THE EFFECTS OF SERVICE RECOVERY SATISFACTION ON CUSTOMER LOYALTY AND FUTURE BEHAVIORAL INTENTIONS: AN EXPLORATORY STUDY IN THE LUXURY HOTEL INDUSTRY Except where reference is made to the work of others, the work described in this dissertation is my own or was done in the collaboration with my advisory committee. This dissertation does not include proprietary or classified information. __________________________________ Kristen A. Riscinto-Kozub Certificate of Approval: ________________________ __________________________ Susan Hubbard Martin O?Neill, Chair Professor Program Director Nutrition and Food Science Nutrition and Food Science ________________________ _________________________ James Wite Anthony Guarino Associate Professor Associate Professor Educational Foundations, Educational Foundations, Leadership, and Technology Leadership, and Technology _________________________ Joe F. Pittman Interim Dean Graduate School THE EFFECTS OF SERVICE RECOVERY SATISFACTION ON CUSTOMER LOYALTY AND FUTURE BEHAVIORAL INTENTIONS: AN EXPLORATORY STUDY IN THE LUXURY HOTEL INDUSTRY Kristen A. Riscinto-Kozub A Dissertation Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Auburn, Alabama May 10, 2008 iii THE EFFECTS OF SERVICE RECOVERY SATISFACTION ON CUSTOMER LOYALTY AND FUTURE BEHAVIORAL INTENTIONS: AN EXPLORATORY STUDY IN THE LUXURY HOTEL INDUSTRY Kristen A. Riscinto-Kozub Permission is granted to Auburn University to make copies of this dissertation at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. __________________________ Signature of Author _________________________ Date of Graduation iv VITA Kristen Ann Riscinto-Kozub, daughter of Nicholas and Anna Marie Riscinto, was born October 1, 1975, in Dubois, Pennsylvania. She graduated from Indiana University of Pennsylvania (IUP) with a Bachelor of Science in Hotel, Restaurant and Institutional Management in 1999. Over a twelve year period in the hospitality industry she was the assistant catering and banquet manager for the Indiana Country Club, a Service Manager for Red Lobster Restaurant, trainer for Olive Garden Restaurants and eventually became a Store Manager for Starbucks Coffee Company. She received a Master of Science in Hospitality and Tourism Management from Roosevelt University, Chicago. In 2003 she accepted a position as Program Coordinator of the Hotel, Restaurant and Tourism Management Program at Ogeechee College. While at Auburn Kristen had the honor of participating in the Epicurean Wine Tour and was selected to represent the College of Human Sciences at the Quality of Life Awards held at the United Nations, New York, New York, December, 2006. She was also inducted in to the Kappa Omicron Nu Honor Society. In accordance with Dr. Martin O?Neill, she was the first graduate student to conduct guest-oriented research with the West Pace Hotel Group; founded by Horst Schulze, Founder and former CEO of Ritz Carlton. Kristen has accepted a position as Assistant Professor in Hospitality Management at Widener University. She and her husband, Jeff, live in Drexel Hill with their two cats; Max and Monty and three dogs; Maddie, Foster, and Patrick. v DISSERTATION ABSTRACT THE EFFECTS OF SERVICE RECOVERY SATISFACTION ON CUSTOMER LOYALTY AND FUTURE BEHAVIORAL INTENTIONS: AN EXPLORATORY STUDY IN THE LUXURY HOTEL INDUSTRY Kristen A. Riscinto-Kozub Doctor of Philosophy, May 10, 2008 (M.S., Roosevelt University, 2005) (B.S., Indiana University of Pennsylvania, 1999) 216 typed pages Directed by Martin O?Neill This study investigated the quality of the complaint management process employed in five luxury hotels around the United States. Over the past two decades, complaint management research has focused on methods to evaluate service failures and identify and measure service recovery attributes. Although research has contributed to the evolution of supporting literature, there appears to be numerous opportunities for additional contributions. In 2000, Andreessen observed that service recovery research only focused on why, to whom, and how customers responded to dissatisfaction. Moreover, Ruyter and Wetzels (2000) revealed that very little research had examined the vi relationship between service recovery and service quality variables (i.e. on-going customer satisfaction, loyalty and behavioral intentions). Current research has centered on the relationship between service recovery satisfaction, loyalty and future behavioral intentions (Boshoff, 1997, 1999; Mattila, 2001; Boshoff & Staude, 2003; Matilla & Wirtz, 2004) with research methods primarily focused on hypothetical service scenarios; which measured customer expectations opposed to customer perception. Furthermore, evaluating hypothetical service scenarios has eliminated the role of post-failure and post-recovery emotion and how the two relate to on-going satisfaction, loyalty and future behavioral intentions (FBI). Therefore, the primary focus of this research was 1) to empirically test the performance of a pre-developed service recovery satisfaction measurement in the luxury hotel industry, 2) to explore the relationship between the customer?s degree of pre-failure loyalty with service recovery satisfaction, 3) to explore the relationship between post- failure emotion, post-recovery emotion on on-going satisfaction, loyalty and FBI and 4) to explore the relationship between on-going satisfaction and the customer?s post- recovery loyalty and FBI intentions. A total of 553 or 48% claimed to have experienced a service failure while staying at one of the selected hotel properties. Results of the factor analysis revealed a two factor structure as drivers of service recovery satisfaction. The factors support the concept that customer satisfaction is achieved through a combination of what is done and how well it is done. Results also revealed a unique relationship between pre-failure loyalty and emotion with on-going satisfaction, post-recovery loyalty and the customer?s FBI toward the organization. vii ACKNOWLEDGEMENTS First and foremost, I am truly grateful to my committee chair, Dr. Martin O?Neill. Thank you for your endless support and encouragement and for creating a standard that was challenging as well as rewarding. You have been a wonderful mentor. To Dr. Susan Hubbard, I am truly grateful for your guidance and support. To Dr. James Witte, I am thankful to have had the opportunity to work with you and Dr. Maria Witte. My sincere thanks to both of you for your guidance and direction. To Dr. Anthony Guarino, I am much appreciative of your statistical guidance and support. To my wonderful friend, JoEllen. Thank you for all of your support during this time. Your friendship has been an amazing blessing and you have enriched my life. To my extended family, Consetta and Chuck, for your love and encouragement. I will never forget how you helped me have the Auburn experience. I am grateful to my Grandmothers, Mary Ann and Rose; and to my Godparents, Aunt Jeannie & John K for their love and support. To my Parents, Nick and Anna Marie, for their endless love, helping me keep things in perspective and for always encouraging me to do better. To my ?kids? Maddie, Foster, Patrick, Max and Monty, for never leaving my side. To my husband Jeff, for your abundant love, laughter and patience. To my Grandfathers, Dominick and John; failure is never an option with the two of you watching over me. Last, but never least, I am most thankful to God, for sending these angels to enrich my life and for always helping me find my faith. viii Style manual or journal used: Publication Manual of the American Psychological Association, 5 th Edition. Computer software used: SPSS 15, SPSS 16, Amos 16 Windows 2000, and Microsoft Word 2000 ix TABLE OF CONTENTS Page LIST OF TABLES ............................................................................................. xi LIST OF FIGURES..........................................................................................xiii CHAPTER I INTRODUCTION ......................................................................................... 1 Overview ........................................................................................... 1 Statement of the Problem ................................................................... 1 Purpose of the Research ..................................................................... 4 Aims and Objectives .......................................................................... 5 Research Questions ............................................................................ 5 Limitations and Delimitations of the Study......................................... 6 Definition of Terms............................................................................ 7 CHAPTER II LITERATURE REVIEW............................................................................. 10 Services Marketing and Relationship Marketing............................... 10 Importance of Service ...................................................................... 13 Satisfaction ...................................................................................... 18 Quality ............................................................................................. 29 Loyalty............................................................................................. 44 Emotion ........................................................................................... 61 Service Failure ................................................................................. 72 Service Recovery ............................................................................. 76 Service Recovery and Justice Theories............................................. 77 Service Failure and Service Recovery............................................... 80 Service Recovery Defined ................................................................ 83 Measurements of Service Recovery.................................................. 85 RECOVSAT .................................................................................... 91 Antecedents of RECOVSAT ............................................................ 92 Hypotheses and Related Literature ................................................... 94 Summary ......................................................................................... 96 CHAPTER III METHODS.................................................................................................. 98 Introduction...................................................................................... 98 Research Considerations .................................................................. 98 Research Hypotheses...................................................................... 100 Theoretical Model.......................................................................... 104 x Methods Overview ......................................................................... 105 Research Setting............................................................................. 108 Research Sample ............................................................................ 111 Research Instrument....................................................................... 113 Measurement of Variables.............................................................. 115 Ethical Considerations.................................................................... 116 Summary........................................................................................ 116 CHAPTER IV RESULTS ................................................................................................. 117 Introduction.................................................................................... 117 Discussion of Returned Questionnaire ............................................ 117 Sample Demographic Characteristics ............................................. 118 Description of Loyalty and Future Behavioral Data........................ 121 Description of Emotion Data .......................................................... 126 Description of Service Recovery Data ............................................ 130 Evaluation of Validity and Reliability............................................. 132 Dimensionality of the Research Instrument .................................... 140 Dimensionality of Emotion Scales.................................................. 140 Dimensionality of Loyalty Scales ................................................... 142 Dimensionality of the Recovery Satisfaction Scale ......................... 144 Testing the Central Research Hypotheses ....................................... 149 Path Analysis of Central Research Hypotheses............................... 156 Summary........................................................................................ 158 CHAPTER V CONCLUSIONS, DISCUSSION, AND RECOMMENDATIONS................................................................ 160 Conclusions.................................................................................... 160 Performance of the Instrument........................................................ 172 Academic Implications................................................................... 175 Practitioner Implications ................................................................ 178 Future Research.............................................................................. 182 Conclusion ..................................................................................... 