the Impact of Perceived Justice on Customer Satisfaction and Intention to Complain in a Service Recovery

Mary Ann Hocutt, Oklahoma State University
Goutam Chakraborty, Oklahoma State University
John C. Mowen, Oklahoma State University
ABSTRACT - This study investigates how service recovery efforts influence the level of dissatisfaction and intention to complain following a service failure. Results from an experiment show that consumers’ satisfaction (intention to complain) are influenced by cause of failure and by perceived distributive and interactional justice in the recovery attempt. Interestingly, when the customer causes the service failure, satisfaction (complaint) levels are higher (lower) after service recovery efforts than in situations where no service failure occurs. Thus a prompt, courteous service recovery effort can have a significant impact on how a customer feels toward a service provider even after a service failure.
[ to cite ]:
Mary Ann Hocutt, Goutam Chakraborty, and John C. Mowen (1997) ,"the Impact of Perceived Justice on Customer Satisfaction and Intention to Complain in a Service Recovery", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 457-463.

Advances in Consumer Research Volume 24, 1997      Pages 457-463

THE IMPACT OF PERCEIVED JUSTICE ON CUSTOMER SATISFACTION AND INTENTION TO COMPLAIN IN A SERVICE RECOVERY

Mary Ann Hocutt, Oklahoma State University

Goutam Chakraborty, Oklahoma State University

John C. Mowen, Oklahoma State University

ABSTRACT -

This study investigates how service recovery efforts influence the level of dissatisfaction and intention to complain following a service failure. Results from an experiment show that consumers’ satisfaction (intention to complain) are influenced by cause of failure and by perceived distributive and interactional justice in the recovery attempt. Interestingly, when the customer causes the service failure, satisfaction (complaint) levels are higher (lower) after service recovery efforts than in situations where no service failure occurs. Thus a prompt, courteous service recovery effort can have a significant impact on how a customer feels toward a service provider even after a service failure.

Firms that provide highr levels of service gain higher levels of profits than those that do not. In the case of service organizations, consumer satisfaction and loyalty may be determined by the quality of a single service encounter (Solomon, Surprenant, Czepiel, and Gutman 1985). One negative service encounter, or service failure, can result in consumers’ becoming dissatisfied. While many researchers have looked at consumer complaint behavior that results from dissatisfaction due to service failures, very little research has explored the impact that service recovery efforts may have on the level of dissatisfaction felt by consumers as a result of service failures (Hart, Heskett, and Sasser 1990).

A service encounter consists of the period of time during which the customer interacts with the employees of the service organization (Czepiel, Solomon, and Surprenant 1985). Most of the dimensions of service quality and/or satisfaction with services identified by Parasuraman, Zeithaml, and Berry (1988) relate directly to the human interaction element of service quality. From a consumer’s point of view, the interaction with service employees defines the service encounter. Therefore, it is important for service employees to be trained and empowered to relate to customers in ways that ensure effective service (Bitner, Booms, and Tetreault 1990; Bitner, Booms, and Mohr 1994).

A service failure is said to occur when the service experience falls short of our expectations (Bell and Zemke 1987). It has been suggested that a service failure is profoundly different from the failure of a tangible product (Albrecht 1988). A service usually provides a psychological and largely personal outcome, whereas a tangible product failure is usually impersonal in its impact on the customer. The notion that service failures have greater consequences than product failures was supported by a survey conducted by Andreasen and Best (1977). After surveying 2,400 metropolitan households, service industries were found to yield the lowest levels of satisfaction. Less than half of the consumers who experienced dissatisfaction complained; and of those who did complain, one-third resulted in resolutions that were unsatisfactory. An impressive stream of research concerning complaint behavior has attempted to look at the ways consumers respond to service failures. It is apparent that many service organizations have developed reactive service failure strategies that focus on complaint management rather than service recovery issues.

