Graduate School
Master of Science in Tourism and Hospitality Management Master Degree Project No. 2011:94
Supervisor: Tommy Andersson
Service Expectations and Perceptions of Exhibitors at TUR 2011 An Application of SERVQUAL Model on Service Quality Evaluation
Junling Qian and Ling Wang
Acknowledgements
We would like to show our gratitude to our Professor and Thesis Supervisor, Tommy Andersson, who provided excellent guidance, ideas, and feedbacks through the construction process of the thesis.
It is an honor for us to conduct the research in the Swedish Exhibition and Congress Centre, and we would like to thank the management team of TUR, especially Per Magnusson, the Exhibition Manager of TUR. He supported the research in a number of ways.
Moreover we want to thank John Armbrecht and Erik Lundberg who made valuable suggestions on literature and helped us to develop the ideas of the thesis topic.
Lastly, we offer our regards and blessings to all of those who supported us in any aspect during the project.
Junling Qian and Ling Wang
Abstract
In stressing the imperative role of service quality in obtaining competitive advantage for events industry, this paper, using the SERVQUAL model, measures the service quality of the annual tourism event, TUR, at the Swedish Exhibition and Congress Centre in the city of Gothenburg by ascertaining the gap between exhibitors‟ expectations and perceptions. It identifies the difference in the importance of the RATER dimensions in exhibitors‟
expectation and perception. Moreover, the study demonstrates the influence of cultural background and perceived changes on exhibitors‟ perception. A pre-event and a post-event online survey as well as an onsite interview were conducted to collect data.
Table of Contents
Abstract ... 1
ChapterⅠ Introduction ... 3
1.1 Research Questions ... 5
1.2 Purpose ... 6
1.3 Limitations ... 6
ChapterⅡ Literature review ... 7
2.1 Service and Service Quality ... 7
2.2 Customer Satisfaction and Service Quality ... 8
2.3 Service Quality Frameworks ... 9
2.3.1 Grönroos (1984): Technical and functional quality model. ... 9
2.3.2 Parasuraman, Zeithaml and Berry (1985): Gaps model of service quality. ... 9
2.3.3 Bateson (1995): The Servuction System. ... 12
2.3.4 Haywood-Farmer (1988): Attribute service quality model. ... 13
2.3.5 Spreng and Mackoy (1996): Model of perceived service quality and satisfaction. 14 2.3.6 Philip and Hazlett (1997): PCP attribute model. ... 16
2.3.7 Dabholkar, Shepherd & Thorpe (2000): Antecedents and mediator model. ... 17
2.4 Summary of Models ... 18
2.5 SERVQUAL Model as the Basis of the Proposed Model ... 19
Chapter Ⅲ Model and Hypotheses ... 21
3.1 Research Model... 21
3.2 Hypotheses Formation ... 21
Chapter Ⅳ Methodology ... 24
4.1 Sampling for Quantitative Study ... 24
4.2 Quantitative Study – Survey Instruments ... 24
4.3 Qualitative Study... 25
4.3.1 Interview. ... 25
4.3.2 Observation. ... 26
4.4 Statistical Techniques Applied ... 27
4.4.1 Summated scale. ... 27
4.4.2 Reliability coefficients. ... 27
4.4.3 Data recoding. ... 27
4.4.4 Compare means – One-way ANOVA. ... 28
4.4.5 Multiple regression. ... 29
Chapter Ⅴ Results, Analysis and Discussion ... 30
5.1 Sample Description ... 30
5.2 Reliability of the Measurement Scale ... 30
5.3 Sample Distribution ... 31
5.4 RATER‟s Influences on Overall Expectations ... 31
5.5 Gap between Expectation and Perception ... 35
5.6 RATER‟s Influences on Overall Perception ... 37
5.7 Satisfaction 2011 vs. Satisfaction 2010 ... 41
5.8 Repeat Exhibitors vs. First-time Exhibitors ... 41
5.9 Domestic Exhibitors vs. Foreign Exhibitors ... 42
5.10 Perceived Changes Influence the Overall Perception ... 43
Chapter Ⅵ Conclusions and Recommendations ... 45
6.1 Conclusions ... 45
6.2 Recommendations for Management ... 47
6.3 Recommendations on Future Research ... 49
Chapter Ⅶ References ... 51
Chapter Ⅷ Appendices ... 57
ChapterⅠ Introduction
TUR, started from 1984, is an annual business fair for the travel, tourism and meeting industries. It provides an international meeting place for exhibitors, from all over the world, to meet trade visitors and the general public, which ensures the fair will attract keen interest among the Nordic countries. TUR has been recognized as the leading fair of the industry and is held at Swedish Exhibition and Congress Centre (the Centre) in Göteborg, a venue with ample experience of holding large and small conferences and congresses. TUR runs for four days at the end of March, consisting of activities and services such as Meetings@TUR for professional buyers and sellers of international conferences and business meetings, BookTUR to offer an opportunity for the exhibitors to create attractive offers only purchasable at the fair, seminars with different topics regarding the future trend of the industry, etc. In 2011 there were 894 exhibitors from 79 countries and regions represented at the TUR (TUR, 2011).
Still the challenges that TUR faces should be seen against all the success. TUR has been held for 27 years and it is stimulating for the organization to be continuously attractive while competing with similar fairs like HORECA Malmo, MATKA Nordic Travel Fair in Helsinki and so on. In addition, when discussing with the management of TUR, we found that the organization only has post-event evaluation on the perception aspect, just as Getz anticipates event managers always rely on visitor surveys to measure customer satisfaction levels and to identify any problem areas, and the input of staff and volunteers can also be important (Getz, Neill, & Carlsen, 2001). But there is no previous study about expectations of exhibitors and thus no comparison between expectations and perceptions has been conducted. Short of identifying the gap between exhibitors‟ expectation and perception would be challenging for the organization to control the service quality and achieve higher level of exhibitors‟ service satisfaction.