183 REFERENCES................................................................................................ 184 APPENDIX A .............................................................................................. 199 APPENDIX B .............................................................................................. 201 APPENDIX C .............................................................................................. 203 xi LIST OF TABLES Table Page 1. Hotel Participation ................................................................................. 118 2. Demographic Profile of Respondents ..................................................... 119 3. Service Failure Details ........................................................................... 120 4. Descriptive Statistics for Pre-Failure Loyalty and FBI............................ 122 5. Descriptive Statistics for Post-Recovery Loyalty and FBI ...................... 122 6. Paired Sample t Test for Loyalty Variables ............................................ 123 7. Pre-Failure and Post-Recovery Customer Loyalty.................................. 124 8. Pre-Failure and Post-Recovery Future Behavioral Intentions.................. 125 9. Descriptive Statistics for Post-Failure Emotion Data.............................. 127 10. Descriptive Statistics for Post-Recovery Emotion Data ......................... 127 11. Paired Sample t Test for Emotion Variables .......................................... 128 12. Emotion Intensity During Failure.......................................................... 129 13. Emotion Intensity After the Service Recovery Process.......................... 130 14. Descriptive Statistics for Service Recovery Data................................... 131 15. Correlation Matrix of Post-Recovery Loyalty and ................................. 137 Post-Recovery FBI 16. Internal Reliability Statistics for Measurement Scales ........................... 139 17. Factor Analysis for Emotion Scales....................................................... 142 18. Factor Analysis for Loyalty Scales........................................................ 143 19. Exploratory Factor Analysis/Service Recovery Satisfaction Attributes. 146 xii 20. Correlation Index for Recovery Satisfaction Components ..................... 151 And On-Going Satisfaction. 21. Correlation Index for Recovery Satisfaction Components ..................... 153 And Expressed Loyalty 22. Independent Samples t Test for Loyalty and Recovery Satisfaction....... 154 xiii LIST OF FIGURES Figure Page 1. Proposed Theoretical Model.................................................................. 105 2. Confirmatory Factor Analysis for Service Recovery.............................. 148 3. Modified Theoretical Model.................................................................. 150 4. Path Analysis of Proposed Relationship ................................................ 157 1 CHAPTER I INTRODUCTION OVERVIEW In 1997, Oliver stated that customer satisfaction must remain a central tenet of all relationship management efforts in the hospitality sector. Hospitality organizations typically strive to make customer satisfaction a focal point of their core values, visions and mission. However, in consideration of steady growth and expansion, global competition, and the influx of seasoned travelers, organizations face cumbersome obstacles in delivering quality service. It appears that quality service is the cornerstone to industry success; thus, those organizations that focus on quality become the flagship for on-going customer satisfaction, engendering loyalty and influencing future behavioral intentions. Shoemaker and Bowen (2003) referred to loyalty through quality service and on-going customer satisfaction as the new Holy Grail; that in which all service organizations should strive to achieve. PROBLEM STATEMENT In 1999, Andreassen observed that service recovery research over the past decade only focused on why, to whom, and how customers responded to dissatisfaction. Prior to that scrutiny Goodwin and Ross (1992) concluded that less attention was directed to corporate responses to the customers? voiced complaints and customers? subsequent attitudinal and behavioral changes were not effectively monitored. Furthermore, Conlon and Murray (1996) argued that most of the existing service recovery studies, at that time, 2 focused on the short-term impact of recovery efforts (i.e. compensation and quality of apology) and failed to explore the true drivers of service recovery satisfaction. Moreover, Ruyter and Wetzels (2000) revealed that very little research had examined the relationship between service recovery and service quality variables (i.e. on-going customer satisfaction, loyalty and behavioral intentions). Consequently, service organizations were not effectively identifying or evaluating the drivers of service recovery satisfaction within the service industry; therefore, the benefits of on-going satisfaction and customer retention were not truly achieved. Within the hospitality industry, Colgate and Norris (2001) suggested that a commitment to a continuous quality improvement process could only be achieved by tracking the number and severity of service failures. Thus, the researchers believed that a customer?s perception of service quality was directly linked to the number and severity of the failures he or she experienced. In addition, their research also showed that the organization?s quality of response and problem resolution was critical to the customer?s continuous satisfaction, loyalty and intentions to return or recommend. Although an extensive and thorough body of literature exists on service quality (Berry, Zeithaml, & Parasuraman, 1985; Parasuraman, Zeithaml, & Berry, 1985, 1988, 1991, 1994; Zeithaml, Berry, Parasuraman, 1993; Zeithaml, Parasuraman & Berry, 1990; Cronin and Taylor, 1992 & 1994; O?Neill, 1992, 2001; Kandampully, 2002; Gronroos, 2001), satisfaction (Oliver, 1980,1993 & 1997; Wirtz & Bateson, 1999;) loyalty (Oliver, 1997; Edvardsson, Johnson, Gustafsson & Strandvik, 2000; McMullan & Gilmore, 2003; McMullan, 2005) emotion (Izard, 1992; Westbrook & Oliver, 1991; Mattila & Wirtz, 2000; Smith & Bolton, 2002; Matilla, 2002) and service recovery (Boshoff, 1997, 1999; 3 Ruyter & Weltz, 2000; Strauss, 2002; Mattila, & Wirtz, 2000; Mattila, 2001; Mattila, & Patterson, 2004; Mueller, Palmer, Mack, & McMullan, 2003), there appears to be areas of opportunity for additional research and new contributions. For example, as the succeeding pages will support, a need has emerged for a systematic approach to track and measure service failures and the overall performance of the service recovery process within the various segments and organizations of the hospitality industry. In addition, to the researcher?s knowledge, a scale has yet to be developed to thoroughly measure the degree of on-going satisfaction, loyalty, emotion and future behavioral intentions in the luxury segment of the hotel industry. While models and scales have contributed to the body of literature (Boshoff, 1997 & 1999; Mattila, 2001; Craighead, Karwan & Miller, 2004; Kau and Loh, 2006), few have been tested empirically throughout various segments of the service industry, let alone in the hospitality industry. For example, Boshoff?s RECOVSAT has received a considerable amount of attention since its development in 1999; however, Strauss (2002) suggested that although Boshoff?s RECOVSAT performed well in a hypothetical scenario it had yet to be tested empirically. Since that time, RECOVSAT has been applied in an empirical nature in the banking and finance industry; however, to the researcher?s knowledge the measure has not been applied empirically to other service industries. Furthermore, to the researcher?s knowledge, the relationship between 1) on-going satisfaction, 2) post-recovery emotion with on-going satisfaction and 3) the customer?s degree of pre-failure and post-recovery loyalty and FBI has not been explored in an empirical sense within luxury hotel properties. In addition to the above, a service recovery model has yet to be developed and confirmed across the luxury segment of the hotel industry. 4 Therefore, the primary focus of this research was to evaluate the effectiveness of the compliant management systems within five selected luxury hotels and to 1) empirically test the performance of RECOVSAT in the luxury hotel segment, 2) explore the relationship between the customer?s degree of pre-failure loyalty and their on-going satisfaction with the organization, 3) to explore the relationship between post-recovery emotion and on-going satisfaction and 4) to explore the relationship between on-going satisfaction with the customer?s loyalty and FBI. PURPOSE OF THE STUDY Although hospitality organizations strive to provide excellent service in the first place, the characteristic nature of the environment makes imperfections unavoidable. With that being said, things can and frequently do go wrong, leading to the need for a systematic recovery effort aimed at 1) resolving customer complaints, 2) influencing on- going satisfaction, and 3) exploring the relationship between on-going satisfaction, emotion, loyalty and FBI. While organizations should strive to provide customers with no reason to complain in the first place, when imperfection strikes, a well-crafted recovery process can turn a weakness into a business opportunity (Barsky, 1996). Therefore, the purpose of this study was to 1) identify attributes of service recovery satisfaction in luxury hotels by empirically testing the performance of RECOVSAT, 2) explore the relationship between the customer?s degree of pre-failure loyalty and their on- going satisfaction with the organization, and 3) to explore the relationship between post- recovery emotion with on-going satisfaction. In addition, this study will also explore the relationship between on-going satisfaction with the customer?s loyalty intentions and FBI. 5 AIMS AND OBJECTIVES Many hospitality organizations now consider the development of a complaint management system to be a vital component to ensure service recovery satisfaction. However, of all the efforts being made to deliver quality service, satisfy customers, engender loyalty and influence FBI, quality service and defusing service problems are still on-going issues in hospitality. Kandampully (2002) stated ?Of all the challenges facing hospitality establishments today- including intense competition, globalization and technological innovation- the single most pervasive and pressing challenge is the ever- increasing demand of customers for service quality? (p. xi). Against this background, the aims and objectives of this study was to evaluate the quality of the service recovery process and the relationship, from the customer?s perspective, with long-term loyalty and behavioral intentions. The intent of this study was to explore 1) the relationship between the organization?s service recovery efforts and the customer?s degree of on-going satisfaction, 2) the relationship between the customer?s degree of pre-failure loyalty, on-going satisfaction and post-recovery loyalty, 3) the relationship between the customer?s post-recovery emotion, loyalty and behavioral intentions and 4) the psychometric performance of Boshoff?s (1999) service recovery satisfaction measure, RECOVSAT, from a selected group of luxury hotels throughout the United States. RESEARCH QUESTIONS The primary questions were: How is the recovery satisfaction construct best defined in the context of the luxury hotel sector? What, if any, is the relationship between 6 an organization?s service recovery efforts; satisfaction and loyalty? And what, if any, role does perceived emotion and loyalty play as a mediator in the recovery process? LIMITATIONS AND DELIMITATIONS Although every attempt was made to eliminate possible limitations, this study does have several limitations. The following section is intended to address some of the limitations surrounding this study so they may be eliminated in future research. One of the primary limitations for this study was the amount of time between the actual service recovery experience and the administration of the scale for evaluation. The sample population consisted of customers who stayed at the selected hotel properties within the 12 months prior to the survey distribution. Unless the service recovery efforts were to an extreme positive or negative, a limitation of the study is certainly the customer?s ability to recall the details surrounding the service provider?s efforts. Another limitation for this study was also related to time. Participants from the sample population were asked to evaluate the type and degree of emotion felt at the time of the service failure as well as the type and degree of emotion felt directly after the service recovery process. Due to the amount of time between the experience and the survey, the accuracy of the emotions involved during the failure and directly after the recovery could be called into question. Delimitations to the study were also assessed. Delimitations are areas of possible emphasis or significance that were not included in this research. Considering the vast scope of the issues surrounding service recovery, there are many delimitations to the study. 