Master Thesis
University of Amsterdam
The relationship between Expectations, Performance Perceptions and Customer Satisfaction.
Nathalie Aleven (10845291) University of Amsterdam
Executive Programme in Management Studies Specialization: Strategy
Academic year: 2016-2017 Supervisor: Mr. Ph.D. B. Flier Amsterdam, 28 June 2017
Statement of Originality
This document is written by Student Nathalie Aleven who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
In this study, we examined the relationship between the differences in Expectations and Perceived
Performance and Customer Satisfaction and investigated the mediating roles of Attribution and Subjective Disconfirmation.
Questionnaires were completed by customers (N=51) and the Board of Directors (N=2) of a company in the Technical Services Sector.
We found evidence for possible mediating roles of (i) Attribution and (ii) Subjective Disconfirmation in the relationship between the differences in Expectations and Perceived Performance and Customer
Satisfaction. Further, as expected the difference between Expectations and Perceived Performance was positively related to Customer Satisfaction. Moreover, the difference between Expectations and Perceived Performance was positively related to Subjective Disconfirmation. We found no significant relationship between (i) Subjective Disconfirmation, (ii) Attribution and Customer Satisfaction. Finally, we found no significant relationship between differences in Expectations and Perceived Performance and Customer Satisfaction.
The study adds to the current literature by furthering our understanding of how the difference between Expectations and Perceived Performance relates to Attribution, Subjective Disconfirmation and Customer Satisfaction. This study has especially contributed to the work regarding the Attributions consumers assign for satisfactory and unsatisfactory product or service performance (Oliver, 2010). Because Attribution mediates Customer Satisfaction and Expectations and Perceived Performance, it needs to be included for further Customer Satisfaction and Service Quality Research. Furthermore, we found evidence that the indirect effect of Subjective Disconfirmation in the relationship between Expectations and Perceived Performance and Customer Satisfaction is significant. This study has shown that Subjective
Disconfirmation can be included as an intervening variable when investigating Customer Satisfaction. However, more research needs to be done on the differences between the direct- and indirect measurements of Subjective Disconfirmation, as this study has not found evidence that the relationship between
Subjective Disconfirmation and Customer Satisfaction is significant.
In addition, this study found a significant relationship between Customer Satisfaction and Customer Retention. Interesting findings include the differences between the Board of Directors and Customers in perceived Satisfaction and perceived Customer Retention. The perceptions of Satisfaction are higher for the Board of Directors than for the customers, and the perceptions of Retention are slightly higher for the Board of Directors than for the customers. When management is able to identify the gaps and knows which levels of Service Quality are deviating, they are able to improve their Service Quality. This also improves the customer understanding and ability to influence the customer.
Key words: Expectations, Perceived Performance, Disconfirmation, Attribution, Customer Satisfaction, Service Quality, Technical Services Sector.
Table of contents
Abstract ... 3
Table of contents ... 4
1. Introduction ... 5
2. Literature review ... 8
2.1 Expectations and Performance Perceptions ... 8
2.2 Expectations, Performance Perceptions and Subjective Disconfirmation ... 9
2.3 Expectations, Performance Perceptions and Attribution ... 10
2.4 Expectations, Performance Perceptions and Customer Satisfaction ... 12
2.5 Subjective Disconfirmation and Customer Satisfaction ... 13
2.6 Expectations, Performance Perceptions, Subjective Disconfirmation and Customer Satisfaction... 14
2.7 Attribution and Customer Satisfaction ... 15
2.8 Expectations, Performance Perceptions, Attribution and Customer Satisfaction ... 15
2.9 Research Model ... 16 3. Methods ... 16 3.1 Research procedure ... 16 3.2 Research context ... 17 3.3 Response rate ... 18 3.4 Measurement of variables ... 18 3.5 Reliability ... 21
3.6 Computing scale means ... 22
4. Results ... 23
4.1 Univariate ... 23
4.2 Bivariate ... 24
4.3 Regression ... 26
4.4 Conclusion ... 28
5. Discussion and conclusion ... 29
5.1 Practical implications ... 31
5.2 Theoretical implications ... 32
5.3 Strengths and limitations ... 32
5.4 Directions for future research ... 33
6. References ... 35
7. Appendices ... 38
1. Introduction
Increasing rapid deregulation and competition have led many retail and service businesses to seek profitable ways to differentiate themselves. Two strategies to differentiate in a profitable way are to deliver high Service Quality and aim for Customer Satisfaction. According to Shemwell et al. (1998), the key to sustainable competitive advantage lies in delivering high quality service that will in turn result in satisfied customers. Research has shown that if firms are able to offer superior Service Quality, they can become more profitable and able to sustain a competitive advantage in their served markets. Profit Impact of Market Strategy (PIMS) research has indicated that companies that offer superior service are able to charge more for their product (Gale, 1992) while achieving higher-than-normal market share growth (Buzzell & Gale, 1987) and profitability (Phillips et al., 1983). Thus, offering superior Service Quality can help firms sustain a competitive advantage in their served markets and can help them become more profitable (Hampton, 1993).
Customer Satisfaction has been a subject of great interest to organizations and researchers. Kotler (2012: 13) defines Customer Satisfaction as: “The extent to which a product’s Perceived Performance matches a buyer’s Expectations.” The interest in Customer Satisfaction has grown enormously, especially since research has shown that there is a direct link between Customer Satisfaction and economic returns. According to Anderson et al. (1994), firms that achieve high levels of Customer Satisfaction also enjoy superior economic returns. Oliver (2010) states that satisfaction has direct effects on profit through its influence on Retention as well as on other facilitating concepts including word of mouth, marketing cost reductions and human capital attraction. Satisfaction results in excess profit returns to the firm that excels among and against its competition. According to Kotler (1991), high Customer Satisfaction ratings are widely believed to be the best indicator of a company’s future profits.
In order to determine the current Service Quality of a company’s customers, Expectations and Performance Perceptions can be explored. In Service Quality research Expectations are viewed as desires or wants of consumers. Parasuraman, Berry, and Zeithaml (1990: 12) noted that the service Expectations concept is “intended to measure consumers’ normative Expectations, and that these Expectations represent an “ideal standard” of performance.” Oliver (2010: 10) defines Performance Perceptions as “the perceived amount of product or service attribute outcome or overall outcome delivered and/or received.” Literature shows that there is a relationship between Expectations and Performance Perceptions. This relation could be positive when customers perceive a performance outcome that exceeds their Expectations. The relationship between Expectations and Performance Perceptions could be negative when customers find that a company is underperforming, where in this
case Expectations exceed the Performance Perceptions. In this study Expectations (E) of a service are compared with Performance Perceptions (P). The equation is therefore: Service Quality = P - E.