However, service quality is a critical concept increasingly gaining its importance in today‟s business environment. The global development of economies leads to expanding of the markets for services and increasing competition within these markets. As a result, the importance of service quality now deserves more attention for organizations wishing to gain competitive advantages. It has been raised to strategy level with guidelines that provide orientations for everyone in the organization (Clement & Selvam, 2006). But unlike product quality, service quality is complicated to define because of the highly transitory and
intangible nature of most services. As Parasuraman, Zeithaml, and Berry (1985) pointed out, service quality is “the degree and direction of discrepancy between customers‟ service perception and expectations”, there are two pivotal concepts always linked with the evaluation of service quality: customer‟s expectations and perceptions. The discrepancy between these two concepts is recognized as the fifth gap in service quality management.
Expectations are dynamic. Customer expectations may differ among people from different countries and cultural background where service treatment standards may differ a lot. Service expectations also derive from many other sources, such as personal needs, perceived service alternatives, customer self-perceived service role, service promises, word of mouth communication, past experience, and situational factors beyond the control of the service provider (Zeithaml, Bitner, & Gremler, 2009).
Perceptions are always considered relative to expectations and vary as significantly as expectations. Evaluation of service is based on customers‟ perception of Reliability, Assurance, Responsiveness, Empathy, and Tangibles. Customers will have perceptions of single, transaction-specific encounters as well as overall perceptions of a company based on their cumulative experiences over a period of time which might include multiple service encounters of the organization, or of the region, or even over the whole industry (Zeithaml et al., 2009). Many factors, like management systems, staff and volunteer actions, and facilities etc. may directly affect the customers‟ perception of quality and therefore their level of event satisfaction. Thus to create a lasting impression in the customer‟s mind, organizations need to add value in the views of the customers that good service quality and satisfaction is achieved, by concentrating on the development and provision of these more relational elements of the service encounter (Neill, Getz, & Carlsen, 1999).
Service quality in event management is our primary interest in constructing this paper. Events, as a unique section in tourism industry, can be described as non-standardized services in which the knowledge, behavior, and commitment of the service providers is crucial (Bejou, Edvardsson, & Rakowski, 1996). There are studies that assess service quality of events by identify predefined targets of service quality in the actually delivery (Chelladurai & Chang, 2000). But Zeithaml et al. (2009) argue that the discussion of service quality and customer satisfaction should be based on what customers actually perceive rather than on a set of standard criteria.
Service quality in tourism industry is an undeniable key factor for long-term business success.
For annual events in particular, deliver service of desirable quality is a vital basis to attract new participants and encourage repeat visits. Studies have been conducted to demonstrate the positive relationship between customer satisfaction and a desire to return in the field of tourism (Kozak, 2001; Yoon & Uysal, 2005). The first satisfactory impression gained from the initial experience has direct impact on the customer‟s decision to become a repeat visitor (Moutinho, 2001). And increasing satisfaction will result in increased repeat visits in the absence of counter moves by competitors (Kozak, 2001).
1.1 Research Questions
Better management of customers‟ realistic expectation can help to narrow the gap between expectations and perceptions of service quality, and thus customers‟ average satisfaction is likely to rise to a level that motivate their repeat visits to the next TUR fair. Therefore, we are interested in the following questions,
1. What is important in exhibitors‟ expectation?
2. Are exhibitors‟ actual perceptions in line with the expectations?
3. What can the management team of TUR do in the future to design their service and minimize the gap between expectation and perception?
TUR is a fair in change (Magnusson, 2011). It constantly self-improves by adapting valuable suggestions from fair participants. There are some new features implemented in TUR 2011, which include invitations sent to more than 3000 VIPs, establishment of Travel Trade Council, and changed opening hours on Friday. When an event organizer introduces a new or improved service, the marketing and sales departments must make the service appealing to build up new customer relationships and meanwhile consolidate existing customer relationships. However, the organization cannot afford to raise the expectations above the level at which it can consistently perform (Zeithaml et al., 2009). If the external communication does not deliver the actual image of the offering, unrealistic expectations are likely to be set up, which subjects the organization to a huge risk that customers are disappointed, satisfaction level is low, and bad impression among those customers is created.
Further consequences such as word-of-mouth effect and reduction in repeat visits can be estimated. Therefore another question is proposed as,
4. Are exhibitors‟ perceptions of changes related to overall perception on service quality of TUR 2011?
1.2 Purpose
The purpose of our study covers the following points,
Discuss the theories behind customers‟ expectations and perceptions of service quality.
Analyze the level of importance of the service dimensions to exhibitors‟ expectations on TUR 2011.
Compare the average gap score between expectations and perceptions.
Evaluate the overall satisfaction of TUR 2011.
Provide recommendations for the management team of TUR and for future research connected to the topic.
1.3 Limitations
This paper is based on data collected before, during, and after the fair mainly through Webropol. There are difficulties to obtain the same number of respondents for expectation and perception. Neither are we able to have the same respondents for the two sets of data.
Conclusions are based on the findings from TUR 2011. Therefore they cannot be generalized to all business fairs in travel, tourism and meeting industries. However TUR is an interesting case which can provide an insight for future studies.
ChapterⅡ Literature review
2.1 Service and Service Quality
Service quality has been studied by many researchers within the field of business management for years as the market becomes increasingly competitive and management has expanded its focus from internal performance to external interests including customer satisfaction and customers‟ perceptions of service quality (Grönroos, 1992). Researchers proposed different views on the definitions and characteristics of service and service quality, however, no consensus has been reached on how service quality is best conceptualized.