7 One of the most significant delimitations was the role of the employee in the service recovery process. Employees play a pivotal role in on-going customer satisfaction; thus, in order to truly gauge the effectiveness of the service organization?s recovery efforts, it would be best to explore employee attitudes and perceptions toward on-going customer satisfaction, complaint handling and service recovery. The hotel organizations selected for this study preferred to first and foremost explore service recovery from the customer?s perspective. Another delimitation to the study was the degree of dissonance the customer may have felt prior to the time of the service failure. In an effort to fully dissect the effectiveness of the organization?s service recovery efforts, it may be of value to understand the customer?s attitude or emotion prior to entering into the service encounter. This element was not included in the study due to the amount of time between the actual service failure and the evaluation. If this indicator of service recovery satisfaction were to be evaluated, the service recovery satisfaction measurement should be administered closer to the actual event. DEFINITION OF TERMS Attitudinal Loyalty: The point in which the customer develops a favorable or unfavorable attitude toward the service or provider. This leads to the customer?s intention to return or recommend (Dick & Basu, 1994). Behavioral Loyalty: When a customer?s intentions are converted into actions; thus, the customer?s motivated intentions in the previous loyalty state has transformed or progressed to the point where the customer is ready to act upon their intentions (Kahl & Beckmann, 1985). Consumption Emotion: The combination of both positive and negative emotions with the service encounter. This plays a significant role in the customer?s degree of satisfaction, loyalty and future behavioral intention (Allen, Machleit & Kleine, 1992; Oliver, 1993; Richins, 1997; Barsky & Nash, 2002). 8 Disconfirmation Theory: Based on cognitive dissonance and the way people respond to their level of dissonance (Festinger, 1957). Distributive Justice: The allocation of benefits and costs; which relates to the degree of compensation associated with recovering the failed service offering (Oliver, 1997). Emotion: ?The arousal and various forms of affect and cognitive interpretations of affect that may be given a single description. One of the most predominant problems in defining emotion is to understand where emotion stops and where cognition begins. The greater the amount of cognitive interpretation required, the more cognitive the emotion becomes? (Oliver, 1997, p.294). Heterogeneity: When the customer?s overall evaluation of a service encounter does not rest solely on the processing of tangible attributes or the intangible elements from the service provider but instead on a combination of the above, paired with the customer?s mood, emotions and attitudes (Zeithaml, Parasuraman & Berry, 1990; Mantel & Kardes, 1999). Inseparability: When customers are physically involved and become a vital part of the service delivery process. Service depends on the provider?s performance as well as the customer?s participation (Kelly, Donnelly, & Skinner, 1990). Intangibility: The interpersonal elements between the people of the service organization and the customer that differentiates service and goods (Zeithaml, Parasuraman & Berry, 1990). Interactional Justice: The degree to which people are treated (Oliver, 1997). Loyalty: ?A deeply held commitment to rebuy or repartonize a preferred product or service consistently in the future, thereby causing repetitive same-brand or same-brand set purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior? (Oliver, 1997, p. 392). Procedural Justice: Resolving conflicts by use of a formal approach in order to be consistent and maintain the customer relationship- even if the outcome is unsatisfactory to one party (Wildes, 2005). Perishability: Services are time sensitive. Service providers have a small, distinct window of opportunity to delivery service excellence and meet the customer?s expectations (Berry, Zeithaml, & Parasuraman, 1985). Quality: Quality can be united into one concept of consumer-perceived quality, where quality can be defined only by customers and occurs where an organization supplies goods and services to a specification that satisfies their needs (O?Neill, 2001). 9 Relationship Marketing: The fundamental feature that differentiates services marketing from goods marketing is the intangible nature of the service industry (Berry, Zeithaml, & Parasuraman, 1985). Satisfaction: ?The summary psychological state resulting when the emotion surrounding disconfirmation expectations is coupled with the consumer?s prior feelings about the consumption experience?. ?A judgment that a product or service feature, or the product or service itself, provided (or is providing) a pleasurable level of consumption-related fulfillment, including levels of under-or-over fulfillment? (Oliver, 1981, p.13). Service: An interpersonal relationship between the organization and the customer (Price, Arnould & Tierney, 1995). A critical moment of truth that involves either satisfying or dissatisfying the customer (Albrecht & Zemke, 1995). Service Failure: Negative encounter occurs and the customer realizes that the service has not met their expectations (Zeithaml, Berry, & Parasuraman, 1993). Service Recovery: An organization?s response to poor quality service (Gronroos, 1988). Zone of Tolerance: The level of service satisfying or dissatisfying service that each individual customer will accept in each service scenario (Zeithaml & Bitner, 2000). 10 CHAPTER II LITERATURE REVIEW INTRODUCTION The purpose of this chapter is to review the literature that supports the constructs outlined in this study. To begin, the importance of relationship marketing to the service industry as well as the hospitality industry will be evaluated. The second section of the literature review will contain a detailed analysis of satisfaction and dissatisfaction and the direct relationship satisfaction has with service quality, engendering loyalty and influencing future behavioral intentions. In addition, emotion will be defined and various measures will be analyzed to support the important contribution emotion makes to this study. Finally, this chapter will conclude with a discussion of service failure and complaint management and provide an extensive examination of the research that surrounds service recovery. SERVICES MARKETING Over the past decade the United States economy has transformed from a manufacturing-based economy to a service-based economy. Along with this transformation the demand to understand and dissect the importance of service has emerged. To truly appreciate the complex nature of the service industry, related research has revolved around service quality (Parasuraman, Zeithaml & Berry, 1985; Gronroos, 1988; Cronin & Taylor, 1992; O?Neill, 1992; Oliver, 1997), satisfaction (Oliver, 1993 & 1997; Wirtz & Bateson, 1999; Zeithaml & Bitner, 2000), loyalty (Dick & Basu, 1994; 11 Oliver, 1997; Bowen & Shoemaker, 1998; Reichheld & Sasser, 1990; Heskett, Sasser & Schlesinger, 1994; McMullan & Gilmore, 2003; McMullan, 2005) and complaint management systems (Boshoff, 1997 & 1999; Mattila, 2001; Boshoff & Staude, 2003; Craighead, Karwan, & Miller, 2004; Mattila & Patterson, 2004). Although an extensive body of research exists today on the constructs surrounding services marketing, the expansion and evolution of the service industry has revealed a gap in the literature regarding customer service and related issues. Services marketing literature suggests that a customer orientation puts the customer at the core of an organization?s purpose and activity; therefore, looking after the customer and maximizing satisfaction allows the firm to remain competitive and profitable (McMullan & Gilmore, 2005). In order for a service provider to make the customer the core of the organization, the provider must first have a detailed understanding of the customer?s needs and expectations. As part of the service industry?s approach to rapid growth and market domination, today many products and services are easily duplicated, making it quiet challenging for service providers to compete by offering unique yet basic services and products. In an effort to differentiate themselves from the competition, many service providers have begun to place additional emphasis on nurturing the intangible characteristic of the service industry; thus in turn, placing their efforts on developing relationships in an effort to engender loyalty and retain customers opposed to continuously marketing to obtain new customers. RELATIONSHIP MARKETING The fundamental feature that differentiates services marketing from goods marketing is the intangible nature of the service industry (Zeithaml, Parasuraman & 12 Berry, 1990). The intangible characteristics of the service industry can simply be defined by the uniqueness and quality that the people bring to the service transaction. Therefore, the core of relationship marketing is developing and maintaining relations between the employees of the service organization and the people in the micro-environment, i.e. the customers (Ravald & Gronroos, 1996). The concept of relationship marketing started to emerge in research in the mid- 1980s, when the notion of developing a long-term, mutually satisfying relationship between customers and service providers was realized to be an asset (Weber, 1992). Over the past 20 years, relationship marketing has become an increasingly popular approach to services marketing and engendering loyalty; especially in the hospitality industry. In fact, it appears that many service organizations strive to infiltrate relationship marketing into the culture of the organization through their mission and vision. Both professional journals and trade publications recommend various concepts and ways to approach relationship marketing; however, the most agreed upon concepts to successful relationship marketing are that the service provider must 1) make the customer?s needs and expectations the core of their philosophy and 2) form an emotional connection based on trust and commitment between the service provider and the customer. To support the above, McKenna (1991) as well as Shoemaker and Bowen (2003) and Shoemaker and Lewis (1999) revealed that a in order to have a true bond between the customer and the organization, that is built on trust and commitment, the customer must be integrated into the design of the service or product. To enhance that theory and take it one step further, Claycomb and Martin (2002) demonstrated that in 13 order for relationship marketing to be effective, organizations must strengthen ties with existing customers in an effort to enhance satisfaction and serve the customer. Hence, it must become the foundation of the organization?s business philosophy. Gittell?s (2002) research in the health care industry revealed that a relationship between the service provider and the customer had a direct connection with customer satisfaction and loyalty. His study also showed that the stronger the relationship between the service provider and the customer, the more likely the customer was to revisit and recommend the organization. In regards to the hospitality industry, since the early 1990s, hotels, restaurants, airlines and cruiselines has focused much of their marketing attention and budget on building customer loyalty through relationship marketing. However, according to Bowen and Shoemaker (1998), the hospitality industry?s approach to relationship marketing has mainly focused on transactional tactics such as frequent user programs and free amenities and services; opposed to 1) building relationships, 2) creating an emotion connection with the customer, and 3) engendering loyalty in an effort encourage the customer to revisit and/or recommend. Although research has found that customers respond favorably to these monetary relationship tactics, their actions are not truly driven by an established, emotional connection with the service provider. Therefore, in the event of a service failure and the customer?s dissatisfaction, the customer?s loyalty is based on a pecuniary relationship and switching behavior is more probable. THE IMPORTANCE OF SERVICE When it comes to providing service, understanding customer expectations and the customer?s perception of the service 14 encounter is a vital component to delivering superior service. Delivering superior service, especially in the hospitality industry creates a myriad of opportunities for the service organization to surpass the competitive and become a recognized leader in the service industry. It only stands to rationalize that the concept of the service encounter directly affects satisfaction, loyalty and future behavioral intensions; which in turn, has a direct affect on the organization?s success and financial stability. In the section that follows, a detailed analysis of the service industry will be presented in the form of 1) the widely accepted definitions of service, 2) the service industry?s economic impact to the United States economy, and 3) a discussion surrounding the service industry?s distinguishing characteristics. SERVICE DEFINED Researchers have explored and clarified the definition and concept of service since the early 1960s. Although a large body of service research exists acknowledging countless definitions and theories that surround service, some of the most widely accepted and citied definitions in more recent literature are: 1) a service encounter is an interpersonal relationship between the organization and the customer (Price, Arnould & Tierney, 1995) and 2) it is a critical moment of truth that involves either satisfying or dissatisfying the customer (Albrecht & Zemke, 1995). In 1985, Czepiel, Soloman, Suprenant and Gutman defined service as an encounter which involves any transaction or exchange of products or services between the customer and a service organization. This researcher?s definition of service is the moment of truth when employee training and expertise is truly tested in an effort to meet or exceed the customer?s expectations of the tangible and intangible components of the service encounter. 