The Customer Satisfaction and Service Quality literature has done little research into the Attributions consumers assign for satisfactory and unsatisfactory product or service performance (Oliver, 2010). This study has extended that work by exploring the effects of Attribution. Management would benefit from knowing what Attributions its consumers assign for satisfactory and unsatisfactory product or service performance. Attributions are the perceived causes of events and influence perceptions of satisfaction. When customers have been surprised by an outcome (the service is either much better or much worse than expected), they tend to look for the reasons, and their assessments of the reasons can influence their satisfaction. Attribution measurement is rarely performed in satisfaction surveys. However, Attribution is a concept which needs to be included when measuring service performance and Customer Satisfaction, as successes are in most instances foregone conclusions in the minds of the consumers and failures may not be processed. Research has shown that Attribution is related to the concepts of Customer Satisfaction and Exceeding Expectations. Therefore, this research studies the relationship between these two concepts and Attribution. We argue that Attribution positively mediates the relationship between the difference in Expectations, Performance Perceptions and Customer Satisfaction as defined in the hypothesis.
This research studies the relationship between Expectations, Performance Perceptions and Customer Satisfaction. According to McKinney et al. (2002), Disconfirmation is the consumer’s subjective judgments resulting from comparing their Expectations with their perceptions of performance received. Oliver (2010) makes a distinction between Objective Disconfirmation and Subjective Disconfirmation. According to Oliver, Objective Disconfirmation provides the basis for a subjective interpretation of this Expectations-performance difference. With Subjective Disconfirmation, customers give their perceptions of whether the service is worse or better than expected. In the literature there are two schools of thought concerning how to measure Disconfirmation. Oliver (2010) states that Subjective Disconfirmation has to be included as a separate variable when investigating Customer Satisfaction, allowing for the direct measurement of Disconfirmation. Pitt et al. (1997) and Swan and Trawick (1980) argue for the subtractive Disconfirmation approach, where Expectations and Perceived Performance are subtracted from each other. In this case, Disconfirmation is not directly measured. We opt for the direct approach as it has been the more established approach in the Expectations-Disconfirmation paradigm. Therefore, this research suggests that Subjective Disconfirmation positively mediates the relationship between the difference in Expectations, Performance Perceptions and Customer Satisfaction.
The purpose of this research is to study the relationship between Expectations, Performance
Perceptions and Customer Satisfaction as well as investigate the mediating roles of (i) Attribution and (ii) Subjective Disconfirmation. This research aims to provide new insights in the Customer
Satisfaction and Service Quality literature by looking at the role that Customer Expectations and Performance Perceptions play as possible antecedents of customers’ Attribution, Subjective
Disconfirmation and Satisfaction. We expect that Attribution and Subjective Disconfirmation act as mediators between Expectations, Performance Perceptions and Customer Satisfaction. We believe that our study contributes to a more comprehensive understanding of Customer Satisfaction and Service Quality. We have formulized the following research questions: a) What is the influence of the difference in Expectations and Performance Perceptions on Customer Satisfaction?
b) What is the mediating effect of (i) Subjective Disconfirmation and (ii) Attribution on the link between Expectations and Performance Perceptions on Customer Satisfaction?
2. Literature review
This chapter discusses the most relevant findings from the current literature about Expectations, Performance Perceptions and Customer Satisfaction and states the hypotheses of this study. This chapter begins with exploring the differences between Expectations and Performance Perceptions. Secondly, the concepts of Expectations and Performance Perceptions and Subjective Disconfirmation will be outlined and the relationship of these variables will be formulated in the first hypothesis. Thirdly, the link between Expectations, Performance Perceptions and Attribution is described.
Fourthly, the link between the differences in Expectations and Performance Perceptions and Customer Satisfaction is discussed. Fifthly, the relationship between Subjective Disconfirmation and Customer Satisfaction is exposed. Sixthly, this chapter outlines how Subjective Disconfirmation mediate the relationship between the difference in Expectations and Performance Perceptions and Customer Satisfaction. Seventhly, the relationship between Attribution and Customer Satisfaction is described. Finally, this chapter outlines how Attribution mediate the relationship between the difference in Expectations and Performance Perceptions and Customer Satisfaction. The chapter ends with a research model which graphically illustrates the stated hypotheses.
2.1 Expectations and Performance Perceptions
Increasing rapid deregulation and competition have led many retail and service businesses to seek profitable ways to differentiate themselves. One strategy to differentiate is delivering high Service Quality. Research has shown that if firms are able to offer superior Service Quality, they can become more profitable and this helps them sustain a competitive advantage in their served markets. Profit Impact of Market Strategy (PIMS) research has indicated that companies that offer superior service are able to charge eight percent more for their product (Gale, 1992) while achieving profitability (Phillips et al., 1983) and higher-than-normal market share growth (Buzzell & Gale, 1987). Thus, offering superior Service Quality can help firms sustain a competitive advantage in their served markets and help them become more profitable (Hampton, 1993). Expectations and Performance Perceptions are explored in this study to determine the current Service Quality of an engineering company.
Oliver (2010: 10) defines Expectations as “a prediction, sometimes stated as a probability or likelihood, of attribute or product performance at a specific performance level.” Parasuraman, Zeithaml and Berry (1988: 17) define Expectations as “desires or wants of customers, i.e., what they feel a service provider should offer rather than would offer.” In the Service Quality research
Expectations” and that these Expectations represent an “ideal standard of performance”. Oliver (2010: 10) defines Performance Perceptions as “the perceived amount of product or service attribute outcome or overall outcome delivered and/or received.”
Oliver (2010) states that there is a relationship between Expectations and Performance Perceptions. This relation could be positive when customers perceive a performance outcome that exceeds their Expectations, and the relationship could be negative when customers find that a company is underperforming. In this case, Expectations exceed Performance Perceptions. Businesses with high Service Quality will meet or exceed Expectations.
In this study Expectations (E) of a service are compared with Performance Perceptions (P). The equation is therefore: Service Quality= P-E.
2.2 Expectations, Performance Perceptions and Subjective Disconfirmation
In the Customer Satisfaction and Service Quality literature Expectations are compared with Performance Perceptions to determine whether a company is underperforming or exceeding their Expectations. According to Oliver (2010), positive Disconfirmation occurs where performance exceeds Expectations and negative Disconfirmation occurs when performance falls short of Expectations.
Oliver (2010: 22) defines Disconfirmation from a technical point of view as “the result of a comparison between what was expected and what was observed.” In current satisfaction theory, Disconfirmation refers more to the psychological interpretation of the expectancy-performance discrepancy. The expectancy-performance discrepancy is “the objective or calculated difference between Expectations levels and performance levels, sometimes referred to as a gap.” This concept would be described by customers in terms of the performance of a service being better or worse than expected. Anderson and Sullivan (1993) define Disconfirmation as the extent to which perceived quality fails to match pre-purchase Expectations. Mohr (1982) states that the
Expectancy/Disconfirmation paradigm provides the grounding for the vast majority of satisfaction studies in process theory and encompasses four constructs: 1) Expectations; 2) Performance; 3) Disconfirmation; and 4) Satisfaction. Dis/confirmation arises from discrepancies between prior Expectations and actual performance. This conceptualization is reflected in the definition of satisfaction by Tse and Wilton (1988: 204): “The consumer's response to the evaluation of the perceived discrepancy between prior Expectations or some norm of performance and the actual performance of the product as perceived after its consumption.”