Therefore it is necessary to start the review from the discussion of the definition of service for our study purpose.
Grönroos (1990, p. 27) described service as “an activity or series of activities of more or less intangible nature that normally, but not necessarily, take place in interactions between the customer and service employees and/or systems of the service provider, which are provided as solutions to customer problems ”. Zeithaml and Bitner (1996, p. 5) mentioned in their book that service are “deeds, processes, and performances”. Yong (2000), based on the definitions offered by different researchers, drew a conclusion on the definition of service: a service, combined with goods products, is experienced and evaluated by customers who have particular goals and motivations for consumers for consuming the service. These three are examples of the diversified definitions of service. There are still many others not listed here.
Similar to service‟s definition, opinions on the definition of service quality and how it should be measured are lack of consensus, exactly as Reeves and Bednar (1994, p. 436) concluded in their work that “there is no universal, parsimonious, or all-encompassing definition or model of quality”. However, as observed by Chang, Chen and Hsu (2002), the traditional notion of service quality by Parasuraman et al. (1985) is most commonly accepted. Shostack (1977) claimed that “intangibility is not a modifier; it is a state”. She means that intangibility is part of the nature of a service. The service is rendered by service provider and experienced by the customer and it cannot be kept in stock, cannot be touched, tasted or tried on for size. Her notion emphasized the role of the experience, which is reflected in the traditional notion:
service quality is viewed as the customer‟s perception of service excellence combined with certain comparison standards such as his past experiences. Therefore, according to this notion,
service quality can be understood as a conception inevitably influenced by customers‟
subjective perceptions.
2.2 Customer Satisfaction and Service Quality
Although the traditional notion is widely accepted, it is not bullet-proof. It has been facing many challenges. Oliver (1997) pointed out the traditional definition failed to distinguish service quality from customer satisfaction. According to Oliver, perception of quality may come from external mediation rather than experience of service, consumers must experience satisfaction in person. Moreover, satisfaction level is evaluated not only based perceptions of excellence, but also involving many factors such as personal needs, product category norms, and expectations of service quality. This challenge was supported by Bitner and Hubbert (1994) who claimed quality are mainly cognitive, satisfaction is an affective experience.
Therefore, it is important to differentiate customer satisfaction and service quality.
Facing many challenges, Zeithaml et al. (2009) updated their work and defined customer satisfaction as the customer‟s fulfillment response. It could be understood as a judgment whether that product or service has met the customer‟s needs and expectations. They further differentiate the underlying causes and outcomes of the two concepts:
Service quality is a focused evaluation that reflects the customers’ perception of reliability, assurance, responsiveness, empathy, and tangibles. Satisfaction, on the other hand, is more inclusive: it is influenced by perceptions of service quality, product quality, and price, as well as situational factors (the uncontrollable circumstances, such as weather condition) and personal factors (customers’
emotional state and their expectations)… Satisfaction is the customer’s fulfillment response… It is the consequence of the comparison between the needs, expectations and the perceptions. (p. 103 – 104)
The two concepts are fundamentally different as well as closely connected. Perceived service quality is a component of customer satisfaction, together with others such as product quality, price, customer emotion, perceptions of equity and attributions for service success or failure.
2.3 Service Quality Frameworks
Several studies present the idea to measure and enhance service quality by identifying its dimensions. This chapter attempts to study different service quality models followed by a brief summary and discussion of the models.
2.3.1 Grönroos (1984): Technical and functional quality model.
According to Grönroos there are two dimensions of service: technical and functional.
Technical service quality refers to the quality of the outcome of a service. Functional service quality concerns with quality of the process of the customer-employee interaction. Customers assess the service quality by comparing the perceived performance with their expectation in terms of reliability, tangibles, responsiveness, assurance and empathy. McDougall and Levesque (1994) added a third dimension to Grönroos‟s model: physical environment, developing their three-factor model of service quality.
Figure 1. Technical and Functional Quality Model
2.3.2 Parasuraman, Zeithaml and Berry (1985): Gaps model of service quality.
The gaps model of service quality proposed by Parasuraman et al. (1985) provides a structured, integrated way of viewing services. The model consists of the following gaps:
Customer expectation vs. management perception (Gap 1), management perception of
Expected service
Techinical Quality
Functional Quality
Perceived Service Perceived
Service quality
Image Traditional Marketing activities
(advertising, field selling, PR, Pricing) and external influence by traditions, ideology and word of mouth
Source: Grönroos (1984)
What? How?
customers‟ expectations vs. service quality specifications (Gap 2), service quality specifications vs. service delivery (Gap 3), service delivery vs. external communications (Gap 4), and finally the gap between customers‟ expectations and service perceived (Gap 5).
Figure 2. Gaps Model of Service Quality
The customer gap (Gap 5) represents the potential discrepancy between the expected and perceived service from the customers‟ view. Because some sources of the customers‟
expectations are marketer-controlled factors (such as pricing, advertising, sales promises) and some factors the marketer has limited ability to affect (for instance, word-of-mouth communications, personal needs and past experience), it is possible to assess the discrepancy between customers‟ perceptions and expectations in the service delivery process (Zeithaml, Bitner, & Gremler, 2006, p. 34). To measure the service quality gap the SERVQUAL method is dedicated, which provides a foundation for research that concerns to the creation of quality among service industries. According to SERVQUAL method, there are five dimensions (tangibles, reliability, responsiveness, assurance and empathy) of service quality that are applicable to service industry (Parasuraman et al., 1985). The customer gap is the fifth one in
the service quality gaps model and also the basis for the gaps model. It is affected by four provider (internal) quality gaps which interact with one another in many ways (Zeithaml et al., 2006, p. 34).
