Tolerance to inferiority and tolerance disconfirmation:
A model of consumer dissatisfaction
Submitting author: Lida Lingling Zhang
Affiliation: University of Macau
OtherAuthors: Chow, Clement S. F.
Affiliations: University of Macau
Abstract
Although the disconfirmation paradigm dominates the consumer satisfaction literature, its current models do not address the specific nature of consumer dissatisfaction. In this paper we propose a model that captures the formation process of consumer dissatisfaction under the disconfirmation paradigm. We draw from dispositional (tolerance to inferiority) and cognitive (tolerance disconfirmation) constructs and assess the model in a laboratory experiment. Results offer evidence for the validity of the model and hold several implications for flexible consumer dissatisfaction management.
Keywords: Tolerance to inferiority; Tolerance disconfirmation; Consumer dissatisfaction; Product performance
Track: Consumer behaviour
1. Introduction
Consumer satisfaction has aroused wide interest in marketing research, which has shown that product performance outcomes affect consumer satisfaction levels. But how? The disconfirmation process is the current dominant paradigm and it offers several explanations. Some of these include the degree to which actual product performance is perceived to be (1) better than predicted performance (e.g., Droge & Halstead, 1991; Oliver, 1980; Tsiros, Mittal, & Ross, 2004), (2) better than one’s normative beliefs about the performance (e.g., Cadotte, Woodruff, & Jenkins, 1987), (3) better than the performance one believes is deserved given invested costs (e.g., Tse & Wilton, 1988), and (4) better than one’s personal desires for performance (e.g., Spreng, MacKenzie, & Olshavsky, 1996). Each of these explanations helps qualify and mediate the relation between actual product performance and consumer satisfaction.
As a construct, consumer satisfaction is an affective state (Giese & Cote, 2000) resulting from fulfillment judgment (Oliver, 1997). Need fulfillment underlies consumer satisfaction and depicts the nature of the construct in light of two basic sets of needs (Herzberg, 1966). One set – need to avoid pain – stems from one’s animal disposition; the other set – need to approach growth – stems from one’s human disposition. Individuals are motivated to avoid pain in a negative psychological state and motivated to approach growth in a neutral or positive psychological state. Whereas the need for growth is germane to consumer satisfaction, the need for pain alleviation is germane to consumer dissatisfaction.
Dissatisfaction is a negative state. Fulfillment of the need for pain alleviation results in less dissatisfaction. The formation process of consumer dissatisfaction has distinct components. A number of studies (e.g., Anderson & Mittal, 2000; Johnston, 1995; Leavitt, 1977; Maddox, 1981) have shown that distinct product attributes fulfilling the need for pain alleviation contribute to consumer dissatisfaction. However, these research efforts are limited to product performance and do not extend to consumer psychological processes. The psychological components mediating product performance and consumer dissatisfaction defines the undeveloped area, which we address in this study. Understanding consumer dissatisfaction is practically important because it can lead to the consumer responses that impinge upon business performance, such as complaints, switching, and negative word of mouth (e.g., Bougie, Pieters, & Zeelenberg, 2003; Hirschman, 1970; Maute & Forrester, 1993; Singh, 1988; Singh & Pandya, 1991). Understanding consumer dissatisfaction enables marketers to design flexible and effective marketing strategies. As such, the purpose of this paper is to build on the disconfirmation paradigm by developing and empirically assessing a model of the principal psychological antecedents of consumer dissatisfaction.
2. Conceptual framework
2.1. Two constructs: tolerance to inferiority and tolerance disconfirmation
Miller (1977) suggests that minimum tolerable performance[1] can be a referent in consumer satisfaction and that it reflects a bottom-level standard of least acceptable performance. Zeithaml, Berry, and Parasuraman (1993) suggest five determinants of the bottom-level standard – situational factors, perceived alternatives, transitory intensifiers, consumer self-perceived role, and predicted performance. Though the five factors are expansive, one other factor for explaining minimum tolerable performance is neglected by that model: consumer’s dispositional characteristics, which may include tolerance of inferior performance – a trait that varies across individuals. Baumgartner (2002) calls for developing dispositional concepts in consumer contexts to integrate the conceptually fragmented consumer behavior research by taking greater advantage of personality psychology. Responding to this call, we offer “tolerance to inferiority” as a sixth dimension of minimum tolerable performance.
Tolerance to inferiority is the tendency to withstand psychological discomfort arising from poor performance. It is a dispositional concept as it describes characteristic responses. Those with high tolerance to inferiority tend to accept poor performance. Zeithaml et al. (1993, p. 6) propose a similar but different concept that represents the difference between the level of desired performance and the level of minimum tolerable performance. As the authors explain, this “zone of tolerance” reflects the “extent to which customers recognize and are willing to accept heterogeneity.” Tolerance to inferiority, by contrast, represents the degree to which a consumer tends to endure and accept inferior product performance.