Oliver (2010) makes a distinction between objective Disconfirmation and subjective Disconfirmation. He states that if Expectations and performance are combined, this forms the objective Disconfirmation level. Objective Disconfirmation provides the basis for a subjective interpretation of this Expectations-performance difference. With subjective Disconfirmation, customers give their perceptions of whether the service is better or worse than expected.
In conclusion, according to Oliver (2010), when Performance Perceptions are more positive than what was expected, positive Disconfirmation occurs. When Performance Perceptions do not meet
Expectations, meaning that performance is less than expected, negative Disconfirmation occurs. Finally, if Performance Perceptions exactly match what was expected, confirmation is said to occur. Exceeding Expectations means Performance Perceptions are higher than Expectations.
Underperforming means that the Perceived Performance is less than expected. This research assumes that positive levels of Expectations have a positive impact on customer Disconfirmation and that negative levels of Expectation, have a negative impact on customer Disconfirmation. Therefore, we hypothesize:
H1. Positive levels of Expectations have a positive effect on Subjective Disconfirmation and vice versa.
2.3 Expectations, Performance Perceptions and Attribution
Attributions are the perceived causes of events and influence perceptions of satisfaction. When customers have been surprised by an outcome (the service is either much better or much worse than expected), they tend to look for the reasons, and their assessments of the reasons can influence their satisfaction. For example, was the company able to control the delay of the delivery of the service, or did the delay occur because of poor weather conditions (Wilson et al., 2016)?
Attribution theory focuses on explaining why a certain event has occurred and proposes that consumers look for the cause of particular experiences when arriving at satisfaction judgments. The Attribution process results in some predictable patterns of blame (failure) or gratitude (success) assessment. Generally, these patterns fall within the structure of the locus, stability and controllability framework. Weiner (1979) defines the dimensions as follows:
1. The locus of causality refers to whether the cause was something about the attributor (internal) or outside the attributor (external). This refers to judgments of who is responsible for an event.
Consumers can assign the locus to themselves or to an external entity such as a service provider. A self-ascribed event occurs when a consumer blames him or herself for a negative event. For example, a
2. Stability is defined as referring to whether the cause was constant over time (stable) or variable over time (unstable) (Weiner, 1979). This is the likelihood that an event will occur again in the future. For example, consumers could ask themselves, “If I buy this product or use this service again, is another negative outcome likely to happen?” If a customer experienced the same negative situation more often at the same company, he or she naturally comes to believe that this is a stable situation, and
satisfaction with the company will be diminished.
3. Controllability is defined by Weiner (1979) as referring to whether the cause was controlled by a company (in one’s control) or not controllable by a company (outside one’s control). A controllable cause was therefore defined as one that could be changed or affected by someone, either by the actor or by other people (Weiner, 1979). This refers to the extent to which an outcome was controllable. For example, was the company able to control the service outcome from the perspective of the customer? If a company delivered a service later than discussed because of weather events, the customer could be dissatisfied about the situation but does not blame the company because weather events are
uncontrollable.
These examples illustrate the point that Attribution requires a motivating stimulus. This research takes Attribution into account because it can have an effect on the reason for the outcome of customer experiences and affects the perceived causes of negative (and positive) perceptions of service experiences of the customer.
Attribution is determined as an inference into what caused observed events (a causal agent) either as a specific reason (e.g. poor quality construction) or as a dimension (e.g. the consumer versus an external entity such as the manufacturer) (Oliver, 2010). The answer to the question “why did this outcome occur?” represents an Attribution.
The concepts associated with Attribution are satisfaction and Disconfirmation. As generally accepted in the literature, Disconfirmation of Expectations is the most prominent causal agent for the onset of Attribution. This observation pertains whether the disconfirming evidence is thought to be discrepant or “familiar,” sharing some characteristics with regret. The Disconfirmation process described in this agency literature is identical to the way it is discussed in the satisfaction literature, which helps support a strong rationale for the study of Attribution in the context of consumer satisfaction and dissatisfaction (Oliver 2010). In this research the relationship between Attribution, Disconfirmation and Customer Satisfaction is studied.
Little research has been done about the Attributions consumers assign for satisfactory and unsatisfactory product or service performance. Attribution measurement is rarely performed in satisfaction surveys. Management would benefit from knowing which Attributions its consumers assign for satisfactory and unsatisfactory product or service performance. Because the Attributional
perceptions of the actor and observer are frequently at variance, the consumer as actor is likely to draw different Attributional conclusions from the manager as observer. Successful performance is
differentially attributed to the customer by the customer and to the firm by the firm’s management. Failure however tends to be attributed to the firm by the customer and to some other entity (e.g. suppliers, but also including the customer) by the firm (Oliver, 2010). Therefore, this research samples both the consumer’s and management’s perspectives of the service performance outcome. It is
possible that management thinks that some Service Quality aspects will be delivered and that the customer disagrees. When the management is able to identify the gaps and knows if and where Service Quality is lacking they are able to improve their Service Quality. This will improve the customer understanding and ability to influence the customer.
To conclude, little research has been done about the Attributions consumers assign for satisfactory and unsatisfactory product or service performance. According to Wilson et al. (2016), Attributions are the perceived causes of events and influence perceptions of satisfaction. When customers have been surprised by an outcome (the service is either much better or much worse than expected), they tend to look for the reasons, and their assessments of the reasons can influence their satisfaction. Therefore, Attribution is a concept which needs to be included when measuring service performance and Customer Satisfaction, as successes are in most instances foregone conclusions in the minds of the consumers and failures may not be processed. Theory has shown that Attribution is consistent with the concepts of Customer Satisfaction and Disconfirmation. This research assumes that Customer
Attribution has a positive impact on the difference between Customer Expectations and Performance Outcome. Therefore, we hypothesize:
H2. Positive levels of Expectations have a positive effect on Attribution and vice versa.
2.4 Expectations, Performance Perceptions and Customer Satisfaction
Kotler (2012: 13) defines Customer Satisfaction as: “The extent to which a product’s Perceived Performance matches a buyer’s Expectations.” Kotler argues that if performance matches Expectations the customer is satisfied, and if performance exceeds Expectations the customer is highly satisfied or delighted.