15 The United States economy has shifted from an economy dominated by the manufacturing of tangible goods to an economy reliant upon the characteristics of the service industry. In 2006, the Bureau of Labor Statistics released an analysis of the service industry which stated that the service sector of the United States economy, at that time, accounted for more than 75 percent of the United States Gross Domestic Product and approximately 80 percent of all jobs in the United States. However, Namasivayam & Hinkin (2003) warned that although the service industry has become a significant part of the United States economy, customers are still experiencing dissatisfaction with the quality of the service encounter received. Research on satisfaction and quality will be thoroughly discussed in an upcoming section; however, the research presented in this next section will demonstrate that customer expectations and demands on the service provider are steadily increasing. On the contrary, the service provider continuously falls short in meeting and/or exceeding customer expectations. UNIQUE CHARACTERISTICS OF THE SERVICE INDUSTRY In consideration of the increasingly competitive nature of the service industry, especially the hospitality industry, service professionals must concern themselves with a minimal increase in the market share in addition to maintaining their existing customer base (O?Neill, 2001). As the service industry continues to expand, customers are constantly presented with new opportunities to find a service provider that is capable of fulfilling their demands and meeting their expectations. Lovelock (2001) believed that the pressure placed on service organizations to achieve service excellence is truly justified. In his research, Lovelock (2001) pointed out that 1) today, customers are more demanding of the products and services they buy, 2) the development of sophisticated 16 technology has made it convenient for all service organizations, large and small, to offer personalized services that are highly valued by the customer, and 3) in an increasingly competitive and international marketplace, providing a quality service encounter is seen as having the advantage over the competition. In addition to the characteristics stated above, service has also been characterized in four dimensions. One of the four, intangibility, has been discussed and defined in a previous section; however, in addition to further discussion of intangibility, this section will also focus on the distinguishing characteristics of inseparability, heterogeneity and perishability. Intangibility once again is a fundamental characteristic of service. As stated earlier, Zeithaml, Parasuraman and Berry (1990) claimed that intangibility is the fundamental factor that differentiates service and goods. In consideration of the tangible objects, i.e. products or goods, involved in a service encounter, a customer can use physical attributes such as sight, touch, sound and taste, to evaluate the quality and to determine their degree of satisfaction with the product or goods. However, a customer is unable to utilize their physical attributes to judge the quality and their satisfaction with the intangible aspect to the service encounter (Berry, 1995). Consequently, the intangible aspect to the service is dependent upon the employee?s ability to interact with the customer and satisfy the customer with the personalization and emotional element that makes the service encounter exceptional. In reference to the inseparability characteristic, in most service encounters, the provider is present when the service is being delivered and consumed. Therefore, customers are often physically involved and become a vital part of the service delivery process. Thus, the quality of the service depends on the provider?s performance as well 17 as the customer?s participation (Kelly, Donnelly, & Skinner, 1990), and the delivery of the service becomes a human performance that the customer is inseparable from (Zeithaml, Parasuraman, & Berry 1990). In service, the evaluation of past service encounters tends to influence the customer?s attitude and intentions when preparing for an upcoming service encounter. Zeithaml, Parasuraman & Berry (1995) refered to heterogeneity as a remarkable phenomenon. The customer?s overall evaluation of a service encounter does not rest solely on the processing of tangible attributes or the intangible elements from the service provider but instead on a combination of the above, paired with the customer?s mood, emotions and attitudes (Mantel & Kardes, 1999). The challenge to heterogeneity is the fact that many service providers strive to engender loyalty by from their existing customer base; therefore, service organizations tend to confuse customer satisfaction with customer loyalty. However, in order to engender loyalty, a customer must be satisfied with the service organization in good times and challenging times. Heskket, Jones, Loveman, Sasser and Schilger (1994) found that relying on the customer?s satisfaction with a specific service transaction does not necessarily mean that the customer will be satisfied enough to return or recommend the service. In fact, their research on the Service Profit Chain revealed that the weakest link in the chain was between satisfaction and loyalty. Thus, suggesting that a satisfied customer does not necessarily develop into a loyal customer. The final characteristic of the service industry is perishability. Simply put, perishability means that the services are time sensitive and service providers have a small, distinct window of opportunity to delivery service excellence and meet the 18 customer?s expectations. Once again, Zeithaml, Parasuraman and Berry?s (1985) definition of perishability stating that, unlike products that can be stored or reinvented for future orders, services can not be returned, recaptured or recreated; only repeated. Industry projections suggest that the service industry will continue to dominate the United States workforce as well as the economy. In consideration of the above, service providers must continuously strive to create a balance in their approach to service excellence. A balance of the characteristics of the service industry seems to be the most reliable approach to delivering quality service, creating satisfied customers and maintaining a competitive advantage. SATISFACTION Satisfaction plays a key role in the service industry and especially in the hospitality industry (Edvardsson, Johnson, Gustafsson & Strandvik, 2000). As the service industry has evolved, researchers have made great strides to define and understand satisfaction from the consumer?s perspective. The emphasis to comprehend what truly creates satisfied customers has lead to an ever increasing body of literature surrounding satisfaction, how service providers create satisfied customers and the effects that satisfaction has on businesses today (Oliver, 1997; Barsky & Nash, 2002). McKenna (1991) suggested that in order to achieve satisfied customers, organizations must forget about market surveys, advertisements, and promotions and focus on developing the right infrastructure that offers the right products and services that meet the customer?s expectations. To a certain degree, relationships may keep the customer loyal but the customer must be fulfilled on other levels too; thus, the customer must be truly satisfied with the quality and value of the services and products. For 19 example, if a customer chooses to stay at a luxury hotel, he or she must receive accommodations and services that correspond with other luxury properties and find value in the price paid to truly be satisfied. Scholars have contributed volumes of research to support the importance of satisfaction in the service industry. Blodgett, Granbois, and Walters (1993) revealed that satisfied customers are far more valuable to a service organization when compared to dissatisfied customers; thus, the organization?s primary objective should be to satisfy the customer. Through the years, research in professional, as well as trade publications, have justified the importance of satisfaction. In response to the vast amounts of research, service providers have slowly come to realize that satisfied customers are truly assets to the organization. On the contrary, in 1996, Luchars and Hinkin acknowledged that managers typically choose to ignore satisfaction and service quality research and suggested techniques since direct operating costs could not be accurately measured or associated with specific instances of good or bad service. This in turn, led justification for a manager to rely on data collected from the tangible or quantifiable measurements in order to make managerial decisions. However, in consideration of the ramifications, when customers are satisfied, they are likely to participate in favorable behavior toward the provider, i.e. positive word of mouth advertising, willingness to recommend, revisit intentions, decreased price sensitivity over time and their willingness to participate in research to help the organization revolutionize. It is no surprise that satisfied customers are truly invaluable to an organization. However, at the opposite end of the spectrum, service providers are still faced with the negative consequences from the dissatisfied customer. One of the most significant 20 problems with dissatisfied customers is that some customers choose not to complain in order to avoid confrontation or to save the employee from being reprimanded. Nevertheless, with each dissatisfied customer who chooses not to complain, the service provider losses the opportunity to rectify the situation and return the customer to a satisfied state. In addition to possibly loosing the customer, the service provider also misses the opportunity to fulfill a commitment to continuous quality improvement. The quality and approach to service recovery will be discussed later in this chapter; however, for now, it is important to recognize that dissatisfied customers can be far more damaging to an organization than satisfied customers can be beneficial. SATISFACTION DEFINED Throughout the 1980s and the1990s, Oliver?s research and definitions of satisfaction revolved primarily around transaction-specific satisfaction. As one of the most widely accepted definitions, Oliver (1981) defined satisfaction as ?an evaluation of the surprise inherent in a product acquisition and/or consumption experience. Thus, satisfaction is the summary psychological state resulting when the emotion surrounding disconfirmation expectations is coupled with the consumer?s prior feelings about the consumption experience? (p.27). In 1997, Oliver further clarified satisfaction when he stated ?satisfaction is the consumer?s fulfillment response. It is a judgment that a product or service feature, or the product or service itself, provided (or is providing) a pleasurable level of consumption-related fulfillment, including levels of under-or-over fulfillment? (p.13). Past research has also presented various theories, models and definitions to explain and justify the importance of satisfaction. Opposed to a transaction-specific 21 approach to satisfaction, Fornell (1992) described satisfaction as the cumulative overall evaluation of a customer?s purchase and consumption experience based on numerous interactions between the customer and the service provider. To support the importance of cumulative satisfaction, Olsen and Johnson (2003) found that customer loyalty and repurchase intentions are based on a broader consumption history and are able to explain more variation in loyalty; therefore, service providers must focus on satisfaction over the course of the service provider-customer relationship. SATISFACTION AND DISSATISFACTION Edvardsson, Johnson, Gustafsson and Strandvik (2000) pointed out that as the hospitality industry continues to grow, the need for having the competitive advantage becomes more important. A dissatisfied customer can be one of the biggest liabilities to a service organization. In 1993, Blodgett et al. reported that a dissatisfied customer told an average of nine people about their negative experience. As a result, his research also found that service organizations lost an average of twelve percent of their annual customer volume each year due to poor quality service. Unfortunately, a large majority of dissatisfied customers choose not to complain and never seek redress for the failed service attempt; thus, leaving the organization dissatisfied with intentions to never return and to share their negative experience with others. Just as satisfied customers reward the organization with loyalty and positive word-of-mouth advertisement the chain reaction of dissatisfaction can also create a ripple effect. Overall, satisfaction or dissatisfaction is typically determined by the combination of various