Tolerance disconfirmation is the degree to which a consumer perceives product inferiority to be beyond his/her maximum tolerable inferiority, or the degree to which a consumer perceives product performance to be worse than his/her minimum tolerable performance. The construct reflects a perception of difference.
2.2. Hypotheses
With the above two constructs, tolerance to inferiority and tolerance disconfirmation, we develop a disconfirmation model that explains the relation between product performance and consumer dissatisfaction (see Fig. 1).
Fig. 1. Conceptual model.
It is well established that the worse a product performs, the more likely it is that a consumer will feel dissatisfied with it.
H1: Product performance negatively affects consumer dissatisfaction.
According to Helson’s (1959, 1948) adaptation level theory, products are judged in relation to referents. Minimum tolerable performance is a diagnostic referent for consumer dissatisfaction as it is informative on whether a consumer feels painful at certain low level of performance. Therefore, tolerance disconfirmation is reasoned to mediate the relation between product performance and consumer dissatisfaction. Moreover, the mediation is more likely to be partial than full as disconfirmation is not the only mediating psychological process (Oliver, 1997).
H2a: Product performance negatively affects tolerance disconfirmation.
H2b: Tolerance disconfirmation positively affects consumer dissatisfaction.
H2c: Tolerance disconfirmation partially mediates the relation between product performance and consumer dissatisfaction.
The minimum difference between a standard and a stimulus necessary for detection is called the “just noticeable difference” (Coren, Porac, & Ward, 1978). According to Weber’s Law, it is not a constant value; it increases as the standard (i.e. original intensity) increases. Tolerant consumers are capable of enduring psychological discomfort arising from poor product performance, which means tolerant consumers have sufficient psychological resources to deal with the discomfort. Thus, product performance needs to increase to a great degree for the consumers to perceive an increase in their psychological resources and a fulfillment of their need for discomfort alleviation. In other words, large performance increase is necessary for the tolerant consumers to feel less dissatisfied. By contrast, intolerant consumers have insufficient psychological resources and thus can perceive the difference even when performance increases a little. In other words, only small performance increase is necessary for the intolerant consumers to feel less dissatisfied. Overall, consumer dissatisfaction is sensitive to performance increase more for consumers with low tolerance to inferiority.
H3: The relation between product performance and consumer dissatisfaction is stronger for consumers with low tolerance to inferiority.
Tolerance disconfirmation is a perception of the degree to which actual product inferiority is beyond one’s maximum tolerable inferiority. For instance, suppose the intensity of product inferiority and that of maximum tolerable inferiority are same in the beginning. Since the maximum tolerable inferiority of tolerant consumers is high, actual performance inferiority needs to increase a lot for them to perceive the difference. By contrast, since the maximum tolerable inferiority of intolerant consumers is low, only small increase in performance inferiority is necessary for them to perceive the difference. Therefore, tolerance disconfirmation is sensitive to performance more for consumers with low tolerance to inferiority.
H4: The relation between product performance and tolerance disconfirmation is stronger for consumers with low tolerance to inferiority.
3. Method
We undertook a laboratory experiment with a 2 (high/low tolerance to inferiority) x 2 (high/low product performance) between-subject full factorial design. A sample of 242 undergraduate students were asked to imagine that they had just graduated from high school, had been admitted into college, and their parents were going to rent an apartments for them. After exposed to manipulation materials, the subjects completed measures.
3.1. Manipulations
A manipulation was developed to strengthen, not to contradict, one’s existing tolerance to inferiority level. Before subjects were exposed to experiment setting, their tolerance to inferiority in apartment consumption contexts was pre-measured. 7-point Likert-type scales were developed for the pre-measures. Based on pre-test scores, half the subjects (tolerant) were assigned to high tolerance treatment groups and the other half (intolerant) were assigned to low tolerance treatment groups.
Subjects were exposed to one of two different contrived market research reports released by the housing bureau about tolerance to inferiority in apartment consumption contexts. The high tolerance to inferiority version described that most young people were very tolerant to living in an inferior apartment. The low tolerance version described that most young people were not tolerant at all to living in an inferior apartment. Subjects were also exposed to a flash file presenting the results of a fictitious questionnaire survey conducted in the university supporting the findings in the housing bureau report.