This study proposes that positive levels of Expectations have a negative effect on Customer Satisfaction and negative levels of Expectations have a positive effect on Customer Satisfaction. Therefore, we hypothesize:
2.5 Subjective Disconfirmation and Customer Satisfaction
Customer Satisfaction is defined in many ways. Oliver (2010: 10) defines Customer Satisfaction as: “Satisfaction is the consumer’s fulfillment response. It is a judgment that a product or service feature, or the product or service itself, provides a pleasurable level of consumption-related fulfillment, including levels of under- or over-fulfillment.” Kotler (2012: 13) defines Customer Satisfaction as: “The extent to which a product’s Perceived Performance matches a buyer’s Expectations.” Kotler argues that if performance matches Expectations the customer is satisfied, and if performance exceeds Expectations, the customer is highly satisfied or delighted. Oliver (2010) agrees with this concept of the Expectancy Disconfirmation theory, which is one of the comparison operations for satisfaction. This concept refers to the psychological interpretation of an Expectations-performance discrepancy. Customers would describe this mechanism in terms of the performance of a product or service being better or worse than expected. Anderson and Sullivan (1993) state that satisfaction is best specified as a function of perception and "Disconfirmation."
Disconfirmation becomes central in explaining consumer satisfaction. In the satisfaction research literature, Disconfirmation occupies a central position as a crucial intervening variable. According to Oliver (2010), positive Disconfirmation occurs when performance exceeds Expectations, and with negative Disconfirmation, performance falls short of Expectations. Anderson and Sullivan (1993) state that satisfaction is best specified as a function of perception and Disconfirmation. Anderson and Sullivan define Disconfirmation as the extent to which perceived quality fails to match pre-purchase Expectations.
McKinney et al. (2002: 299) define Disconfirmation as “Consumer subjective judgments resulting from comparing their Expectations and their perceptions of performance received.” Oliver (2010) makes a distinction between objective Disconfirmation and subjective Disconfirmation. Oliver states that if Expectations and performance are combined, this forms the objective Disconfirmation level if quantified. Whereas objective Disconfirmation provides the basis for a subjective interpretation of this Expectations-performance difference. With subjective Disconfirmation, customers give their
perceptions of whether the service is worse or better than expected. Figure 1 shows that subjective Disconfirmation is an antecedent to satisfaction.
In the literature, there are two schools of thought regarding how to measure Disconfirmation. Oliver (2010) states that Subjective Disconfirmation has to be included as a separate variable when
investigating Customer Satisfaction in order to directly measure Disconfirmation. Pitt et al. (1997) and Swan and Trawick (1980) argue for the subtractive Disconfirmation approach where Expectations and
Perceived Performance are subtracted from each other. In this case, Disconfirmation is not directly measured. We opt for the direct approach because it has been the more established approach in the Expectations-Disconfirmation paradigm.
Thus, research has shown that Disconfirmation becomes central in explaining Consumer Satisfaction and that satisfaction is best specified as a function of Subjective Disconfirmation. In the satisfaction research literature, Subjective Disconfirmation occupies a central position as a crucial intervening variable. Customers would describe this mechanism in terms of the performance of a product or service being better or worse than expected. This research assumes that positive Subjective Disconfirmation has a positive effect on Customer Satisfaction. Therefore, we hypothesize: H4:Higher levels of Subjective Disconfirmation have a positive effect on Customer Satisfaction.
Figure 1: The Complete Expectancy Disconfirmation With Performance Model, adapted from Oliver (2010).
2.6 Expectations, Performance Perceptions, Subjective Disconfirmation and Customer Satisfaction
According to Oliver (2010), when Performance Perceptions are more positive than what was expected, positive Disconfirmation occurs and leads to consumer satisfaction. On the other hand, negative Disconfirmation is when performance falls short of Expectations and leads to customer dissatisfaction. We argue that subjective Disconfirmation mediates the relationship between the difference in
customer Expectations, Performance Perceptions and Customer Satisfaction. Therefore, we hypothesize:
H5: Higher levels of Subjective Disconfirmation positively mediate the relationship between exceeding Expectations and Customer Satisfaction.
2.7 Attribution and Customer Satisfaction
Little research has been done about the Attributions consumers assign for satisfactory and
unsatisfactory product or service performance. According to Wilson et al. (2016), Attributions are the perceived causes of events and influence perceptions of satisfaction. When customers have been surprised by an outcome (the service is either much better or much worse than expected), they tend to look for the reasons, and their assessments of the reasons can influence their satisfaction. Attributions are the perceived causes of events and influence perceptions of satisfaction (Oliver, 2010). This research assumes that customer Attribution has a positive impact on Customer Satisfaction. Therefore, we hypothesize:
H6: Higher levels of Attribution have a positive effect on Customer Satisfaction.
2.8 Expectations, Performance Perceptions, Attribution and Customer Satisfaction
The concepts associated with Attribution are satisfaction and Disconfirmation. As generally accepted in the literature, Disconfirmation of Expectations is the most prominent causal agent for the onset of Attribution. This observation pertains whether the disconfirming evidence is thought to be discrepant or “familiar,” sharing some characteristics with regret. The Disconfirmation process described in this agency literature is identical to the way it is discussed in the satisfaction literature, which helps support a strong rationale for the study of Attribution in the context of Customer Satisfaction and dissatisfaction (Oliver 2010). We argue that Attribution positively mediates the relationship between the difference in Expectations, Performance Perceptions and Customer Satisfaction. Therefore, we hypothesize:
H7: Higher levels of Attribution positively mediate the relationship between exceeding Expectations and Customer Satisfaction.
2.9 Research Model
Figure 2 demonstrates the research model of the relationship between the difference in Expectations and Performance Perceptions and Customer Satisfaction. We test this in a study on customers of an engineering agency. We aim to provide new insights in the Customer Satisfaction and Service Quality literature by looking at the relationship between differences in Expectations and Performance
Perceptions, Subjective Disconfirmation, Attribution and Customer Satisfaction and by answering the following research questions: a) What is the influence of the difference in Expectations and
Performance Perceptions on Customer Satisfaction?
b) What is the mediating effect of (i) Subjective Disconfirmation and (ii) Attribution on the link between Expectations and Performance Perceptions on Customer Satisfaction?
Figure 2: Conceptual model
3. Methods
3.1 Research procedure
The type of research is a questionnaire given to customers and the Board of Directors of an
engineering agency. The questionnaire is distributed at a single point in time through Cross-sectional study. Data were collected by means of an online survey distributed among 178 customers and two members of the Board of Directors. The online survey was sent to every customer and the members of the Board of Directors of an engineering agency. The surveys consist of questions that measure the four different variables being researched in their relationship to each other, and the same set of questions was sent to the Board of Directors. However, the instructions for the Board of Directors were to give their perceptions on how they think customers give their answers. According to Oliver
differences between the perceptions of the Board of Directors of the engineering agency and the customers are analyzed.
The questions are answered by making use of a seven-point Likert scale. This method is used because the respondents (customers and the Board of Directors of an engineering agency) of this questionnaire are likely to be busy professionals, and making use of a Likert scale reduces the time a respondent must spend filling out the survey.