Provider gap 1 states the difference between customers‟ expectations of service and the expectations perceived by manager. The gap occurs because managements‟ inaccurate understanding of what customers exactly expect. There are some conceptual factors may contribute to this gap: lack of marketing research orientation, inadequate upward communication, too many levels of management which separates contact personnel from top managers (Parasuraman et al., 1985, p. 52), in sufficient relationship focus, or inadequate service recovery (Zeithaml et al., 2006) etc. Zeithaml also poited out the strategy for closing this gap which includes multiple marketing reserch method, building strong relationships with customers and implying recovery strategies for service failure (Zeithaml et al., 2006).
The second provider gap takes place when translating customers‟ expectations into service quality specifications. Gap 2 mainly reflected by poor service design, lack of customer-driven standards, and inappropriate pysical evidence and servicescape etc. (Zeithaml et al., 2006, p.
38) Authors across the literature suggest some implements, such as service blueprinting and quality function deployment, to design services without oversimplification and incompleteness in order to avoid this gap (Zeithaml et al., 2006). Besides, they indicate that the importance of managers‟ commitment to improve service quality should not be ignored while setting the sevice standards to meet customers‟ perceptions (Zeithaml, Parasuraman, &
Berry, 1990).
The discrepancy between service specifications and actural service performance by company employees is the provider gap 3. This gap happens commonly in service business when empolyees are unable or unwilling to perform the service at the desired level (Zeithaml et al., 1990, p. 89). There are four main factors that may lead to gap 3: deficiencies in human resource policies, customers who do not fulfill roles, problems with service intermediaries and failure to match supply and demand (Zeithaml et al., 2006). Solutions to reduce this gap are provided, such as provide role clarity to employees and minimize role conflict, improve employee-technology-job fit, measure and reward service performance, empower service employees and encourage teamwork, manage extermal customers (Zeithaml et al., 1990).
Provider gap 4 illustrates the difference between what the company promises about a service and what it actually delivers (Zeithaml et al., 1990, p. 115). It happens because a service company might overpromise or misrepresent customers through its media advertising and other external communications, and thus may potentially raise customers‟ expectations about service delivery. Broken promises can be made due to many factors: overpromising in advertising or personal selling, inadequate horizontal communication, ineffective management of customer expectations and lack of integrated service marketing communications (Zeithaml et al., 2006, p. 42). To escape gap 4, the service company should manage all communications to customers so that keep the promises be accurate and realistic.
2.3.3 Bateson (1995): The Servuction System.
According to “the Servuction System” by Bateson (1995), service is based on the delivery of a bundle of service benefits which may come from a variety of sources. Two aspects are identified: invisible aspect and visible aspect. Visible aspect is represented by the inanimate environment in which the service encounter takes place and the contact personnel who actually deliver the service. Invisible aspect, on the contrary, refers to employees‟ back-stage performance and supporting systems. The invisible aspect of the business is inextricably linked to the visible aspect. Delivery of service quality can only be ensured when the invisible aspect and the visible aspect are well integrated. The combination of the two aspects influences customers‟ experience and their perception of service quality which consequentially affect their spending patterns and future decisions. In this model, the interrelationship between customers is another influential factor on the focal customer‟s experience. But we consider this factor as one outside of the scope of our study.
Figure 3. The Servuction System
2.3.4 Haywood-Farmer (1988): Attribute service quality model.
Haywood-Farmer proposes that a service organization has “high quality” if it meets customer preferences and expectations consistently. He differentiates service organizations according to their relative degree of service customization, labor intensity, contact and interaction between the customer and the service organization. His model categorizes service quality components into three groups: physical facilities and processes, behavioral aspects, and professional judgment. Each of these components contains multiple factors shown in the following figure. Services of law interaction and labor intensity and high customization, such as legal consulting, are closer to the component of professional judgment. He highlights that too much emphasis on any one of the components while let others be excluded may lead to negative impact on customer‟s perceptions. For example, if a service organization emphasizes heavily on procedure, customers are likely to perceive that the organization is rigid and inflexible.
Figure 4. Attribute Service Quality Model
2.3.5 Spreng and Mackoy (1996): Model of perceived service quality and satisfaction.
This model is based on the model proposed by Oliver (1993) who claims the distinction between service quality and satisfaction. According to Oliver, service quality is formed by a comparison between ideals (desires) and perceptions of performance regarding quality dimensions, while satisfaction is a function of the disconfirmation of predictive expectations regarding both quality dimensions and non-quality dimension. Oliver‟s model does not explain the relationship between desire congruency/disconfirmation and overall satisfaction, neither does it specifies the effect of expectation on perceived performance. Spreng and
Mackoy‟s study confirms the distinction between the two concepts and further improves the model by demonstrating that satisfaction is directly influenced by desire congruency/disconfirmation, which means when the desire is met by perceived performance the overall satisfaction level will be higher than under the situation where the desire is not met. They also claim the influence of expectation on perceived performance. Moreover, Spreng and Mackoy‟s model points out the dual effects of expectation on satisfaction. The explanation is:
The practice to lower expectations and over-deliver service to create higher satisfaction is based on the negative impact of expectations on satisfaction through disconfirmation (i.e., lower expectations cause higher positive disconfirmation, which causes higher satisfaction).
However, expectation could also produce positive impact on perceived performance and desire congruency and by this path it could have positive impact on satisfaction. If expectations are lowered, the firm is also risky to lower the perceptions of performance, which will consequently lower the satisfaction. Therefore managers should make a balance between the positive and negative impacts of expectations.