fundamentals. First, in order for the guest to decide if he or she is truly 22 satisfied with the product or service, the guest must process the activities surrounding the service encounter as well as the quality of the service or product itself. In addition, satisfaction typically is collective observations over a period of events. Thus, to determine a customer?s satisfied or dissatisfied state a customer should experience a product or service over a period of time and numerous service encounters. DISCONFIRMATION THOERY To summarize the various definitions of satisfaction; customer satisfaction starts at the moment consumption begins and carries through during the entire service encounter. In addition, satisfaction is also the end result of post-consumption evaluations and judgments of a product or service over the service provider-customer relationship. Martin (2007) also indicated that although a customer could be satisfied at the end of a service encounter, the customer could still be unhappy with the end result, alluding to the fact that expectations play a large role in the evaluation of customer satisfaction. The determination of what truly satisfies a customer is a fundamental element to successful service. For some service providers it appears that the organization has a natural ability to satisfy customers. For some, lines of communication, pleasant gestures and perfecting the intangible elements appears to be less cumbersome. For a service provider capable of delivering this level of service, evaluating a customer?s degree of satisfaction may seem to a valued practice of continuous quality improvement. Yuksel and Rimmington (1998) justifed that service providers who evaluate a customer?s satisfaction level and apply that knowledge to improve on service delivery are typically positioned to have a greater advantage over the competition. In a more detailed explanation, they continue by stating that the benefits of customer satisfaction evaluation 23 is central when identifying product and service differentiation, increasing customer retention, creating intentions to recommend, and encouraging positive word of mouth communication (Yuksel & Rimmington, 1998). While early research considered the satisfaction construct a form of attitude, considerable research in the late 1980s revealed that satisfaction is notably more complex and that measurement issues are numerous (Cooper, Cooper & Duhan, 1989). Researchers have developed various competing models and theories of measurement approaches for satisfaction (Yuksel & Rimmington, 1998); however, the most commonly used representation of customer satisfaction is the disconfirmation approach (Ramaswamy, 1996). Disconfirmation theory has shown that satisfaction is directly related to the variation between a customer?s pre-purchase expectations and their post-purchase perceptions of the actual service encounter. Festinger (1957) suggested that Disconfirmation Theory is based on cognitive dissonance and the way people respond to their level of dissonance. Festinger (1957) explained that dissonance is a psychological state where a person feels uncomfortable with his or her decision or situation. When dissonance is present in a service scenario, the service provider has the responsibility to reduce the level of dissonance and reassure the customer that his or her decision was the correct decision; and deliver a satisfactory experience. The researcher also explained that dissonance tends to be a motivating factor; thus, new events or new information may become apparent to a person; this could possibly defuse or magnify the level of dissonance. This process aids a customer?s evaluation of the service scenario, which in turn, determines the customer?s level of satisfaction or dissatisfaction. 24 Oliver?s (1997) perspective on disconfirmation theory is the extent of satisfaction or dissatisfaction that a customer has with a particular service encounter. Thus, it is the process the customer goes through to determine the disconfirmation between the customer?s expectations of performance and the actual perceived performance of the service encounter. Whether the product or service meets the customer?s expectations or fails to meet expectations will determine the customer?s level of satisfaction or dissatisfaction. Differences between the expected service and the actual service experience are referred to as disconfirmation. When comparing expectations with performance, three outcomes are possible; 1) if the service experience exceeds the customer?s expectations, the customer will experience positive disconfirmation; which in turn, high levels of satisfaction will be the end result, 2) if the service experience simply meets the customer?s expectations and the service provider does not have a unique or over-zealous approach to service; zero disconfirmation occurs, and 3) if the service experience fails to meet the customer?s expectations, the customer will experience negative disconfirmation; which in turn, some degree of dissatisfaction will be the end result. The various components that correspond with disconfirmation such as the degree of disconfirmation, a customer?s prior experience, tolerance level and the attributes that surround service failures and recovery efforts will be evaluated later in this chapter; however, it is important to recognize that disconfirmation will differ from customer to customer and in the severity that surrounds the disconfirmation. For example, a customer dissatisfied over a poor quality cup of coffee will have a different degree of disconfirmation in comparison to a customer being dissatisfied over the poor quality or cleanliness of a luxury hotel room. 25 As seen in Figure 1, Wirtz and Bateson (1999) demonstrated a model to signify the path of disconfirmation and the zones of tolerance that accompany disconfirmation. As stated earlier, the disconfirmation model is one of the most recognized representations of customer satisfaction in literature today (Ramsaswany, 1996). Wirtz and Bateson (1999) explain that ?almost all customer satisfaction models are based on some sort of comparison process, and the most widely used conceptualization, the disconfirmation-of- expectations model, which compares pre-consumption expectations with performance perceptions (p.83). Figure1. Disconfirmation-of-Expectations Model Source: (Wirtz & Bateson, 1999) In addition to the disconfirmation theory, a customer?s demand for quality also hinges greatly on the equity theory. Equity theory suggests that consumers evaluate service experiences by assessing the balance between what they receive and what they expect (Susskind, 2001). Since customer expectations play a significant role in service quality, Expected Quality Perceived Quality Disconfirmation Satisfaction 26 the disconfirmation theory and equity theory serves as key elements in measuring how customer needs and expectations are changing in response to the changing market. Customers enter into a service scenario, especially in the hospitality industry, with predetermined expectations and interests. In addition to their predetermined expectations, customers tend to evaluate a service organization based on various attributes such as the quality of food, ambiance, d?cor, and personal attention. In many situations, a customer tends to follow their own personal unwritten rules of service quality, thus, if expectations are not fulfilled, a guest may consider the service encounter to be lacking in quality; therefore, leaving dissatisfied. ZONES OF TOLERANCE The existing body of literature on satisfaction suggests that the customer?s expectations prior to the service encounter are necessary factors in determining their degree of satisfaction or dissatisfaction. The zones of tolerance are defined by Zeithmal and Bitner (2000) as the level of service that each individual customer will accept in each service scenario. For some customers, zones of tolerance may be minimal; demanding a more perfected approach to service. Other customers may have a larger zone of tolerance where he or she may be more tolerant to imperfections. Research suggested (Zeithmal & Bitner, 2000; Hoyer & MacInnis, 2001) that customers with demanding schedules or who are limited for time may have a minimal zone of tolerance compared to customers who have more flexibility or who are seeking a relaxed service experience. In addition, Hoyer and MacInnis (2001) cautioned that customers place more emphasis on certain service encounters than others. 27 The underlying message to satisfaction research is that satisfaction plays a vital role in the success of any service organization; especially in the hospitality industry. As most organizations will attest, customer satisfaction is the key to repeat customer visitation, customer loyalty, their willingness to recommend and the ability to obtain and retain quality employees. In respect to the above, organizations would be remiss to disregard the impact on the financial stability, additional resources, higher wages and, last but not least, a healthy bottom line. THE SERVICE PROFIT CHAIN In the early 1990 researchers explored the impact of service quality, customer and employee satisfaction and loyalty. As an end result the Service Profit Chain was created and the relationship between satisfaction, dissatisfaction and the organization?s success emerged. In 1994, researchers, Heskett, Jones, Loveman, Sasser and Schlessinger researched the potential relationships between service quality, customer loyalty, profitability, employee satisfaction and productivity. The end result of their research was the creation of the Service Profit Chain. Their justification for the creation of the model was as follows: Heskett, Jones, Loveman, Sasser and Schlessinger (1994) stated: ?The economics of service requires innovative measurement techniques. These techniques calibrate the impact of employee satisfaction, loyalty and productivity on the value of products and services delivered so that managers can build customer satisfaction and loyalty and assess the corresponding impact on profitability and growth. In fact, the lifetime value of a loyal customer can be 28 astronomical, especially when referrals are added to the economics of customer retention and repeat purchase of related products? (p.164). The Service Profit Chain, featured in Figure 2, demonstrates the relationship between profit and growth, customer satisfaction and loyalty, the value of services and goods delivered to the customer, employee satisfaction, loyalty and productivity and what is labeled as internal service quality signifying that employees have to deliver results to customers in every service encounter. Figure 2. The Service Profit Chain Source: (Heskett, Jones, Loveman, Sasser and Schlessinger, 1994, p.19) Based on their research, the strongest link in the Service Profit Chain was that between customer satisfaction and service value. Typically, customers tend to seek service quality processes that correspond with the price paid. If the customer feels that the value exceeds the price paid, the customer is, most often, satisfied with the service Internal Service Quality Employee Satisfaction Employee Productivity Employee Retention External Service Value Customer Satisfaction Revenue Growth Profit Customer Loyalty 29 encounter. From a different perspective, customer loyalty linked to customer satisfaction exhibited to be the least reliable link in the Service Profit Chain. Reichheld (1996) suggested that the services that satisfy customers may not always be the same services that engender loyalty to service organizations. Loyalty to the service organization will be covered later in this chapter; however, once again, it is important to draw attention to Heskett?s et al. (1994) findings. Their research revealed that short-term measures of the loyalty-satisfaction relationship could be disrupted by situations such as competitive price reductions, special trail offers, or even creative marketing schemes. This in turn, may entice customers to patronize the competitor and remove the current service provider from their radar, regardless of the level of satisfaction the customer has received in the past. Reichheld and Sasser (1990) tested relationships between market share and profitability in service organizations and found that service organizations with higher levels of customer loyalty also enjoyed higher profitability in a given industry, and loyal customers also became more profitable over time. In conclusion, research regarding the Service Profit Chain also revealed that employee productivity is linked to loyalty. Heskett et al. (1994) explains that traditional measures of the losses incurred by employee turnover concentrate only on the cost of recruiting, hiring, training and replacing employees. However, the focus of this study is to explore the significant loss in quality and customer satisfaction. QUALITY Service providers have a direct influence and control over the customer?s degree of satisfaction through the quality of the products and services provided. As global 30 competition and market saturation changes the expanding service sector, quality plays an increasingly pivotal role in both attracting and retaining service customers (Jenson & Markland, 1996). Service quality research has addressed a variety of fundamentals to assist organizations in creating satisfied customers. Research topics have focused on the implementation of a service quality strategy, anticipation of customer expectations