Survey Administration started on April 26th, 2017. The survey was closed on June 2nd, 2017. The Statistical Software Package for Social Sciences (SPSS) was used to perform the statistical analyses. We chose an online survey because we asked customers across the country. The data collection used to test these hypotheses is quantitative in nature. The data are collected through a questionnaire distributed among the targets of Customer Satisfaction. In addition, we used the tool Qualtrics for the data collection. This is an effective provider for collecting data and is especially a safe tool for analyzing data and integrating it into SPSS.
3.2 Research context
The research is conducted at an engineering agency named QING Groep. A company in the profit sector is chosen because this research studies the relationship between Expectations, Performance Perceptions and Customer Satisfaction. As mentioned in the literature study, research has shown that there is a direct link between firm profit and Customer Satisfaction and Service Quality. The
engineering agency is located in Arnhem, Borne, Utrecht and Wageningen. Different professionals in various disciplines work or at the customer’s location. QING Groep contains three Business Units: QING Engineering, QING Mechatronics and QING Sustainable.
QING Sustainable supports its clients with the sustainability of buildings and processes. That means looking at the whole picture of energy and energy efficiency. How can energy be saved or recovered? And what are the possibilities, both technically and financially, for sustainable generation?
QING Engineering uses its knowledge and abilities to achieve the objectives of its clients. The engineers of QING Groep work both at the client’s location or at our own engineering department in Arnhem or Borne.
QING Mechatronics supports its clients in optimizing its production by offering customer-specific total or partial solutions. QING Mechatronics develops, builds, installs and maintains machinery, equipment, and performs tests and measurements. The combination of designing and developing
concepts with assembly and commissioning creates great problem solving skills. That means QING Mechatronics is often utilized in situations where standard concepts are not enough.
3.3 Response rate
Of the 178 customers that received the online survey, 54 fully completed the questionnaire (response rate 30.34%). Three of the respondents turned out to be former colleagues of QING Groep. Therefore, we chose to remove these respondents from the data analysis. 51 respondents remained (response rate 28.65%). We also sent the questionnaire to the two Board of Directors of the engineering agency. They both completed the questionnaire (response rate 100%).
3.4 Measurement of variables
The items used in the questionnaire were derived from English studies. Because all respondents to the survey spoke Dutch as their first language, the original items were translated into Dutch.
Operationalization of variables
Expectations
Expectations of Service Quality are measured trough the SERVQUAL (Service Quality) instrument. The items are adapted from Parasuraman, Zeithaml, and Berry (1988) who define perceived Service Quality as "a global judgment, or attitude, relating to the superiority of the service”. The five dimensions as drivers of Service Quality are: Reliability, Responsiveness, Assurance, Empathy and Tangibles. SERVQUAL is used to measure customers’ perceptions of Service Quality based on these five dimensions. SERVQUAL is a 22-item instrument for assessing customer perceptions of Service Quality in service and retail organizations. Customers are instructed to answer questions about what they expect from an excellent engineering agency, and the questions are answered by making use of a seven-point Likert scale. A sample item includes: “Excellent engineering agencies will provide its services at the time they promise to do so.” The respondents filled in a seven-point scale ranging from (1) “Strongly Disagree” to (7) “Strongly Agree.”
Performance Perceptions
Performance Perceptions of Service Quality are measured in the same way as Expectations, except the questions are specified for the engineering agency QING Groep. A sample item includes: “QING
scale ranging from (1) “Strongly Disagree” to (7) “Strongly Agree.”
Subjective Disconfirmation
According to Oliver (2010), positive Disconfirmation occurs where performance exceeds Expectations, and negative Disconfirmation occurs when Performance Perceptions falls short of Expectations. A score of 0 would indicate that performance met desired/expected Service Quality. For Subjective Disconfirmation there are two main approaches described in the literature concerning this measurement. Oliver (2010) states that Subjective Disconfirmation has to be included as a separate variable when investigating Customer Satisfaction in order to directly measure Disconfirmation. Pitt et al. (1997) and Swan and Trawick (1980) argue for the subtractive Disconfirmation approach where Expectations and Perceived Performance are subtracted from each other. In this case, Disconfirmation is not directly measured. In this study we opted for the direct measurement of Disconfirmation. Customers describe the performance of a service as being better or worse than expected (Oliver, 2010).
Attribution
To measure causal Attributions the Causal Dimension Scale of Russells (1982) is used. This scale is designed to assess how the attributor perceives the causes she or he has stated for an event. Moreover, this scale assesses causal perceptions in terms of the locus of controllability, causality and stability dimensions described by Weiner (1979). Russells (1982) described a set of items to separately assess each of the three causal dimensions. The set of Russells’ items are adapted for the questionnaire. The questions are answered by making use of a seven-point Likert scale. Low scores represent greater internal locus, greater stability and greater controllability. High scores on these subscales indicate that the cause is perceived as internal, stable and controllable.
Answers to locus dimension questions were re-coded into dummy variables (presented in the same Table 3) in order to use them in a regression model. A sample item includes: “The service outcome was controlled by QING Groep/ was uncontrolled by QING Groep.” The respondents filled in a seven-point scale ranging from (1) “The service outcome was controlled by QING Groep” to (7) “The service outcome was uncontrolled by QING Groep.”
Customer Satisfaction
According to Shemwell et al. (1998) the key to a sustainable competitive advantage lies in delivering high quality service that will in turn result in satisfied customers. In our survey we used four items adapted from Taylor and Baker (1994) to measure Customer Satisfaction A sample item includes: “Overall, in purchasing services, I believe that I would be pleased with QING Groep’s services.” The respondents filled in a seven-point scale ranging from (1) “Very Unimportant” to (7) “Very
Dissatisfied” to (7) “Very Satisfied.” A sample item includes: “My feelings toward QING Groep’s services can best be characterized as.”
Control variables
To test the relative relationship of the dependent and independent variables the following control variables take into account:
Business Unit
The first control variable is the Business Unit of QING Groep. There could be differences between the Business Units (Engineering, Mechatronics or Sustainable) of QING Groep. For example, customers of Engineering could be more satisfied than customers of the Business Unit Mechatronics. Therefore, the Business Units of QING Groep are suggested as a control variable.
Age
The second control variable is age. Age has attracted considerable research attention. Research has concentrated on the differences in the information-processing abilities which are needed to evaluate a product, by compering young and elderly customers (Moscovitch, 1982; Walsh, 1982). Most of these studies have shown that information processing declines with age (Gilly & Zeithalm, 1985). Older people seem to have restricted information-processing capabilities. As a result, their reactions to satisfaction shifts might also change. For that reason, age is suggested as a control variable.
Gender
The third control variable is Gender. Kolodinsky (1993) states that differences in the complaining behavior of women and men may indicate a difference in how each group evaluates failures. Therefore, gender is suggested as a control variable.
Frequency of used services
The fourth control variable is Frequency of used services (per given project), of an engineering agency. This is the number of previous encounters with all providers within a particular service industry. Zeithaml et al. (1993) states that customers possess higher Expectations if they have previous experience. When customers use services for an industry on a frequent basis, they should also have a greater depth of knowledge for recoveries and judging failures and thus may have stronger
Expectations in both cases. In our survey we used one item to measure frequency of used services. The sample item includes: “How often did you use the services (per given project) of an engineering bureau?” The respondents filled in one of the following groups: ≤ 5, 6-10, 11-15, 16-20, ≥ 21.