Service recovery consists of all the actions people may take to move a customer from a state of dissatisfaction to a state of satisfaction. Zemke (1993, p. 463) defined planned service recovery as:

. . . a thought-out, preplanned process for returning aggrieved customers to a state of satisfaction with the company or institution after a service . . . has failed to live up to expectations or promised performance.

Few service firms take a proactive approach in service recovery with respect to consumer satisfaction (Albrecht 1988). In addtion, many companies are not prepared to recover from service failures because service employees have not been trained in how to handle disgruntled customers during a service encounter (Berry and Parasuraman 1992).

In the case of a service failure and subsequent recovery, a consumer assesses satisfaction with the outcome of resolution efforts and then reevaluates the overall consumption experience in light of the degree of success or failure in obtaining redress (Andreasen 1977). Thus, the consequences of a service provider’s failure to resolve service delivery problems can be severe. Bitner et al. (1990) found that from the customer’s point of view, the largest percentage (42.9%) of dissatisfactory outcomes in service encounters was due to the employee’s response to service delivery system failures. Bitner et al. (1994) also found that from the service employee’s point of view, the vast majority (51.7%) of dissatisfying service encounters were the result of inadequate responses to service failures. Inadequate response to service failures also increases the likelihood that these disgruntled customers will complain about the incident. TARP (1986) research findings suggest that the average customer who has had a problem with an organization tells 8 to 10 people about it, and 90 percent say they will never do business with that company again.

The present study investigates both the level of satisfaction and the likelihood of complaint behavior that may result after a service recovery effort is undertaken in response to a service failure. Specifically, the empirical results of our research would enhance managerial understanding of the impact of three factors on the effectiveness of service recovery efforts: (1) the cause of the failure, (2) the outcome resolution resulting from the recovery attempt, and (3) the manner in which the recovery effort is executed.

THEORETICAL DEVELOPMENT AND HYPOTHESES

According to the disconfirmation paradigm, consumer satisfaction is the result of an evaluative process whereby a consumer compares his/her expectations of how the service should be performed with the actual performance of the service (Oliver 1980). Confirmation (leaving the consumer in a neutral state) occurs when the service is performed as expected. Negative disconfirmation (dissatisfaction) occurs when the service performance does not live up to prior expectations. During service encounters, consumers expect "zero-defects" in service delivery. Despite the service provider’s attempts to offer consistent, high-quality service to consumers, service failures may still occur because service delivery is heterogeneous across service encounters due to the variability in situational factors and individual differences between consumers and service employees (Singh 1991). Therefore, based on the disconfirmation paradigm, the following hypothesis is proposed:

H1: The level of consumer satisfaction (likelihood of complaint behavior) with a service encounter will be lower (higher) in the case of a service failure than in the case of no service failure.

Attribution theory provides a basis for understanding how consumers respond to service failures (Folkes 1984). Attribution theory predicts that the perceived reason for a service failure influences the level of consumer satisfaction. The service failure itself is not the only factor influencing the level of satisfaction; consumers try to determine why the failure occurred. The three causal dimensions of attribution theory include: stability (i.e., is the service failure likely to occur very often?), controllability (i.e., could the service failure have been avoided?), and locus (i.e., is the service failure the fault of the consumer or the service provider?). If a customer realizes that a service failure is due to his/her own fault, he/she would be less upset than if the service provider is to blame for the service failure. Therefore, based on the locus dimension of attribution theory, we would expect the locus of causality (other versus self) to influence the level of consumer satisfaction and intention to complain in the following way:

H2: After a service failure, the level of consumer satisfaction (likelihood of complaint behavior) will be higher (lower) if the consumer perceives the locus of causality for the failure to be self versus the service provider.