Figure 5. Model of Perceived Service Quality and Satisfaction
In agreement with their contributions, we doubt the necessity to single out the concept of desire in the model, and find ourselves more agree with the classification of expectations by
Wilson et al. (2008), where expectations are identified as five levels that a customer may have towards a firm or a particular service: ideal expectations/desires, normative expectations, experience-based norms, acceptable expectations, and minimum tolerable expectations.
Therefore desire is actually one type of expectation.
2.3.6 Philip and Hazlett (1997): PCP attribute model.
Figure 6. PCP Attribute Model
Philip and Hazlett proposed a hierarchical structure formed model, which based on three main classes of attributes: Pivotal, Core and Peripheral. The whole model is built on the premises of using a combined scale instead of separate scales to measure the gap between expectations and perceptions and attaching different weights to individual dimensions to indicate the importance. According to the model, every service consist three albeit overlapping areas which includes all the dimensions and attributes that used to define service quality. The three areas also division into three hierarchical levels: pivotal (representing the outputs of a service organization), core and peripheral (conjointly representing the inputs and processes of a service organization).
Pivotal attributes, located at the apex of the oval-shaped model, exert the highest influence on customers‟ consuming in the first instance and customers‟ satisfaction level. That is to say, whether a service is satisfactorily delivered mainly depends on whether the customer receives
the output for which he/she originally approached the organization. Core attributes, together with the third level of the model: peripheral attributes, assume lower degree of influence on customers‟ satisfaction level in the first instance but the influence may grow rapidly during repeat service consuming. In short, the customer‟s satisfaction level may be determined more by the output (pivotal attribute) of the service, and (relatively) less by the personnel and the organizational structures (core and peripheral attributes) involved.
Thus, the challenge that any service confronted with, is to delight the customer in all three areas (pivotal, core and peripheral attributes) so that they could become 100 per cent satisfied of the service. However, the P-C-P model does not and cannot provide working attributes for each of the three categories of attributes. Instead it depends on individual service sectors to select the attributes that fit best into each of the three categories. Just as the authors point out:
there is no magic recipe, formula or blueprint which can be applied to the service sector as a whole.
2.3.7 Dabholkar, Shepherd & Thorpe (2000): Antecedents and mediator model.
Figure 7. Antecedents and Mediator Model
According to Dabholkar et al. (2000), factors related to service quality should be viewed as antecedents to overall evaluations of service quality rather than its components. That is to say, customers form a separate overall evaluation of the service quality at the same time of evaluating different factors related to the service, rather than form a straightforward sum of the components. This research also find out that customer satisfaction strongly mediates the
effect of service quality on behavioral intentions, while service quality is more closely related to specific factor evaluations about the service. At the same time, it illustrates the importance of measuring customer satisfaction separately from service quality when determine customers‟
evaluations of service.
2.4 Summary of Models
Having all these service quality models in mind, it is clear that no consensus has been reached on the conceptual model of service quality and the method of measuring service quality.
Putting difference aside, it can be observed that the descriptions of service quality dimensions in those models are to some extent overlapping each other. Grönroos (1984) offers two dimensions of service quality: technical dimension and functional dimension, one from the perspective of service result and the other from that of service process, which is further developed by McDougall and Levesque (1994) with the dimension of physical environment.
Parasuraman et al. (1985) include five dimensions into their SERVQUAL model: tangibles, reliability, responsiveness, assurance, and empathy. Bateson (1995) discusses the visible and invisible sides of service which imply the dimensions of physical environment, service process and service outcome. Haywood-Farmer (1988) categorizes service quality components into three groups: physical facilities and processes, behavioral aspects, and professional judgment. Philip and Hazlett (1997) also include service environment, service process and service outcome in their Pivotal-Core-Peripheral model. Dabholkar et al. (2000) included physical aspects, reliability, and personal interactions.
Another similarity among these models is that they all support the view that service quality should be evaluated by comparing service quality expectations with actual perceptions of customers who have experienced the service offerings.
2.5 SERVQUAL Model as the Basis of the Proposed Model
Among the discussed models, most of them do not offer an explanation on how to measure service quality nor do they provide ideas and procedures for identifying service quality problems and directions for future improvements. However the gap model offered by Parasuraman et al. (1985) is an exception. This model is an analytical tool which allows researchers to identify possible service quality gaps in a systematical way and assess service quality by analyzing the variables affecting the expectations and perceptions. The model has been tested by Parasuraman et al. (1985) in several cases, including credit card companies, securities broker, retail banks, etc.
But it does not mean that the theory of Parasuraman et al. (1985) is bullet-proof.
SERVQUAL model has faced challenges from many researchers. Babakus and Boller (1992) fail to replicate the five-dimensional factor structure of SERVQUAL and produce only two dimensions by their study. According to Brown, Churchill, & Peter (1993), the use of difference between scores in the model may cause problems such as reliability, discriminate validity, spurious correlations, and variance restriction. Buttle (1996) criticizes on SERVQUAL that the five dimensions are not universals and the model fails to draw on established economic, statistical and psychological theory. Oliver (1997) points out that the model has difficulty to distinguish service quality from customer satisfaction. Carmen (1990) claims that the model‟s five dimension are inadequate to meet the needs of service quality measurement and it may be problematic to measure expectation with SERVQUAL. This opinion is agreed by Yong (2000) who highlights that customers‟ needs are not always easy to identify, and incorrectly identified needs may result in measuring conformance to a specification that is improper. Yong (2000) mentions that due to the fact that expectations may fluctuate greatly over time, the validity of the outcome from SERVQUAL measurement needs to be questioned.