Level of importance QING Groep as partner.
In our survey, we used one item to measure the level of importance QING Groep as partner. It could be the case that a customer from for example the Business Unit QING Sustainable is more satisfied with the Service Quality than a customer from the Business Unit QING Mechatronics. The sample item includes: “How important is QING Groep as a partner?”. The respondents filled in a seven-point scale ranging from (1) “Very Unimportant” to (7) “Very Important.”
Education level
The sixth control variable is Education level. The level of education may have a significant effect on Attribution, Disconfirmation, Exceeding Expectations and Customer Satisfaction.
Years of experience with QING Groep
In our survey we used one item to measure years of experience with QING Groep. The sample item includes: “How many years of experience do you have with QING Groep?” The respondents were asked to fill in one of the following groups: ≤ 1, 2, 3, ≥ 4.
Customer Retention
Customer Retention is conducted as an extra variable and further evaluated in the discussion section. Customer Retention examines repeated purchase behavior (Hennig-Thurau, 2004). According to Kotler (1994), the key to Customer Retention is Customer Satisfaction. Oliver (2010) states that satisfaction has direct effects on profit through its influence on Retention. Therefore, we want to investigate Customer Retention as an extra variable. We suggest: The higher the level of satisfaction, the higher the level of Customer Retention. In our survey, we used four items to measure Customer Retention. The items are adapted from Hennig-Thurau (2004). A sample item includes: “In the future, I will buy most services at QING Groep.” The respondents filled in a seven-point scale ranging from (1) “Strongly Disagree” to (7) “Strongly Agree.”
3.5 Reliability
Reliability enables the examination of the consistency of measurements. Reliability checks were performed for Expectations, Performance Perceptions, Customer Satisfaction, Attribution and
Customer Retention. The Cronbach’s alpha represents the estimator of the internal consistency and has been tested to verify if all the items in one scale measure the same, or if some questions should not be used for analysis. As exhibited in Table 1, all five variables have a Cronbach’s alpha of >.7, which indicates a high level of internal consistency.
Table 1 Cronbach’s Alpha
__________________________________________________________________________________________
Variable Cronbach’s Alpha*
__________________________________________________________________________________________ Expectations .884 Performance Perceptions .947 Customer Satisfaction .919 Attribution .871 Customer Retention .893 __________________________________________________________________________________________ *Note that Subjective Disconfirmation is not included in the measurement, this variable does not include items.
3.6 Computing scale means
The final preliminary step is to create new variables as a function of existing variables for hypothesis testing. In order to describe a variable we calculated the mean of all items. Means and standard deviations of all variables are exhibited in Table 2. The new variables are: ∆ Expectations and Performance Perceptions and Subjective Disconfirmation.
4. Results
4.1 Univariate
In the univariate data inspection, the variables are examined separately. First of all, A frequency table was used to examine the general characteristics of the variables. Then, the mean, standard deviation, minimum and maximum, standardized skew and kurtosis are analyzed. Further, different graphical overviews are analyzed. Assumptions are also checked in the univariate data inspection. In particular, the assessment of normality. This is done by looking at the histograms and Q-Q of the numeric variables. Furthermore, all variables under investigation have been checked for missing data. A frequency test was run for all variables. No data was missing nor did errors occur in the data. Finally, in the univariate analysis, possible outliers were looked into. There were no possible outliers. Looking at the means and standard deviations, the respondents seem overall satisfied. The mean of Customer Satisfaction is M=5.652 and the standard deviation is SD=.944. Furthermore, the mean of Attribution is relatively high at M=3.623, and the standard deviation is SD=.836. This indicates that there could be other reasons that influence the satisfaction of customers.
Table 2 Mean, Standard deviation and Correlations and Reliabilities Mean Standard Deviation 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 1.Business Unit 2.000 0.748 2.Age 3.550 1.045 .179 3.Gender 1.120 .325 .329* .042 4.Education Level 5.120 .516 .207 .026 -.084 5.Years of Experience 2.000 1.483 .126 .168 -.166 .052 6.Frequency of used services 2.510 1.155 -.046 .062 -.110 .065 .058 7.Level of importance QING Groep as partner 4.730 1.297 .268 .305* .220 .019 -.177 .202 8.∆ Expectations and Performance Perceptions .352 .759 -.062 -.013 .011 -.090 .289* -.109 -.336* 9.Subjective Disconfirmation 1.343 .746 -.045 -.006 .017 -.063 .304* -.128 -.356* .996** 10.Attribution 3.623 .836 -.007 0 -.161 .301* -.233 -.170 -.265 .006* .040 (.871) 11.Customer Satisfaction 5.652 .944 .220 .187 .136 .168 -.125 .249 .537** -.711** -.708** -.196 (.919) 12.Customer Retention 4.544 1.084 -.031 .278* .098 .008 -.075 .241 .446** -.417** -.419** -.277* .649** (.893)
Note: N=51 customers. Reliabilities are reported along the diagonal. **. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
4.2 Bivariate
In the bivariate data analysis, the cohesion of two variables has been analyzed as well as the distribution of the variables. This was done by scatter plots for the numerical variables and by box plots when analyzing a numerically dependent variable and categorically independent variables. The assumptions were also analyzed in the bivariate data inspection, which involved checking whether there is a linear relationship between the variables and whether there is homoscedasticity, where variance in residuals should be the same at each level. For this reason, scatterplots were used.
SPSS provided us with a table of correlation coefficients called a correlation matrix for all
the combinations of variables. There is a high positive relation between Subjective Disconfirmation and the difference between Expectations and Perceived Performance. The relationship has a
significance value less than .01 and a Pearson correlation coefficient of r=.996. Further, there is a high negative relation between Customer Satisfaction and the difference between Expectations and
Perceived Performance. The relationship has a significance value less than .01 and a Pearson
correlation coefficient of r=-.711. There is also a high negative relation between Customer Satisfaction and Subjective Disconfirmation. The relationship has a significance value less than .01 and a Pearson correlation coefficient of r=-.708. Finally, there is a positive relation between the difference between Expectations and Perceived Performance and Attribution. The relationship has a significance value less than .05 and a Pearson correlation coefficient of r=.006.
When we look at the control variables, there is a correlation between Business Unit and Gender. The relationship has a significance value less than .05 and a Pearson correlation coefficient of r=.329. Further there is a correlation between Level of importance QING Groep as partner and Age. The relationship has a significance value less than .05 and a Pearson correlation coefficient of r=.305. In addition, there is also a correlation between Attribution and Education level. The relationship has a significance value less than .05 and a Pearson correlation coefficient of r=.301. Moreover, there is a correlation between the difference between Expectations and Perceived Performance and the Years of Experience. The relationship has a significance value of .05 and a Pearson correlation coefficient of r=.289. Furthermore, there is a correlation between the difference between Subjective Disconfirmation and the Years of Experience. The relationship has a significance value of .05 and a Pearson correlation coefficient of r=.304. Finally, there is a correlation between Customer Satisfaction and the Level of importance QING Groep as partner. The relationship has a significance value less than .01 and a Pearson correlation coefficient of r=.537.
