990), was used to discover and formulate different definitions of consumer satisfaction. First, because of the plethora of satisfaction research, a literature review was conducted to glean general consistencies and specific inconsistencies in the definitions used by marketing researchers. Second, group interviews provided exploratory data to refute, confirm, and/or augment the evolving definition. Our goal was to use consumer perspectives to expand and question emerging definitions from the literature. Group interviews were conducted to elicit descriptions of satisfaction pertinent to consumer-specified situations. In addition, group interviews provided direction for the next interviewing phase. Third, personal interviewees

provided their generic definitions of consumer satisfaction. Consistent with the procedures of grounded theory, personal interviews were designed to verify, refute, and further refine the emerging components of consumer satisfaction. Personal interviewees also provided data needed for assessing the generalizability of the emerging definitional framework of satisfaction across different contexts. The input of actual consumers is critical for developing a meaningful definitional framework since ultimately, it is the consumers whom we will ask to answer questions about satisfaction. Research Design The literature review consisted of an examination of 20 definitions used during a 30-year period of consumer satisfaction that satisfaction involved a se