However, SERVQUAL model also has much support from numerous researchers. Avkiran (1994) demonstrates the model in banking service. Babakus and Mangold (1992) and Saleh and Ryan (1991) test the model in hospitality industry. Finn and Lamb (1991) evaluate the model in retailing service. Johns and Tyas (1996) use the gap model to differentiate the service quality in food outlets. Johnson and Sirikit (2002) find the model helps the telecommunication industry to achieve sustainable competitive advantage by measuring the service quality and suggestions can be drawn based on the assessment. These studies
demonstrate that SERVQUAL model is able to provide a comprehensive conceptualization of service quality and it functions as a sufficient instrument for the assessment of service quality.
Chapter Ⅲ Model and Hypotheses
3.1 Research Model
The design of the model is based on the combination of the research problems and research purpose proposed in Chapter Ⅰ, and the Gap Model proposed by (Parasuraman et al., 1985) with the emphasis on the Gap 5 between expectations and perceptions. The aim of the model is to present the relationships identified by previous researches and visualize the relationships to be examined in this research. Moreover, it highlights the areas that are not explained by previous researches. For example, the indirect influences of past experience and cultural background on expectations have been long established, but no research specifies their direct relationships with perceptions. In the following section the formulation of the hypothese is described in the order marked in the model.
Figure 8. Research Model
3.2 Hypotheses Formation
To answer the first research question, “What is important in exhibitors‟ expectation?” The first hypothesis is formulated as below. The intention is to analyze RATER1 and their relationships with exhibitors‟ overall expectation of TUR 2011.
1 RATER is the abbreviation of reliability, assurance, tangibles, empathy and responsiveness.
H1: Expected RATER positively influence exhibitors‟ expectations of TUR 2011.
To answer the second question, “Are exhibitors‟ actual perceptions in line with the expectations?”, findings need to be based on the analysis of the gap score between expectation and perception. To gain a deeper insight of issue, it is necessary to identify the quality levels of reliability, assurance, tangibles, empathy and responsiveness and their relation with customers‟ perceptions. Taking these into account, the second and third hypotheses are to be:
H2: There is no gap between exhibitors‟ expectation and perception on TUR 2011.
H3: Perceived RATER influence exhibitors‟ perceptions of TUR 2011.
Customer expectation can be influenced by multiple factors, such as word of mouth communications, personal needs, and past experience (Parasuraman et al., 1985). In the research, we found that many of the exhibitors have participated in TUR before and many of them even participate since the very first TUR. These exhibitors have seen the growth of TUR in the past years. Therefore, we would like to know to which extent the past experiences influence the gap between expectation and perception. Consequently hypotheses are derived to assess the situation in TUR 2011:
H4: Exhibitors‟ satisfaction level on TUR 2011 has no difference with the satisfaction level on TUR 2010.
H5 a: Expectations of TUR differ between repeat participants and first-time participants.
H5 b: Perceptions of TUR differ between repeat participants and first-time participants.
Cultural background may also significantly impact the formation of expectation. Customer expectations may differ among people from different countries and cultural background where service treatment standards may differ a lot. Expectations are dynamic, so evaluations may also shift over time – from person to person and from culture to culture (Zeithaml et al., 2009). Therefore the next hypotheses are:
H6 a: Expectations of TUR differ between domestic exhibitors and foreign exhibitors.
H6 b: Perceptions of TUR differ between domestic exhibitors and foreign exhibitors.
In addition, the new features implemented in TUR 2011 are meant to provide better service to the exhibitors. To discover the effect of the changes on the exhibitor‟s perceptions and answer research question 4, “Are exhibitors‟ perceptions of changes related to overall perception on service quality of TUR 2011”, so the last hypothesis is formulated as:
H7: Exhibitors‟ satisfaction of the perceived changes influences their overall perception of the service quality of TUR 2011.
Chapter Ⅳ Methodology
To assess the service quality of the Centre at TUR 2011, both quantitative and qualitative studies were carried out. Four specific instruments in temporal sequence: a pre-event online survey of exhibitors‟ expectations on TUR 2011 (Appendix 1 & 2); observation logbooks;
several onsite interviews (Appendix 5); and a post-event online survey of exhibitors‟
perceptions on the fair (Appendix 3 & 4). Relevant statistical techniques are utilized in analysis.
4.1 Sampling for Quantitative Study
Sample frame of potential respondents and their email addresses was compiled according to the following criteria:
They had to be the exhibitors who will present on TUR 2011.
They had to be selected randomly.
A list of 500 email addresses was provided to the researchers by the TUR management team, thus representing 55.9% of the total number of exhibitors that participated in TUR 2011.
4.2 Quantitative Study – Survey Instruments
In this paper, customers‟ expectation and perception of the service they receive is established by using a multiple-item scale, SERVQUAL, for assessing the service quality of TUR 2011.
The original SERVQUAL finalized by Parasuraman, Berry and Zeithaml (1988) contains two sets of Likert-type scales, each set having twenty-two questions. One of the two sets is intended to measure customer‟s expectations and the other is to measure customer‟s perceptions. The questions are designed in correspondence to the RATER dimensions embodying all the ten factors mentioned by Parasuraman et al. (1985). The structure is summarized by Van Iwaarden et al. (2003) and Shahin (2006) as follows,
1) Tangibles: physical facilities, equipment and appearance of personnel.
2) Reliability: ability to perform the promised service dependably and accurately.
3) Responsiveness: willingness to help customers and provide prompt service.
4) Assurance: competence, courtesy (knowledge and courtesy of employees and their ability to inspire trust and confidence), credibility and security.
5) Empathy: accessibility, communication, understanding the customer, caring and individualized attention that the firm provides to its customers.