Researchers have found a relationship between equity and consumer satisfaction (Oliver and Swan 1989) and complaint behavior (Brown and Beltramini 1984). Building on the foundations of equity theory, researchers have suggested three dimensions of equity, which are based on perceptions of justice (Greenberg 1996). Two of these three dimensions are explored in the present study: (1) distributive justice that focuses on the perceived fairness of the outcome of the service encounter (Homans 1961), and (2) interactional justice that focuses on the perceived fairness of the manner in which the customer is treated throughout the service encounter (Bies and Moag 1986). We will not manipulate the third dimension of equity, procedural justice (the perceived fairness of the process used to rectify the service failure) since this dimension has been found to be difficult to manipulate in an experimental situation (Goodwin and Ross 1992).

According to Blodgett, Granbois, and Walters (1993), distributive justice refers to the perceived fairness of the redress offered by the service provider (i.e., whether the customer is offered an exchange, monetary refund, etc.). In the case of a service failure, consumers would expect the service provider to compensate them for any tangible loss they suffered as a result of that service failure. Customers may expect different levels of compensation depending on how severely the service failure affects them. An annoyed customer would expect a "fair fix" for the problem, while a consumer who feels "victimized" as a result of the service failure may expect some value-added atonement (Bell and Ridge 1992).

A distinction needs to be drawn between the actual service recovery outcome and the process of service recovery, which relates to how the service recovery is attempted. According to Berry (1986, p. 49),

Service firms that are great at problem resolutionCthat are accessible and respond with quickness . . . are far more likely to repair the damage done to their quality reputations than are the firms that take a casual . . . attitude.

It has also been shown that prompt complaint resolutions have resulted in more consumers continuing to do business with the organization (Albrecht and Zemke 1985). In addition to prompt service recovery, Goodwin and Ross (1992) found that one aspect of interactional fairness, an apology, seems particularly relevant to complaint resolution.

Therefore, based on the distributive and interactional justice dimensions of equity theory, one would expect that both the tangible outcome and the manner in which the problem was resolved would impact how satisfied customers would be and the likelihood that they would complain about the incident. Consequently, we believe that the level of distributive and interactional justice would impact consumers’ satisfaction and intention to complain in the following manner:

H3: After a service failure, levels of satisfaction (likelihood of complaint behavior) will be higher (lower) if consumers perceive high rather than low distributive justice in the service recovery attempt.

H4: After a service failure, levels of satisfaction (likelihood of complaint behavior) will be higher (lower) if consumers perceive high rather than low interactional justice in the service recovery attempt.

RESEARCH METHODOLOGY

Subjects

A convenience sample of 251 undergraduate marketing students (54% women and 46% men) from a large midwestern university participated in this study. Care was taken to choose a service encounter that would be relevant to a student sample. Specifically, we used a service failure in a Mexican restaurant for developing the stimuli in this study. Earlier research has shown that dissatisfying service encounters are particularly prevalent in restaurant settings (Bitner et al. 1990).

Experimental Design

The study used a 2 X 2 X 2 between-subjects factorial design. The factors are cause of failure (customer or restaurant), distributive justice (high or low), and interactional justice (high or low).

Procedure

Subjects were randomly assigned to one of the eight experimental conditions. In addition, there was a control condition that described the same restaurant experience as the other scenarios, but did not include a service failure situation. The subjects were asked to read one of the nine (eight plus one control) versions of a scenario describing a dining experience at a Mexican restaurant. The context of the core service failure described a situation in which a bowl of salsa was spilled onto the customer’s table and some of the salsa was also spilled onto the sleeve of the customer’s new jacket. The scenarios were pretested for believability, and the mean believability was 5.6 on a 7-point scale (1=not at all believable, 7=highly believable).

TABLE 1

MANOVA WITH SATISFACTION AND COMPLAINT BEHAVIOR AS THE DEPENDENT VARIABLES

Independent Variable Manipulations

Locus of causality attribution, distributive justice, and interactive justice manipulations were pretested. The manipulations were achieved by changing the scenario descriptions as follows:

Customer’s fault: "In your rush to try the extra-hot salsa, you accidentally drop the bowl on the table. Unfortunately, most of the salsa is spilled onto your table."