and perceptions, and customer satisfaction and dissatisfaction. In addition to the above, researchers also continue to contribute to the body of knowledge regarding the most effective and reliable forms of service quality measurement. The object of this section is to reveal and discuss the various definitions and models surrounding service quality. In addition, this section will also evaluate the important role that quality plays in service failures and service recovery. QUALITY DEFINED Namasivayam and Hinkin (2003) revealed that the service industry has experienced a continuous transformation over the past decade. As a part of the transformation, creating a quality culture within the organization has become apart of the core values and mission of many organizations. Although most strive to provide quality service and products, customer dissatisfaction, developing customer relationships and creating loyalty appears to be a momentous challenge; especially in the hospitality industry. As customer demands steadily increase and new players in the industry continue to raise the bar, the need for a quality approach to service and products intensifies. There are numerous definitions that facilitate service quality. The nature of a service encounter is defined as the customer?s perception and expectations of all activities 31 involved in the service delivery process; therefore, defining quality can be relatively challenging. O?Neill (2001) acknowledged that the definitions of quality can be united into one concept of consumer-perceived quality, where quality can be defined only by customers and occurs where an organization supplies goods and services to a specification that satisfies their needs. The process involved in maintaining and measuring the quality of goods can be easily defined and implemented. Goods quality is tangible and can be measured by objective indicators like performance, features, and durability (Najjar & Bishu, 2006). However, service quality is intangible; therefore, the definition, strategic process and implementation of a service quality concept become more complex in nature. Lovelock (2001) suggested that the rationalization for defining and determining levels of quality in service corresponds with the ever-changing customer perception and expectations and how well the organization can accomplish service quality standards. In addition to the above, research also suggests that successful organizations are able to diagnose their customer?s expectations fully and satisfy them completely, during each and every service encounter (Zemke & Schaaf, 1990; O?Neill, 2001). Oliver (1997) explained that quality depends primarily on two aspects 1) the product or service being evaluated and 2) the person who is evaluating. Thus, the evaluation and judgment of quality lies in the eye of the beholder. Gronroos?s (2001) research presented a practicum or hands-on approach to understanding quality, suggesting that quality is evaluated in two forms 1) technical quality or what is delivered and 2) functional quality or how is it delivered. THE CHALLENGE WITH SERVICE QUALITY IN HOSPITALITY 32 Although research provides a plethora of ways to evaluate service or product quality, the various definitions stated above shed light on the true challenge in evaluating quality. The challenge in the process of service quality is that various customers, organizations, researchers and educators utilize different fundamentals to assess quality. In 1998, Zeithaml defined service quality as the consumer?s judgment of the overall excellence and superiority of the service encounter. Although the approach to service quality is constantly changing the essential foundation of service quality remains. Managers need to understand quality and how to encourage their staff to implement quality standards and services on a daily basis. However, given the complex nature of quality, the challenge remains not only in the definition but in the approach and the continuous process implemented to maintain quality. O?Neill (2001) explains that given the increasingly competitive nature of the hospitality industry, professionals must concern themselves with a minimal increase in the market share and ways to maintain their existing customer base. The intense competition and market saturation deems pressure placed on service organizations to strive for quality and excellence. Nonetheless, Lockwood (1996) believed that the pressure on organizations to achieve quality comes from three main sources; 1) consumers are more demanding of the products and services they buy, 2) the development of sophisticated technology allows organizations to offer convenient, personal services that are highly valued by the customer, and 3) in an increasingly competitive and international marketplace, quality is seen as providing the edge of competitive advantage. In various sections throughout this chapter, as well as throughout this research, the importance of the intangible elements in the hospitality industry has been discussed. 33 It has been suggested that a number of characteristics in tourism, hospitality and leisure services distinguishes the intangible characteristics from physical goods (Berry, Zeithaml & Parasuraman, 1985; O?Neill, 1992; Reisinger, 1992,). O?Neill (1992) continued by stating that not only do these differences make the service industry unique, each entity comes with its own set of challenges that must be overcome by the service organization. As each entity is faced with the challenge of balancing the intangible characteristics along with the physical goods, maintaining quality standards is more cumbersome. Reisinger (1992) explains that one of the most significant implications for the intangible nature of the hospitality industry is that the services can not be displayed, sampled, tested or evaluated before coming in contact with the customer. The researcher continues to explain that hospitality services are typically produced, sold and consumed in the same location; thus, it is difficult to attain inseparability between production and consumption. Many service organizations struggle when trying to deliver the same quality service and product regardless of the time of day, employee or employees responsible for the task or the manager on duty. In consideration of various attitudes toward service and quality, in addition to the labor intensive environment, it is often cumbersome to have every employee of a hotel or restaurant performing in unison; doing things right the first time and every time (Berry, Zeithaml & Parasuraman, 1985; Resinger, 1992). As heterogeneity has been discussed in a previous section of this chapter, it only seems fitting to revisit the heterogeneity characteristic once again in regard to quality. O?Neill (1992) explained that the differences within each hospitality service entity are unique in nature and are forced to confront their own quality service obstacles on a daily 34 basis. To add to the vast business or operation variations, heterogeneity forces hospitality professionals and researchers to explore the quality dilemmas created by people. The service industry, in particular the hospitality industry, has been built on people delivering a service or creating a product for other people. Thus, since the service experience varies from employee to employee and customer to customer each service experience is unique. Just as the concept of perishability was discussed in service, it also applies to service quality in the hospitality industry. Services in the hospitality industry can not be reserved for a later date or frozen in time. Services, tangible or intangible, are time sensitive and short-lived; thus, managers and employees have a small window of opportunity to create the service quality experience, satisfy the guest and create a loyal customer. The perishability characteristic makes the delivery of quality services and products urgent to customer satisfaction, loyalty and a healthy bottom line. For example, a hotel room that is not filled for the night or a seat on an airline flight not filled at the time of take-off is lost potential profit. SERVICE QUALITY AND SATISFACTION IN HOSPITALITY Ramaswamy (1996) suggested that a lack of quality service or meeting customers? expectations will result in a lack of customer satisfaction (O?Neill, 2001). When customers are satisfied with the organization?s tangible goods and service efforts, customers are more inclined to be loyal to the organization and continuously show their support through repeat visits and positive word-of-mouth advertising. When service quality and customer satisfaction are analyzed research tends to aim toward the disconfirmation theory. As outlined in a previous section, the difference between the customer?s expectations and perception is defined as positive or negative 35 disconfirmation. There appears to be an on-going debate in quality research today. Some researchers suggest that service quality is a vital antecedent to customer satisfaction (Berry, Zeithaml & Parasuraman, 1985; Cronin & Taylor, 1992; O?Neill, 1992). However, on the opposite side of the debate, many researchers suggest that customer satisfaction is a vital antecedent of service quality (Oliver, 1981; O?Neill, 1992). Either way, it appears that both customer satisfaction and service quality are vital to the overall success of today?s hospitality professionals (O?Neill, 1992). In the hospitality industry today service quality and customer satisfaction are two critical elements to most organizations. Professionals are constantly searching for new and exciting ways to promise more than the competition and deliver on their promise. O?Neill and Palmer (2004) explained that service quality and the degree of satisfaction derived from service quality is becoming the single most important differentiating factors in almost every hospitality environment. For the hotel industry, the increasing competition and expansion of unique services and amenities has forced hoteliers to continuously search for the competitive advantage. Today, customers have an abundance of options when traveling. Customers are educated, well traveled and notorious for conducting extensive research before selecting that perfect hotel, restaurant, resort, or cruise line experience. Individual hospitality organizations face a significant challenge when they attempt to deliver quality service to create satisfied customers; customers who will hopefully demonstrate their loyalty to one exclusive brand. In an effort to obtain service quality standards and strive for customer satisfaction, many hospitality organizations have implemented continuous quality improvement strategies to deliver on their promises. The measurement and testing of service quality 36 signifies the organization?s commitment of quality to the customer. Customers often view service quality as the organization?s demonstration of respect and appreciation. In cultivating a continuous quality improvement effort, one of the most important forms of analysis is to focus on the customer?s opinions and feedback. An integral part of any organization?s attempt to instill a ?quality culture? is a commitment to a process of ?continuous improvement? (Witt & Muhlemann, 1995; O?Neill, 2001). In order to remain competitive in the marketplace and to be recognized as a leader in service quality, an organization must continue to utilize different forms of formal and informal measurements. As O?Neill (1992) explained in detail, ?in an attempt to achieve a sustained competitive advantage, hospitality organizations are inventing heavily into a host of service quality improvement initiatives. Whether formal or informal styles of analysis, most initiatives stem from the British Standards Institute, the European Quality Award, the Malcolm Baldrige National Quality Award, the Edwards Deming Prize or derivatives thereof. In addition, the hospitality industry has also been investing quite heavily in raising quality standards through human resource development. Such initiatives include the Investors in People Award, the Welcome Host initiative, and various vocational qualification schemes? (p.167). THE MEASUREMENTS OF SERVICE QUALITY As the need for service quality escalates, the need for continuous quality improvement measurement becomes prominent. Gronroos (1983) provided the groundwork for service quality models utilized in research today. The Service Quality Model consists of two main parts, technical quality and functional quality. Gronroos 37 (1983) explained that technical quality is what the customer actually receives from the interactions with the service organization. Functional quality is the way in which the technical quality is delivered to the customer. The two combine create Image or the service organization?s reputation. Finally the three elements together create experienced quality. The combination of technical quality, which is tangible in nature, paired with functional quality, which are intangible in nature, leads to the overall satisfaction level (Gronroos, 1983). Gronroos (1983) further explained that the overall level of satisfaction depends on the customer?s expectations when entering into the service experience and how the experienced quality measures up in the end. As expressed in Figure 3, Gronroos?s (1983) Service Quality Model shows that technical quality and functional quality are interrelated and directly affect the Image portrayed to the customer. This in turn leads to Experienced Quality. Within experienced quality there are two areas; 1) experienced service as it relates to technical quality and experienced service as it relates to functional quality. Figure 4. Service Quality Model Experienced Service Image Functional Quality Technical Quality Experienced Service Experienced Quality 38 Source: (Gronroos, 2001). In 1985 Berry, Zeithaml, and Parasuraman revealed a model, presented in Figure 4, which demonstrates the dimensions of service quality as it relates to the customer?s role in the service encounter. Most often referred to as SERVQUAL, the model revealed five dimensions of satisfaction leading to the development of their RADER Model (1985). The five dimensions outlined in their research represent the dimensions that Barry, Zeithaml and Parasuraman found to be important in the eyes of the customer (O?Neill, 1992). Martin (2007) described SERVQUAL as one of the most commonly used constructs when measuring the relationship between service quality and satisfaction. The five elements of RATER include: ? Reliability: The ability to perform the promised service dependably and accurately. ? Assurance: Knowledge and courtesy; their ability inspire trust and confidence. ? Tangibles: The condition of the facility, equipment and appearance of the facility. ? Empathy: Caring, individuated attention; appearance of the personnel. ? Responsiveness: Willingness to help, provide prompt and attentive service. The model is believed to be a concise representation of the core criteria that customers employ in evaluating service quality and is a representation of the relationship between the dimensions of service quality, customer expectations and the customer?s overall perception of service quality. In addition, Berry, Zeithaml and Parasuraman (1985) also signifed the correlation between a customer?s expected service and attributes that directly effect a customer?s expectations. The attributes are word of mouth advertising, personal needs, past experiences, and external communications. 