The design of the online survey is based on this structure and each question is presented with a seven-point measurement scale describing the level of agreement from customer‟s perspectives. To adapt to the situation of TUR, the twenty-two items of the SERVQUAL scale are condensed into twenty. TUR has no convention on insisting on error-free records. It is less appropriate to include it into the survey. Furthermore, it is of no practical meaning to ask their expectations and perceptions on the personal attention given by TUR‟s employees.
The objective of the research contains the mission of providing suggestions for future improvement of TUR. Even if “personal attention” is identified of huge importance, it is not realistic to provide personal attention to every single participant, considering the huge number of exhibitors. The exhibitors in TUR are all registered as a single organization or as a company under a big association. Therefore it is more appropriate to use the item of
“individual attention”. As a result, the researchers decide to exclude “error-free record” under reliability and “personal attention” under empathy from the survey. In addition, a number of questions were asked to demographic profile of respondents, overall impression of service quality on TUR 2011, suggested improvement of the office equipment and further commands of the fair etc. The goal is to compare the scores of different service dimensions and determine the average gap score between customers‟ perceptions and expectations for each dimension.
The expectation survey was sent out on 10th of March, 2011, two weeks before the opening of TUR and the perception survey was sent out on 4th of April, 2011, one week after the ending of TUR.
4.3 Qualitative Study 4.3.1 Interview.
In-depth interview is a qualitative research technique that involves conducting intensive individual interviews with a small number of respondents to explore their perspectives on the research topic (Boyce & Neale, 2006). This instrument is ideal for investigating personal and confidential information which is unsuitable to discussion in a group. It is valuable for doing research on people with busy lifestyles who would be unlikely to attend a focus group.
Taking account of the fact that exhibitors at TUR are exposed to a large amount of visitors during the four days, it is more practical to conduct individual interviews instead of arranging a focus group. Moreover, the interviews conducted were semi-structured and probing techniques were employed to evoke additional information from the respondents (Blumberg et al., 2008).
Interviewees were selected at random, and the balance between Swedish exhibitors and exhibitors from the other countries was kept to be more representative. Interview technique was improved after the second interview. The second interview actually started very casually.
It was outside of the interview list and the researchers did not even mean to interview that exhibitor at the very beginning, but then got inspired by the casual chat and mixed interview questions in the chat. It was then found much easier to get the needed information by this kind of casual interview than doing it in a standard interview way. Therefore strategy changed: chat and memorize the key information instead of taking notes and recording the voice; discuss, summarize and write down the information immediately after each interview when the memory is fresh. The interviews turned out to be a pleasant rewarding experience for both interviewers and interviewees. Six interviews were conducted from 25th Mar. to 27th Mar. This approach ended with five concrete interviews, and one refusal because of the sixth interviewee‟s inability to provide objective opinions due to the cooperation relationship between their company and the Centre.
4.3.2 Observation.
Observation is one of the research methods in which researcher getting to know the people they‟re studying by entering their world and participating either with or without subjects‟
awareness (Lewis-Beck, Bryman, & Liao, 2004). Through observation, the researchers are able to gain first-hand knowledge of the activities, processes, operations of the service system and service environment. It can produce a depth of detailed information about all aspects of customers‟ behavior which facilitates the researchers‟ evaluation of the service quality by adding a facet in the analysis. Based on the above reasoning, researchers were walking around the venue every other hour and complete the observation forms during the four-day fair.
4.4 Statistical Techniques Applied
To answer the research questions and test the seven hypotheses, various statistical techniques are utilized in the analysis process, including summated scales, reliability coefficients, data recoding, mean comparison, multiple regression, and one-way ANOVA. SPSS 19 and Excel are employed to process the collected data.
4.4.1 Summated scale.
The use of a single variable to represent a concept only extracts a particular aspect of that concept and the measure of that concept may be inaccurate. In order to increase validity and reliability and reduce measurement errors, multiple variables are joined in a composite measure to represent the concept (Hair et al., 2008). There are five dimensions of service quality. Therefore in linear regression analysis five summated scales are computed with the names of the five dimensions: Tangible, Reliability, Responsiveness, Assurance, and Empathy. Each summated scale is represented by four questions which are called indicator variables. This is done through SPSS‟s „Compute Variable‟ feature. The summated scale value is the mean of the indicator variables. For example, the summated scale of Tangible is the mean of the questions regarding to Equipment, Accessibility, Employee appearance, and Instruction signs in the working questionnaire. Overall expectation is also created with summated scale by calculating the mean of all the twenty items.
4.4.2 Reliability coefficients.
Cronbach‟s alpha is a measure of internal consistency, which ranges in value from 0 to 1 and may be used to describe the reliability of factors extracted from multi-point formatted questionnaires or scales (i.e., rating scale: 1 = poor, 5 = excellent) (Blumberg, Cooper, &
Schindler, 2008). In order to describe a measure or scale as having good inter-item reliability or internal consistency, the value of Cronbach‟s alpha should be at least .70. The purpose of calculating the Cronbach alpha is to assess the internal consistency of measurement of the concept.
4.4.3 Data recoding.
Data recoding was used when comes to mean comparison between two variables that are measured with two different Likert scales, one with 7 points and the other with 5 points. For
example, on a Likert scale of the variable “overall impression of TUR 2011”, 1=very bad, 2=bad, 3=somewhat bad, 4=neutral, 5=somewhat good, 6= good, and 7= very good. We want to recode that into 5 values, the same as in variable “overall impression of TUR 2010”:
1=very bad, 2=bad, 3=neutral, 4=good and 5=very good. So we went to SPSS‟s “recode into different variables” feature and changed some of the categories into a single category. We recoded the original variable into a new variable, since we want to retain the original variable status for checking that the recoding has been done correctly. The instructions for each recoded variable were written in the code book.