Both tolerant subjects and intolerant subjects were randomly assigned to high or low performance treatment conditions. They were exposed to one of two video and audio descriptions of an apartment. The high-performance version presented a good apartment with some amenities. The low-performance version presented a poor apartment with few amenities.
3.2. Measures
Consumer dissatisfaction was measured with twelve items adapted from Cadotte et al. (1987) and Spreng et al. (1996). Subjects were asked to express their feelings about the presented apartment. 7-point Likert scales were used (Cronbach’s Alpha = .98, n = 214; see A.1).
Tolerance disconfirmation measures were adapted from Oliver (1997). Subjects were asked to compare the presented apartment performance and the worst apartment performance they could tolerate, and then evaluate their difference. 7-point Likert-type scales for attribute evaluation were used (Cronbach’s Alpha = .96, n = 217; see A.2).
Tolerance to inferiority measures for the manipulation check (Cronbach’s Alpha = .84, n = 215; see A.3) were same as those used for the pre-measures (Cronbach’s Alpha = .77, n = 277; see A.3). Perceived performance was also measured for the manipulation check. Subjects were asked to express the degree to which they thought the apartment performed well with good features. 7-point Likert-type scales for attribute evaluation were developed (Cronbach’s Alpha = .99, n = 217; see A.4).
4. Major Results
Both tolerance to inferiority manipulation and product performance manipulation are effective.
ANOVA results of consumer dissatisfaction reveal a significant main effect of performance treatment (F(1,210) = 1337.51, p < .01), a significant main effect of tolerance treatment (F(1,210) = 13.89, p < .01), and a significant interaction effect (F(1,210) = 21.82, p < .01). Simple effect analyses (Keppel & Zedeck, 1989) show that (1) the subjects with high tolerance to inferiority are less dissatisfied when apartment performs relatively good (M = 1.95) than when it performs relatively bad (M = 4.96, F(1,210) = 493.25, p < .01), (2) the subjects with low intolerance to inferiority are also less dissatisfied when apartment performs relatively good (M = 1.86) than when it performs relatively bad (M = 5.76, F(1,210) = 878.23, p < .01), and (3) the impact of performance on dissatisfaction is stronger for the subjects with low tolerance to inferiority (MS = 413.60) than for those with high tolerance inferiority (MS = 232.30) (see B. 1). Therefore, H1 and H3 are supported.
ANOVA results of tolerance disconfirmation reveal a significant main effect of performance treatment (F(1,213) = 501.09, p < .01), a significant main effect of tolerance treatment (F(1,213) = 6.02, p = .02), and a significant interaction effect (F(1,213) = 13.63, p < .01). Simple effect analyses of tolerance disconfirmation show that (1) the subjects with high tolerance to inferiority are less likely to perceive tolerance disconfirmation when apartment performs relatively good (M = 2.50) than when it performs relatively bad (M = 4.60, F(1,213) = 172.52, p < .01), (2) the subjects with low tolerance to inferiority are also less likely to perceive tolerance disconfirmation when apartment performs relatively good (M = 2.37) than when it performs relatively bad (M = 5.28, F(1,213) = 344.38, p < .01), and (3) the impact of performance treatment on consumer disconfirmation is stronger for the subjects with low tolerance to inferiority (MS = 231.84) than for those with high tolerance to inferiority (MS = 116.14) (see B.2). Therefore, H4 is supported.
Baron and Kenny’s (1986) approach was used to test the mediating effect of tolerance disconfirmation in the relation between product performance and consumer dissatisfaction. ANCOVA results of consumer dissatisfaction show that tolerance disconfirmation is a significant covariate (F(1,209) = 5.55, p = .02; tolerance disconfirmation and consumer dissatisfaction are positively correlated: r = .82, p < .01). Tolerance disconfirmation reduces the main effect of performance treatment (from F(1,210) = 1337.51, p < .01 to F(1,209) = 332.15, p < .01; 76% reduction in the mean squares from 629.90 to 153.11). Therefore, H2a, H2b, and H2c are supported.
5. Implications
Results provide evidence supporting our conceptual model (see Fig. 1). Consumer tolerance to inferiority moderates the relation between product performance and tolerance disconfirmation and so moderates the relation between product performance and consumer dissatisfaction. Although not hypothesized in the conceptual model, it is also found that there is no any relation between consumer tolerance to inferiority and consumer dissatisfaction if product performs relatively well (F(1,210) = .48, p = .49) and there is significant negative relation between them if product performs relatively bad (F(1,210) = 32.94, p < .01). A possible explanation is that when product performs well, there is nothing inferior to tolerate and so one’s capacity to tolerate inferiority does not make any difference; when product performs bad, those tolerant to the inferior performance feel less dissatisfied.