Restaurant’s fault: "As the waiter hands you the bowl, you seem to notice a crack in the handle of the bowl. Sure enough, as you are taking the bowl, the handle suddenly breaks off and most of the extra-hot salsa is spilled onto your table."

High distributive justice: ". . . the waiter brings a new bowl of extra-hot salsa."

Low distributive justice: ". . . the waiter is unable to bring you more extra-hot salsa because that was their last bowl."

High interactional justice: ". . . the waiter shows empathy and takes steps to help you immediately."

Low interactional justice: an irritated waiter must call the manager (which causes a delay in response) instead of helping you immediately.

Dependent Variable Measurements

The two dependent variables, satisfaction and intention to complain, were assessed using multiple measures to increase reliability. Eight Likert-type (strongly disagree=1, strongly agree=7) satisfaction scales were adapted from Oliver (1980) and Bitner (1990). Seven measures assessing intention to com plain were adapted from Singh (1991) and Blodgett et al. (1993). The multi-item satisfaction and intention to complain scale reliabilities (Cronbach’s alpha) were 0.93 and 0.87, respectively. Indices for satisfaction and complaint behavior were formed by averaging the respective multi-item measures.

RESULTS

Manipulation Checks

Checks of the perception of attribution for the service failure and the interactional and distributive justice questions revealed that subjects perceived these independent variables as intended. All manipulations were assessed on 7-point semantic differential scales. Results showed statistically significant differences (all with p<0.001) in the expected direction between group means on questions about perceptions of cause of failure (restaurant=4.85, customer=2.17), interactional justice (high=5.88, low=4.34), and distributive justice (high=5.50, low=4.07).

MANOVA

Results from MANOVA indicate that all three main effects are statistically significant at the multivariate as well as the univariate levels. In addition, the two-way interaction effect between the locus of causality and interactional justice are significant at the multivariate and the univariate levels. The relevant F-statistics and p-values for these effects from the univariate analysis are reported in Table 1. In Table 2, we report the means corresponding to the statistically significant main effects.

Hypothesis 1 states that the level of consumer satisfaction (likelihood of complaint behavior) with a service encounter will be lower (higher) in the case of no service failure than in the case of service failure. This hypothesis is partially supported. Results show that satisfaction is significantly higher (F=16.12, p<0.001) and the intention to complain is significantly lower (F=16.18, p<0.001) when there is no service failure versus when the service failure is the fault of the restaurant. However, there are no significant differences in levels of satisfaction or levels of intention to complain between the "no failure" condition and the situation where the service failure is the customer’s fault (satisfaction: F=0.26, p<0.610; intention to complain: F=0.09, p<0.768).

TABLE 2

MAIN EFFECT MEANS

Hypothesis 2 states that after a service failure, consumers would be less satisfied if the service failure is the fault of the restaurant rather than their own fault. Likewise, consumers would be more likely to complain if the service failure is the fault of the restaurant. The means reported in Table 2 are in the predicted direction for both satisfaction and intention to complain. The main effects of the locus of causality for comparing these means are statistically significant for both dependent variables. Predictions from hypothesis 2 are supported.

Hypothesis 3 states that after a service failure, levels of satisfaction (likelihood of complaint behavior) will be higher (lower) if consumers perceived high distributive justice. The means reported in Table 2 are in the predicted direction for both satisfaction and intention to complain. The main effects of distributive justice for comparing these means are statistically significant for both dependent variables. Thus, predictions from hypothesis 3 are supported.

Hypothesis 4 states that after a service failure, levels of satisfaction (likelihood of complaint behavior) will be higher (lower) if consumers perceived high interactional justice. The means reported in Table 2 are in the predicted direction for both satisfaction and intention to complain. The main effects of interactional justice for comparing these means are statistically significant for both dependent variables. Thus, predictions from hypothesis 4 are also supported.

Locus of Causality and Interactional Justice Interaction: While we did not predict any interaction between locus of causality, distributive justice, and interactional justice for either satisfaction or intention to complain, results in Table 1 indicate that the interaction between locus of causality and interactional justice is statistically significant for both dependent variables. The means for interpreting this interaction effect are shown in Figure 1.