39 Figure 5. RATER Model Source: (Zeithaml, Parasuraman & Berry, 1985). Research suggests that SERVQUAL measurement is the most common method used in academia and the service industry. According to Shahin (2005), SERVQUAL is the most reliable option of service quality measurement for three reasons 1) it allows for comparison before and after changes, 2) the location of quality related problems can be identified, and 3) organizations can establishment clear standards for service delivery. In addition, SERVQUAL has become the most often used form of measurement due to its capabilities to examine the customer?s expectations before the service encounter and to compare the customer?s expectations with their perception after the service encounter (Shahin, 2005). Dimensions of Service Quality Tangibility Reliability Responsiveness Assurance Empathy Word of Mouth External Communic ations Past Experience Personal Needs Expected Service Perceived Service Perceived Service Quality 40 Although SERVQUAL is applied in numerous studies, there appears to be numerous challenges with the theory of SERVQUAL. As Parasuraman, Zeithaml and Berry (1985) recognized in their research, SERVQUAL was developed with the belief that the measurement of service quality in the eyes of the customer is paramount. When customers evaluate the service process, they evaluate both the tangible and intangible elements of their service experience; however, the ways in which customers evaluate the two differ dramatically (Parasuraman, Zeithaml & Berry, 1985; Oliver, 1997). Coulthard?s (2004) evaluation revealed that SERVQUAL has generated more criticism that any other service quality model. The criticisms span across various elements including: 1) conceptual basis, 2) process orientation, 3) dimensionality, 4) the use of gap scores, 5) different scores, 6) problems in interpretation of the midpoint on Likert scales, and 7) order effects (Coulthard, 2004). Coulthard (2004) continues by stating: ?it is apparent from recent research that SERVQUAL encapsulates only certain aspects of service quality, and that it fails to capture other potentially less controllable components that may have a greater impact upon evaluations of the quality of the service provision? (p. 483). To evaluate the challenges with SERVQUAL further, Cronin and Taylor (1992) found that SERVQUAL confuses satisfaction with attitude. Their research revealed that service quality can be conceptualized as attitude and can be operationalized by the adequacy-importance model. In addition, they also found that performance-expectation determines service quality opposed to performance alone. This in turn led to the development of SERVPERF. Martin (2007) explains that where SERVQUAL appeared to have a good fit in only two of four industries examined, SERVPERF had an excellent 41 fit across all four industries. Lee, Yoo and Lee (2000) found a similar result from their regression analysis research on SERVPERF. O?Neill and Wright (2002) revealed that when customer expectations are not met, the customer will experience a gap between their experience and their expectations; thus, resulting in a lower level of service satisfaction. O?Neill and Wright (2002) argued that the gap between the customer?s expectations and the customer?s experience can directly affect the customer?s desired level of satisfaction or force the customer to be completely dissatisfied. As stated earlier, the perception of service quality is one of the most prominent challenges in a continuous quality improvement process. This sheds light on another common model applied in service quality research referred to as the Gap Model. The Gap Model, Figure 6, was developed by Parasuraman, Zeithaml and Berry in 1985. The model identifies five gaps or factors that define and explain problem situations in service quality and customer satisfaction. Parasuraman, Zeithaml and Berry (1985) presented the Gap Model as follows: ? Gap 1: Consumer Expectations ? Management Perception. Gap 1 occurs when management fails to properly perceive the customer?s service expectations. Rosen (2003) outlines three situations that may create the Consumer Expectations ? Management Perception Gap. First, managers assume they understand the customer?s needs without conducting the necessary market research. Second, managers fail to dedicate the needed time to talk with the customer and truly dissert their needs or expectations. Finally, with some service organizations, the numerous layers of management often cause a communication breakdown; thus, 42 the right managers or employees are not truly aware of the customer?s needs or expectations. ? Gap 2: Management Perception ? Service Quality Specifications. Gap 2 occurs when management has the wrong perception of the intended service quality standards. Service organizations rely on the members of management to implement and maintain service standards. Often service organizations struggle with sub-standard expectations for tangibles, the service process or employee qualifications; thus, managers may expect more of their employees than what they are trained or qualified to deliver. Contrary to that theory, managers tend to focus on short-term goals, profits and cost reduction opposed to quality standards and at times this creates a lackadaisical approach to maintaining quality standards set by the organization. Therefore, allowing employees to maintain lesser standards. ? Gap 3: Service Quality Specifications ? Service Delivery Gap. Gap 3 constitutes a breakdown in employee ? management communication. Some organizations suffer from significant communication breakdowns where employees are not properly trained or qualified to meet the standards; therefore, employees do not correctly translate the standards or guidelines into action. In addition to poorly trained or qualified, problems with teamwork, poor or inadequate equipment, or poor technology also lead to Gap 3 in the service quality breakdown. ? Gap 4: Service Delivery ? External Communication Gap. Gap 4 represents a breakdown in external communications between the service organization and the customer. This is a service situation where the customer?s needs and expectations are not truly understood and do not match the service delivery. Some 43 organizations have a tendency to over promise on their service quality ideals through their mission statement, formal goals and objectives; however, they fall short in delivering on their promises. Gap 4 can be created by two separate breakdowns in communication. As discussed above, a communication breakdown between the service organization and the customer is not the only problem area to consider. In addition, Gap 4 refers to the communication breakdown between the marketing and public relations departments and the front line management or employees. Either interruption in communication or service delivery can cause customer dissatisfaction. Figure 6. Gap Model of Service Quality Word-of-Mouth Communication Perceived Service Service Delivery Translation of Perception into Service Quality Management Perceptions of Consumer Expected Service Personal Needs External Communications to customer Past Experience 44 Source: (Parasuraman, Zeithaml and Berry, 1985). There are numerous reasons for evaluating the importance of service quality in a hospitality setting. With the constantly changing demands and expectations of the customer, many service organization focus on service and product quality as a way to differentiae themselves from the competition. As a general rule, most service organizations strive for service excellence in an effort to truly satisfy the customer, offer value and become a leader within their segment of the industry. However, over the past decade service organizations have repositioned their thinking on quality to include customer satisfaction and loyalty (Johnson & Gustafsson, 2000). Johnson and Gustafsson (2000) explained that quality, customer satisfaction and loyalty form a chain of cause and effect that build on one another and can not be treated or managed successfully as individual segments of the business. LOYALTY Many service organizations allocate substantial resources to measure and monitor quality, satisfaction and loyalty in an effort to retain customers and improve service offering. Yet, in the service industry, customer satisfaction and service quality does not always lead to the ultimate aspiration of loyalty. Due to the benefits or retaining existing customers, the development of customer loyalty has become an important focus for marketing strategy research in recent years (Gwinner, Gremler, & Bitner, 1998; Hagen- Danbury & Matthews, 2001; McMullan, 2005). According to Olorunniwo, Hsu and Udo (2006), loyal customers impact the profitability and overall success of the organization in three ways. First, a customer?s repeat business generates income for the company, second, due to the cost of marketing and advertising, an organization makes less of a 45 financial commitment in retaining customers compared to recruiting new customers and third, loyal and satisfied customers often spread the good news and recommend the services to others. Although the study of customer loyalty has been of interest over the past decade, most service organizations? approach to measuring customer loyalty has changed significantly. At one time, customer loyalty research concentrated primarily on buying behavior; promoting the vast development of reward and points systems; however, more recent research efforts focus on customer?s expectations, perceptions and attitudes toward loyalty (Mason, Tideswell & Roberts, 2006). In the on-set of loyalty research, the behavioral perspective of customer loyalty was less cumbersome to measure. However, Reichheld (2003) explained that the problems associated with behavioral research are that loyalty is much more than repeat purchases. Some customers may consider themselves loyalty to an organization but their buying may be out of inertia, or exit barriers opposed to loyalty. This in turn has created the need for researchers to explore the significance of customer expectations, perceptions and attitudes and how the three directly influence the customer?s future behavioral intentions. In 1999, Shoemaker and Lewis introduced a concept that the service organization must execute three functions of in order to engender customer loyalty. The contribution of their research resulted in the introduction of The Loyalty Triangle; an equal lateral triangle with each side representing one of the functions needed to execute loyalty. Featured in Figure 7, a summary of Shoemaker and Lewis?s (1999) description is explained below. The first side of the triangle is labeled as process or how the service works. For the customer, the process includes everything that happens from the time they 46 enter into the service experience to the time the service experience ends or the customer leaves the property. The second side of the triangle is Value Creation. Value Creation has been divided into two components; value added and value recovery. Value added includes strategies that increase the long-term value of the relationship with the service organization. Value Recovery was designed to rectify a service problem or failure that may occur during the service experience. The last side of the triangle is Database Management and Communication. This side incorporates the importance of communication between the service organization and the customer. Communication can take many forms such as