In conclusion, there is a high positive relation between Subjective Disconfirmation and the difference between Expectations and Perceived Performance. Therefore, we expect to accept H1: Positive levels of Expectations have a positive effect on Subjective Disconfirmation and vice versa. Further, there is a high negative relation between Satisfaction and Subjective Disconfirmation. Therefore, we expect to accept H4: Higher levels of Subjective Disconfirmation have a positive effect on Customer
Satisfaction. Because there is a high positive relation between Expectations, Perceived Performance and Satisfaction we expect to accept H3: Positive levels of Expectations have a negative effect on Customer Satisfaction and vice versa.
As mentioned before, there is no correlation between Attribution and Expectations, Perceived Performance and Satisfaction, and therefore we aspect to reject H2: Positive levels of Expectations have a positive effect on Attribution and vice versa as well as H6: Higher levels of Attribution have a positive effect on Customer Satisfaction.
4.3 Regression
In order to test the seven hypotheses a regression analysis was conducted. We used “process model 4” to test the mediation effect of Attribution and Subjective Disconfirmation in the relationship of Expectations, Perceived Performance and Customer Satisfaction. The outcome model is presented in Table 3 (Note that a more detailed model can be found in Appendix 2: Table detailed Regression Analysis).
We regressed the difference between Expectations and Perceived Performance on Subjective Disconfirmation, Attribution and Customer Satisfaction. In line with hypothesis 1, we found that positive levels of Expectations have a negative effect on Subjective Disconfirmation and vice versa (b = .968, p = .000). .968 means that two customers that differ by one unit on the difference between Expectations and Perceived Performance are estimated to differ by 0.968 units on Subjective Disconfirmation. The sign of b is positive, meaning that those relatively higher in exceeding Expectations estimated to be higher in their subjective Disconfirmation. This effect is statistically different from zero, t= 67.122 p = .000, with a 95% confidence interval from -0.940 to 0.998.
Contrary to hypothesis 2, no significant relationship was found between the difference in Expectations and Performance Perceptions and subjective Disconfirmation (b =-.058, p=0.731).
In line with hypothesis 3, we found that the total effect of the difference between Expectations, Perceived Performance and Customer Satisfaction is b = -.629, meaning two customers who differ by one unit in the difference between Expectations and Perceived Performance are estimated to differ by 0.629 units in their Satisfaction. The negative sign means that customers who perceive greater levels of exceeding Expectations experience lower levels of satisfaction. This effect is statistically different from zero, t=-5.244 p=.000 or between -.871 and -.387, and with 95% confidence.
Besides that, hypothesis 4 and 6 are rejected. No significant relationship was found between
Subjective Disconfirmation and Customer Satisfaction (b = -.330, p = .816) nor between Attribution and Customer Satisfaction (b = -.080, p = .518).
In line with hypothesis 5, the indirect effect of subjective Disconfirmation in the relationship the difference between Expectations and Perceived Performance and Customer Satisfaction is significant. The indirect effect of .320 (Appendix 2) means that two customers who differ by one unit in their exceeding Expectations are estimated to differ by .320 units in their Customer Satisfaction as a result of the tendency for those who perceive subjective Disconfirmation, which in turn translates into a higher satisfaction. This indirect effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence.
In line with hypothesis 7 concerning the indirect effect of Attribution in the relationship, the difference between Expectations and Perceived Performance and Customer Satisfaction is significant. The
exceeding Expectations are estimated to differ by .046 units in their Customer Satisfaction as a result of the tendency of those who perceive Attribution, which in turn translates into higher satisfaction. This indirect effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence.
Tabel 3 Regression Model outcome Satisfaction
1) H3: ∆ Expectations and Performance Perceptions predicts Customer Satisfaction – path c
a. F (9, 41) =12.468 p=<.001 𝑅"=.856
b. b =-.629 t (41) =-5,244 p<.001
2) H2: ∆ Expectations and Performance Perceptions variable predicts Attribution – path a
a. F (9, 41) =2.420 p=.030 𝑅"=.347
b. b =-.058 t (41) =-.347 p=.731
3) H7: ∆ Expectations and Performance Perceptions and Attribution together predicting Customer Satisfaction
a. F (11, 39) =9.848 p = <.001 𝑅"=.735
b. H6: Attribution predicts Customer Satisfaction – path b i. b = -.080 t (39) = -.652 p=.518
c. ∆ Expectations and Performance Perceptions no longer predicts Customer Satisfaction or is lessened predicting Customer Satisfaction – path 𝑐%
i. b = -.953 t (39) =-.697 p =.490
4) H1: ∆ Expectations and Performance Perceptions variable predicts Subjective Disconfirmation
a. F (9, 41) =729.126 p <.01 𝑅"=.994
b. b =.968 t (41) = 67.122 p=<.01
5) H5: ∆ Expectations and Performance Perceptions and Subjective Disconfirmation together predicting Customer Satisfaction
a. F (11, 39) =9.848 p=<.001 𝑅"= .735
b. H4: Subjective Disconfirmation predicts Customer Satisfaction – path b
i. b =.330 t (39) =.235 p=.816
c. ∆ Expectations and Performance Perceptions no longer predicts Customer Satisfaction or is lessened predicting Customer Satisfaction – path 𝑐%
i. b = -.953 t (39) =-.697 p =.490 N=51 customers. Unstandardized regression coefficients are reported.
Table 4: Descriptive Statistic Management Perception
__________________________________________________________________________________
Antecedent Mean Standard Deviation
__________________________________________________________________________________ Independent variables:
Subjective Disconfirmation 1.325 .106
Attribution 3.833 .707
∆ Expectations and Performance Perceptions .325 .106 Dependent variables: Satisfaction 6.250 .707 Retention 5.000 1.414 __________________________________________________________________________________ *Note N=2 4.4 Conclusion
The current study investigated the relationship of the difference between Expectations and Perceived Performance and Customer Satisfaction and examined the mediating roles of Attribution and
Subjective Disconfirmation in this relationship. Figure 3 provides the conceptual model from the literature section with the results as discussed in the Chapter 4.3 regression.
5. Discussion and conclusion
The purpose of this study was to examine the relationship between Expectations, Performance Perceptions and Customer Satisfaction and investigate the mediating roles of Attribution and Subjective Disconfirmation. The aim of the research was to provide new insights in the Customer Satisfaction and Service Quality literature by looking at the different roles customer Expectations and Performance Perceptions play as possible antecedents of customers’ Attribution, Subjective
Disconfirmation and Satisfaction.