The interaction plot shows that regardless of who is to blame for the service failure, there are statistically significant differences in the mean levels of satisfaction between subjects in the low and high interactional justice conditions (cutomer’s fault: F=38.91, p<0.001; restaurant’s fault: F=12.48, p<0.001). However, the difference between low and high interactional justice conditions is higher when the failure is due to customer’s fault rather than restaurant’s fault. Intention to complain varies depending on whether the restaurant or the customer is at fault for the service failure. When the service failure is the restaurant’s fault, there is no significant difference between average intention to complain between subjects in the low and high interactional justice conditions (F=1.17, p<0.280). However, when the service failure is the customer’s fault, the average intention to complain is significantly lower for subjects in the high interactional justice condition (F=18.71, p<0.001).

To aid in interpreting the results in Figure 1, we have also plotted the average satisfaction and intention to complain for the subjects in the control (no failure) group. In Table 3 below, we report the statistical tests for difference in means between the control condition and the four experimental conditions shown in Figure 1. The results of the comparison of means show that all four conditions are statistically significantly different from the control condition. The directionality of these differences suggests that if the failure is the fault of the service provider, in spite of high perceived interactional justice, the average satisfaction of consumers is still lower than the average satisfaction of those who do not experience a failure. However, if the failure is the customer’s fault and if customers perceive high interactional justice, their average satisfaction will be higher than the average satisfaction of those who do not experience a failure. The conclusions for intention to complain are similar to the conclusions for average satisfaction.

In summary, we find that after a service recovery attempt is made following a service failure, consumers’ perceived interactional and distributive justice have significant positive (negative) effects on satisfaction (intention to complain). Also as expected, we find that consumers are less satisfied (more likely to complain) if the service failure is the provider’s fault rather than their own. Interestingly, when failure occurs due to the customers’ fault, the average satisfaction (intention to complain) is the same as when there is no failure. Another interesting finding relates to the interaction effect of the locus of causality by perceived interactional justice. The pattern of this interaction suggests that providers can enhance (reduce) customer satisfaction (intention to complain) by simply treating them nicely during the service recovery attempt.

DISCUSSION

Not only does the disconfirmation paradigm predict, but it also seems intuitively obvious, that we would be less satisfied after we experience a service failure than if there had been no failure at all. Hoever, the aggregate results in Table 2 indicate that when the customer blames him/herself for the service failure, there is no difference between the levels of satisfaction or likelihood of complaint behavior between the failure and the no-failure condition. The reason for this apparent discrepancy could be explained by the fact that these means in the failure conditions are also influenced by the service recovery efforts.

FIGURE 1

LOCUS OF CAUSALITY AND INTERACTIONAL JUSTICE INTERACTION PLOTS

TABLE 3

DIFFERENCES IN MEANS BETWEEN LOCUS OF CAUSALITY X INTERACTIONAL JUSTICE AND CONTROL CONDITION

It is no surprise that our results suggest that consumers will be less satisfied and more likely to complain when the restaurant is at fault rather than the consumer. It is possible that this impact could be reduced by the front-line service employee, however. In this research, we only looked at one dimension of attribution theoryClocus of causality. The two other dimensions of attribution theory, controllability and stability, could provide means for managing the negative effects felt by the customer as a result of a service failure. Customers may be more understanding and/or forgiving if the service employee, after offering a sincere apology to the customer, would also add a comment such as "this sort of thing rarely happens here" (stability) or "I had no idea the bowl was defective" (controllability). These conjectures could be investigated in future research.

Results shown in Tables 1 and 2 confirmed our prediction that there is a difference in the level of satisfaction and the likelihood of complaint behavior depending on whether consumers perceive the outcome to be fair (i.e., high versus low level of distributive justice). The high level of distributive justice in our study was operationalized as just putting the customer back to his/her original state. Instead of being worse off than before the accident (low distributive justice), the service recovery effort replaced the customer’s loss (in this case a bowl of extra-hot salsa). Although most people appear to be content to receive what seems to be fair, i.e., just return them to their original state, other levels of distributive justice (e.g., offer them something extra) could be investigated in future research.