4.4.4 Compare means – One-way ANOVA.
The Means procedure of SPSS calculates subgroup means and related univariate statistics for dependent variables within categories of one or more independent variables (SPSS 19). By comparing means, we could get summary statistics for dependent variables within the levels of one or more independent variables. For example, survey results, collected from two different groups of respondents, could be measured and performed in a one-way analysis of variance to see whether the means differ.
Mean is calculated according to the formula: =
Since we have both pre-event and post-event online surveys, which are considered to base on two different samples due to the anonymous and voluntary nature of the survey processes, comparison in SPSS and Excel was conducted to identify the difference between expectations and perceptions.
One-way ANOVA is one-way analysis of variance, a technique only for numerical data, which is used to compare means of two or more groups drawn from the same population (Howell, 2002). A probability value (Sig) associated with the difference between the groups is provided to determine whether the difference is significant or not. If Sig is lower than .05, the null hypothesis should be rejected. It means no difference is observed. On the contrary, if it is higher than .05, it fails to reject the null hypothesis and the correspondent alternative hypothesis should be rejected. Hypothesis 5 and 6 in this research paper are alternative hypothesis, which means if Sig value is higher than .05, they are to be rejected.
In addition, a statistic named F ratio is produced. In an ANOVA, the F-ratio, similar to Sig value, is the statistic used to test the hypothesis, whether the means are significantly different from one another or not and if the difference is due to chance (Stockburger, 1998).
4.4.5 Multiple regression.
Multiple regression analyzes the relationships between one metric dependent variable and several metric independent variables. The objective of this technique is to predict the changes in the dependent variable in response to the changes in the independent variables (Hair et al., 2008). It enables the researchers to determine how much the variation in the dependent variable can be explained by the independent variables and indicates which determinants are most important and which are insignificant (Niekerk & Hendrik, 1996). It also makes it possible to determine each determinant‟s contribution to the linear correlation after its links with other variables have been discounted (Oppenheim, 1992, p. 26). To determine the relative importance of different independent variable, R square is utilized as a reference index.
R2 is the square of correlation coefficients which ranges from 0 to 1. If Rsquare is equal to 0, it means there is no linear relationship. The closer R square is to 1, the stronger the relationship is.
The purpose of applying multiple regression is to test Hypothesis 1 and 3, so that to identify the relative importance of the independent variables (the five dimensions of service quality) and determine their individual contribution to the relationship with the overall service quality.
Furthermore, Hypothesis 7 is tested with regression in the assessment of the relationship between perceived changes and overall perception of the service quality.
Chapter Ⅴ Results, Analysis and Discussion
In this chapter, the results based on the online survey will be presented and analyzed according to the order of the hypotheses proposed. Key statistical signals from the survey are presented and supported by observation and interview findings. The aim is to conduct statistical analysis on the collected data, test the hypotheses, and find the evidence for rejection or confirmation of the tested hypotheses.
5.1 Sample Description
The total number of data list contains email addresses of 500 respondents, while 3 of them were not reachable because of technical problems such as email system error. After sending out reminders, 93 respondents conducted the pre-event survey of exhibitors‟ expectations on TUR 2011 and 96 respondents conducted the post-event survey of exhibitors‟ perceptions on TUR 2011. The response rates are 18.71% and 19.32% respectively. On the basis of Frequencies Analysis, the geographical overview about the respondents is concluded as shown in Table 1 below.
Table 1.
Geographical Information of the Respondents
Expectations Perceptions
Frequency Valid
Percent Frequency Valid
Percent
Sweden 58 62.4 52 54.2
Non-Sweden 35 37.6 44 45.8
Total 93 100.0 96 100.0
Repeat Participants 76 81.7 72 75
First-time Participants 17 18.3 24 25
Total 93 100.0 96 100.0
5.2 Reliability of the Measurement Scale
When using Likert-type scales it is imperative to calculate and report Cronbach‟s alpha coefficient for internal consistency reliability for any scales or subscales one may be using
(Gliem & Gliem, 2003). The Cronbach‟s alphas of all items lie well above the threshold of .70. The Cronbach‟s alpha for total scale reliability of expectation is .882 and of perception is .963 (refer to Appendix 6), which exhibits the internal consistency of the survey instrument and the SERVQUAL scale demonstrates a high degree of reliability for the measurement of service quality in TUR 2011.
5.3 Sample Distribution
Before regression and one-way ANOVA analysis is conducted, it is important that normal distribution of the data is ensured. Sample distribution is tested by Descriptive Statistics in SPSS for all the items under expectation survey and perception survey. It shows that almost all of the data is normally distributed with few exceptions under expectation survey.
According to Central Limit Theorem, if a random sample is selected from a population, then, when the sample size is sufficiently large (n>=30) the sampling distribution of the mean tends to approximate the normal distribution (Arjomand, 2002). The sample size of the case in analysis is considered large enough to cancel out the detrimental effects of nonnormality (Hair et al., 2008).
5.4 RATER’s Influences on Overall Expectations
To identify the five dimensions influence on the overall expectations, a new variable labeled as “overall expectation” is computed by calculating the mean of all the twenty items.
Moreover, in order to view the variables in a structural way, the analysis is conducted at two levels. First, summated scales are used to group the variables according to the five dimensions, so that they can be viewed for what they represent collectively in describing a particular dimension. Later, the variables‟ relationships with the overall expectation are examined at the most detailed level (individual variables themselves) to view their individual influences.
Pearson correlation shows the degree to which the five dimensions are correlated individually with the overall expectation. The results imply that Assurance and Empathy correlate stronger with Overall expectation than other dimensions do, and the Pearson correlations are .858 and .848 respectively.