We expected that Attribution and Subjective Disconfirmation act as mediators between Expectations, Performance Perceptions and Customer Satisfaction. We believe that our study has contributed to a more comprehensive understanding of Customer Satisfaction and Service Quality. As mentioned before in the literature study, little research in the Customer Satisfaction and Service Quality literature has been done about the Attributions consumers assign for satisfactory and unsatisfactory product or service performance (Oliver, 2010). This study has extended that work by exploring the effects of Attribution.
The research findings do support most of our hypotheses. We found that when customers perceive higher levels of exceeding Expectations, they experience higher levels in their Subjective
Disconfirmation (H1). This finding corresponds with the literature, when performance perceptions are more positive than what was expected, positive Disconfirmation occurs Oliver (2010). Moreover, we found that the customers who perceive greater levels of exceeding Expectations experience lower levels of satisfaction (H3). This also corresponds with the literature, Kotler (2012) describes Customer Satisfaction as: “The extent to which a product’s Perceived Performance matches a buyer’s
Expectations”. Additionally, we found evidence that the indirect effect of subjective Disconfirmation in the relationship between Expectations and Perceived Performance and Customer Satisfaction is significant (H5). Because this aspect mediates Expectations, Perceived Performance and Customer Satisfaction, this study has shown that Subjective Disconfirmation can be included as an intervening variable when investigating Customer Satisfaction. There needs to be attention to include Subjective Disconfirmation as a separate variable in further Customer Satisfaction and Service Quality research. Besides that, we also found evidence that the indirect effect of Attribution in the relationship
concerning the difference between Expectations and Perceived Performance and Customer Satisfaction is significant (H7). Attention to Attribution needs to be included in further Customer Satisfaction and Service Quality Research, as this aspect mediates Customer Satisfaction, Expectations and Performance Perceptions.
We did not find support for H2: “Higher levels of Attribution have a positive effect on Customer Satisfaction”, H4: “Higher levels of Subjective Disconfirmation have a positive effect on Customer Satisfaction” and H6: “Higher levels of Attribution have a positive effect on Customer Satisfaction.”
Common source bias could play a role for Attribution. The method used was self-report, which involves asking a participant about their feelings, attitudes and beliefs without researcher interference (Favero & Bullock, 2015). With Attribution, customers look for the cause of particular experiences when arriving at satisfaction judgments. It could be the case that customers do not want to revise their judgements and therefore are not completely honest.
The two schools of thought concerning how to measure Disconfirmation could play a role for not finding evidence for the significant relationship between Subjective Disconfirmation and Customer Satisfaction. Oliver (2010) states that Subjective Disconfirmation has to be included as a separate variable when investigating Customer Satisfaction in order to directly measure Disconfirmation. Pitt et al. (1997) and Swan and Trawick (1980) argue for the subtractive Disconfirmation approach, where Expectations and Perceived Performance are subtracted from each other. In this case, Disconfirmation is not directly measured and is not included as a separate variable. Thus, for Disconfirmation it depends how the measurement takes place. In this study we opted for the direct measurement of Disconfirmation. Customers described the performance of a service in terms of being better or worse than expected (Oliver, 2010). The outcome could differ if the indirect measurement is chosen. More research needs to be done on the differences between the direct- and indirect measurements of Subjective Disconfirmation.
With this information, we are able to give proper answers to the research questions:
a) What is the influence of the difference in Expectations and Performance Perceptions on Customer Satisfaction? We found that the total effect of the difference between Expectations and Perceived Performance is B = -.629, meaning two customers who differ by one unit in the difference between Expectations and Perceived Performance are estimated to differ by 0.629 units in their Satisfaction. The negative sign means that customers who perceive greater levels of exceeding Expectations experience lower levels of satisfaction.
b) What is the mediating effect of (i) subjective Disconfirmation and (ii) Attribution on the link between Expectations and Performance Perceptions on Customer Satisfaction?
The indirect effect is significant for both the (i) subjective Disconfirmation and (ii) Attribution in the relationship concerning the difference between Expectations and Perceived Performance and
Customer Satisfaction. Subjective Disconfirmation has an indirect effect of .320 (P=.000) (Appendix 2). Attribution has an indirect effect of .005 (P=.000) (Appendix 2).
In addition, this study has done research on Customer Retention and on the management perception of the Board of Directors of an engineering agency. Customer Retention concerns repeated purchase behavior (Hennig-Thurau, 2004). According to Kotler (1994), the key to Customer Retention is Customer Satisfaction. There is a positive relation between Customer Retention and Customer Satisfaction. The relationship has a significance value less than .01 and a Pearson correlation
difference between Customer Retention are estimated to differ by .288 units in their Satisfaction. So, the higher the level of satisfaction, the higher the level of Customer Retention.
Another finding is the differences between the perceptions of the Board of Directors and the customers of the engineering agency. According to Oliver (2010) the perceptions of the actor and observer are frequently at variance, and the consumer as actor is likely to draw different Attributional conclusions from the manager as observer. Interesting findings are the differences in perceived satisfaction and perceived Customer Retention. Looking at the means and standard deviations, the Board of Directors perceptions of satisfactions are higher than the perceptions of satisfaction of the customers. The mean of Customer Satisfaction by customers is M=5.652 (SD=.944) and the mean of Customer Satisfaction by the Board of Directors is M=6.250 (SD=.707) (Table 4). The Board of Directors perceptions of Customer Retention are slightly higher than the perceptions of Retention of the customers. The mean of Customer Retention by customers is M=5,652 (SD=.944) and the mean of Customer Retention by the Board of Directors is M=5.000 (SD=1.414) (Table 4). The management perceives some of the Service Quality aspects to be higher than the customer. When management knows exactly where Service Quality is deviating, this will improve their customer understanding and ability to influence the customer. Further, management would benefit from knowing which attributions its consumers assign for satisfactory and unsatisfactory product or service performance. The reason is that successful performance is differentially attributed to the customer by the customer and to the firm by the firm’s management. Failure however tends to be attributed to the firm by the customer and to some other entity (e.g. suppliers but also including the customer) by the firm (Oliver, 2010).
5.1 Practical implications
This research has indicated that there is a significant relationship between Customer Satisfaction and Customer Retention. Higher levels of satisfaction will provide higher levels of Customer Retention. If a company is able to increase the Customer Satisfaction, an organization is more likely to increase the repeated purchase behavior of the customers.
This research has indicated that there is a difference in perceptions of satisfaction and Customer Retention between the Board of Directors and the customers. Successful performance will be differentially attributed to the customer by the customer and to the firm by the firm’s management. Failure however tends to be attributed to the firm by the customer and to some other entity (e.g. suppliers but also including the customer). It is possible that the management thinks that some Service Quality aspects will be delivered and that the customer disagrees. When the management is able to identify the gaps and knows what levels of Service Quality are deviating, they are able to improve their Service Quality. This will improve the customer understanding and ability to influence the customer.