There are also relatively large differences in satisfaction in the high versus low interactional justice conditions. As mentioned earlier, this suggests how important prompt, courteous service may be to customers. An important implication of this finding is that front-line employees should be empowered by the service organization to handle most routine service failures. Both the customer and employee are frustrated when the simplest transactions have to be monitored or approved by supervisors. Not only does the customer have to wait, but the employee is made to feel inferior. An employee who is otherwise qualified for the job of serving customers should be trained and trusted to use discretion in dealing with routine service failure situations. Empowerment may give employees the best avenue to succeed in transforming a potentially unhappy customer into a delighted consumer.

Abrams and Pease (1993) conjectured that it is possible that after a service failure has been successfully resolved, consumers may feel a stronger commitment to the service provider than if no failure had occurred in the first place. We find evidence of this effect under a very specific conditionCwhen service failure occurs due to customers’ fault and when customers perceive high interactional justice in the service recovery attempt by the provider. Heskett, Sasser, and Hart (1990) contend that it is necessary for a service firm to attempt a service recovery to create a satisfied customer, irrespective of the cause of the problem or the location of blame. A major implication of this finding is that prompt, courteous service recovery actions on the part of the service provider when the customer has caused a service failure may increase levels of customer satisfaction and decrease the likelihood of complaint behavior above and beyond the levels that would ave been reached without any failure.

CONCLUSIONS AND FUTURE RESEARCH

The way in which the service recovery is handled is likely to have an important influence on whether the customer returns to the service firm. There is no reason why service encounters that begin with a service failure can not end as a service recovery success. The inability to respond to a service failure "adds insult to injury"Ccustomers are disappointed twice, first when the service breakdown occurs and later when there is a lack of effort on the part of the service provider to recover from the service failure.

Our research in "the art of service recovery" suggests that the "means" (process or interactional justice) may be more important than the "ends" (outcome or distributive justice) when it comes to service recovery. However, this conclusion is dependent on the particular manipulations of distributive and interactional justice used in this study. Further research with different types of service encounters and different manipulations of means and outcomes related to service recovery are needed to generalize these results. Conditions of an experiment describing a recovery from a core service failure rather than a delivery system failure could be manipulated. Core service failures have been reported most often as being dissatisfactory (Bitner et al. 1990) and as the reason consumers switch service firms (Keaveney 1995).

Future research is needed for other types of service encounters that occur in long-term rather than discrete exchange relationships, such as those experienced for professional services. It would also be interesting to look at a service failure from the employee’s point of view. Bitner et al. (1994) suggest that service employees and consumers see service failures in much the same way. It is quite possible that empowering employees to respond appropriately to service failures will result in a win-win-win situation: the customer, the employee, and ultimately the service firm itself will benefit.

REFERENCES

Abrams, M. and M. Pease (1993), "Wining and Dining the Whiners," Sales and Marketing Management (February), 73-75.

Albrecht, K. (1988), At America’s Service, Homewood, IL: Dow Jones- Irwin.

Albrecht, K. and R. Zemke (1985), Service America!, Homewood, IL: Dow Jones-Irwin.

Andreasen, A.R. (1977), "A Taxonomy of Consumer Satisfaction/Dissatisfaction Measures," The Journal of Consumer Affairs, 11(Winter), 11-24.

Andreasen, A.R. and A. Best (1977), "Consumers ComplainCDoes Business espond?" Harvard Business Review (July-August), 93-101.

Bell, C.R. and K. Ridge (1992), "Service Recovery for Trainers," Training and Development (May), 58-63.

Bell, C.R. and R.E. Zemke (1987), "Service Breakdown: The Road to Recovery," Management Review (October), 32-35.

Berry, L. (1986), "Big Ideas in Services Marketing," Journal of Consumer Marketing, 3 (Spring), 47-51.

Berry, L. and A. Parasuraman (1992), "Prescriptions for a Service Quality Revolution in America," Organizational Dynamics, 20 (Spring), 5-15.

Bies, R.J. and J.S. Moag (1986), "Interactional Communication Criteria of Fairness," Research in Org. Behavior, Greenwich, CT: JAI Press.

Bitner, M.J. (1990), "Evaluating Service Encounters: The Effects of Physical Surroundings and Employee Responses," Journal of Marketing, 54 (April), 69-82.

Bitner, M.J., B.H. Booms, and L.A. Mohr (1994), "Critical Service Encounters: The Employee’s Viewpoint," Journal of Marketing, 58 (October), 95-106.

Bitner, M.J., B.H. Booms, and M.S. Tetreault (1990), "The Service Encounter: Diagnosing Favorable and Unfavorable Incidents," Journal of Marketing, 54 (January), 71-84.

Blodgett, J.G., D.H. Granbois, and R.G. Walters (1993), "The Effects of Perceived Justice on Complainants’ Negative Word-of-Mouth Behavior and Repatronage Intentions," Journal of Retailing, 69 (Winter), 399-428.

Brown, S.P. and R.F. Beltramini (1984), "Consumer Complaining and Word of Mouth Activities: Field Evidence," Adv. in Consumer Research, 16,9-16.

Czepiel, J., M.R. Solomon, and C.F. Surprenant (1985), The Service Encounter, New York: Lexington Books.

Folkes, V.S. (1984), "Consumer Reactions to Product Failure: An Attributional Approach," Journal of Consumer Research, 10 (March), 398-409.

Goodwin, C. and I. Ross (1992), "Consumer Responses to Service Failures: Influence of Procedural and Interactional Fairness Perceptions," Journal of Business Research, 25, 149-163.

Greenberg, J. (1996), The Quest for Justice on the Job: Essays and Experiments, Thousand Oaks, CA: Sage Publications.

Hart, C.W.L., J.L. Heskett, and W.E. Sasser,Jr.(1990),"The Profitable Art of Service Recovery," Harvard Business Review (July-August), 148-156.

Heskett, J.L., W.E. Sasser, Jr., and C.W.L. Hart (1990), Service Breakthroughs: Changing the Rules of the Game, New York: Free Press.

Homans, G.C. (1961), Social Behavior: Its Elementary Forms, New York: Harcourt, Brace and World.

Keaveney, S.M. (1995), "Customer Switching Behavior in Service Industries: An Exploratory Study," Journal of Marketing, 59(April), 71-82.

Oliver, R.L. (1980), "Theoretical Bases of Consumer Satisfaction Research: Review, Critique, and Future Direction," Theoretical Developments in Marketing, Chicago: American Marketing Assoc., 206-210.

Oliver, R.L. and J.E. Swan (1989), "Consumer Perceptions of Interpersonal Equity and Satisfaction in Transactions: A Field Survey Approach," Journal of Marketing, 53(April), 21-35.

Parasuraman, A., V.A. Zeithaml, and L.L. Berry (1988), "SERVQUAL: A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality," Journal of Retailing, 64(Spring), 12-40.

Singh, J. (1991), "Understanding the Structure of Consumers’ Satisfaction Evaluations of Service Delivery," Journal of the Academy of Marketing Science, 19, 223-244.

Solomon, M.R., C. Surprenant, J.A. Czepiel, and E.G. Gutman (1985), "A Role Theory Perspective on Dyadic Interactions: The Service Encounter," Journal of Marketing, 49(Winter), 99-111.

TARP (986), Technical Assistance Research Program, Consumer Complaint Handling in America: An Update Study.

Zemke, R. (1993), "The Art of Service Recovery: Fixing Broken Customers And Keeping Them on Your Side," The Service Quality Handbook, New York: American Management Association, 463-476.

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