newsletters, advertisements, special promotions, or courtesy calls to notify the customer of special events, featured favorites or limited availability. Figure 7. The Loyalty Triangle Source: (Shoemaker & Lewis, 1999) LOYALTY DEFINED There are many definitions of loyalty that make the overall concept and the ability to measure customer loyalty somewhat burdensome. In 1969, Day revealed that the concept of customer loyalty was the combination of repeat purchases from one service Customer Process Value Creation (Added and Recovery) Database Management/ Communication 47 provider or brand coupled with the customers? psychological attachment toward the provider. Newman and Werbel (1973) defined loyalty as those who repurchased a brand, considered only that brand, and did no brand-related information seeking. In 1978, Jacoby and Chestnut stated that the only way to understand brand loyalty was to examine the belief, attitude, and conative structure of the customer?s orientation of the focal brand. The basis of their research proposed that all three evaluative stages had to be in place for true loyalty to exist; therefore, the customer was able to become loyal first in a cognitive sense, then later in an affective (attitude) sense and finally in a conative or behavioral sense. In 1994, Dick and Basu found that customer loyalty was viewed as the strength of the relationship between a customer?s relative attitude and repeat patronage. Their research examined relative and attitudinal loyalty in four dimensions; true loyalty, latent loyalty, spurious loyalty and no loyalty. Although Dick and Basu?s research has been citied in numerous textbooks and professional articles, their loyalty segmentation does not appear to be applied in current research (East & Sinclair, 2000). Just as Oliver has made significant contributions to satisfaction research, his research interests and expertise also incorporate customer loyalty. In 1997, Oliver states: ?loyalty is described as a deeply held commitment to rebuy or repartonize a preferred product or service consistently in the future, thereby causing repetitive same-brand or same-brand set purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior? (p. 392). To contradict Newman and Werbel?s definition of customer loyalty, Oliver (1997) further stated that a customer could only be loyal to a service or brand when the customer is 48 aware of the competitor?s services, amenities, and value and yet ?fervently desires to rebuy a product or service; will have no other, against all odds and at all costs? (p. 392). In addition to his differences in the definition of loyalty, Oliver?s (1997) research successfully modified and applied the cognition-affect-conation pattern developed by Jacoby and Chestnut. Oliver (1999) justified his modification by suggesting that although Jacoby and Chestnut (1978) explored the psychological meaning of loyalty to distinguish behavioral (repeat purchase) definitions. Jacoby and Chestnut (1978) analysis concludes that consistent purchasing as an indicator of loyalty could be invalid because of happenstance buying or a preference for convenience and that inconsistent purchasing could mask loyalty if customers were miltibrand loyal. Oliver?s (1999) research differs in his viewpoint that customers can progress in their degree of loyalty at each phase and, in order for the customer to truly be loyal, the end result must be accompanied by a final phase of action. Oliver (1999) suggested that customers first become loyal in a cognitive sense, then in an affective sense, later in a conative manner and finally in a behavioral or ?action loyalty? manner. Action (behavioral) loyalty, or the fourth phase of loyalty, as suggested by Oliver (1999), defined in 1985 by Kuhl and Beckmann, as the point in time when a customer?s intentions are converted into actions; thus, the customer?s motivated intentions in the previous loyalty state has transformed or progressed to the point where the customer is ready to act upon their intentions. Action is perceived as a necessary result of engaging three phases and is accompanied by an additional desire to overcome obstacles that may prevent a customer from patronizing the service organization (Oliver, 1999). 49 McMullan and Gilmore (2003) explored the relationship between four phases of customer loyalty described by Oliver. The researchers explored the customer?s attitude over the four phases of loyalty despite the impact that marketing tactics or situational influences have on customer loyalty. As outlined in Figure 8, McMullan and Gilmore (2003) explained that a customer begins at level 0, when he or she holds no information or have not developed an attitude toward the product or service. During level one, the customer develops a set of beliefs that may find a certain product or service superior to others. Level two is the point in which the customer develops a favorable or unfavorable attitude toward the service or provider; which in turn, leads to the third phase; the customer?s intention to return or recommend. The fourth and final phase requires the customer to take action and act on his or her beliefs and attitude. McMullan and Gilmore (2003) suggested that a customer may progress through each phase of loyalty, disregarding situational and marketing tactics. Situational tactics include poor service, quality or service failures. Marketing tactics include advertising, promotions or other influences the competition may create to influence a customer?s purchase. Figure 8. The Four Phases of Loyalty Level 0 Level 4 Action Level 3 Conative Level 2 Affective Level 1 Cognitive Phase Situational Influences Marketing Tactics 50 Source: (McMullan & Gilmore, 2003). McMullan (2005) suggested that although numerous studies have distinguished between attitudinal and behavioral dimensions of loyalty (Jacoby & Kyner, 1973; Jacoby & Chestnut, 1978; Dick & Basu. 1994; Knox & Walker, 2001), ?these have not adequately explored the complex inter-relationships between the two dimensions and the dynamic processes by which loyalty is initiated and sustained? (p. 470). As customers progress through the phases of loyalty development, the sustainers and vulnerability elements change to reflect the degree of involvement (McMullan, 2005). This in turn led to the need to further explore and define this element of customer loyalty; therefore, McMullan?s (2005) research sought to once again apply Oliver?s (1997) framework to further explore and define the dimensions of customer loyalty and the process in which customer loyalty is initiated and sustained. The findings of the research revealed that a customer must have a sense of loyalty to the service organization during all four phases of the cognition-affect-conation-action pattern in order to overcome vulnerabilities introduced by the competition. In addition, McMullan (2005) highlighted the fact that a customer may prefer one brand or service over the others; however, if an alternative brand or offer is more appealing and the customer feels that the preferred service organization is not worthy of their loyalty, the possibility of switching to the competition increases. McMullan?s (2005) contribution led to the construction of a multi-item scale to measure customer loyalty development. McMullan (2005) supported the belief of Reichheld (2003) when he stated: ?a loyal customer is one who values the relationship with the company enough to make the company a preferred supplier. Loyal customers do not switch for small 51 variations in price or service, they provide honest and constructive feedback, and provide enthusiastic referrals? (p. 126). CUSTOMER LOYALTY AND LOYALTY PROGRAMS Differentiating between true customer loyalty and incentive loyalty poses another challenge when trying to measure the effectiveness of customer loyalty. In recent years there has been a rise in the use of loyalty programs by marketers and academic professionals in measuring loyalty (Sharp & Sharp, 1997). Loyalty programs and incentives were perceived to be one of the most successful marketing tools of the 1990s and have received considerable attention in both trade and academic journals (O?Malley, 1998). Although numerous organizations in the service industry have experienced much success with loyalty reward programs, formal loyalty reward programs have not proven to be as beneficial in the hospitality industry. Even though evidence suggests that hotel, restaurant, airline and cruise line customers respond favorably to loyalty programs, the ?hotel rewards? and ?frequent flyer programs? were developed around transactional tactics (Bowen & Shoemaker, 1998). Unfortunately, research has acknowledged that true customer loyalty cannot be obtained solely by rewarding customers with such monetary-based rewards (Dowling & Uncles, 1997; Duffy, 1998). Shoemaker and Lewis (1999) indicated that true customer loyalty is earned through the development of nonmonetary strategies that focus on the personalization of customer services, individual recognition of the customer, and the formation of a level of trust and commitment between the service organization and the customers. 52 Dube and Renagham (1999) extended the notion that essentially, for a service organization to ensure continued commitment and patronage; customers need to feel that they are receiving additional value in return for their loyalty. To support the effectiveness of loyalty programs in the hospitality industry Gilbert (1996) and Bejou and Palmer (1998) reviewed the effectiveness and development of frequent flyer programs in the airline industry. Interestingly, both studies revealed that the expansive offerings of loyalty programs among the major airlines paired with the limitations and difficulty in using the loyalty program resulted in a decline in traveler loyalty (Bejou & Palmer, 1998). LOYALTY IN THE HOSPITALITY INDUSTRY Shoemaker and Lewis (1999) defined loyalty as it pertains specifically to the hospitality industry: ?loyalty is when the customer feels so strongly that you can best meet his or her relevant needs that your competition is virtually excluded from the consideration set; these customers buy almost exclusively from the preferred service organization- referring to the organization as their hotel or their restaurant? (p. 349). Loyalty in the hospitality industry has been one of the most researched areas in recent years (Tideswell & Fredline, 2004) and over the last two decades, customer loyalty in hospitality research has focused on various aspects. Knutson (1988) investigated a customer?s likelihood to return to a hotel property based on the most important features offered at the hotel and found that a clean and comfortable room was the determining factor. Even though their research considered the impact of return visitation, Knutson 53 (1988) failed to investigate the effects on other loyalty related attitudes such as the impact on word of mouth or price sensitivity issues (Tideswell & Fredline, 2004). In terms of service quality and its effect on customer loyalty, Dube and Renaghan (1999) explored the effect of several functional areas of the hotel on customer loyalty outcomes and found that service quality and the quality of the employees were two of the most important factors effecting loyalty and intentions to repurchase and recommend. In 1999, Shoemaker and Lewis reviewed a range of loyalty-generating strategies used in the hotel industry. Their research revealed that for true loyalty to be achieved, service organizations should focus on customer recognition, service quality, personalized service, emotional rewards and tailor made offers. Once again, Shoemaker and Lewis (1999) recognized that reward programs in the hospitality industry, specifically in the hotel segment, result in loyalty to the frequency program opposed to the actual brand or hotel. Further contributions of their research also revealed six types of value-added or value-recovery strategies used when generating customer loyalty. The six loyalty- generating strategies in the hotel industry are: ? Social: Interpersonal links with the hotel via regular communications. ? Emotional: Providing a sense of recognition using guest?s names. ? Experiential: Providing extra services such as turndown services. ? Functional: Providing extra facilities such as Internet access. ? Temporal: Time savings through quick check-in services. ? Financial: Monetary savings through discounts on services. Although their loyalty-generating strategy research does not reveal the degree of loyalty associated with each type of strategy listed, it is reasonable to assume that the highest 54 degree of loyalty is likely to result from the Social, Emotional and Experiential strategies with the Functional, Temporal and Financial strategies generating a lesser degree of loyalty (Tideswell & Fredline, 2004). As stated previously in this chapter, service quality in a vital antecedent to customer loyalty. In 2001, Fu and Parks analyzed the relationship between service quality and customer loyalty in elderly restaurant customers in the United States. Their findings revealed that elderly diners were more likely to revisit a specific restaurant because of the friendliness and empathy of the restaurant employees opposed to the speed of service or the quality of the product. Bowen and Shoemaker?s (1998) research on service quality and customer loyalty revealed that customer?s searched for service quality in terms of receiving hotel upgrades, being able to request a specific hotel room or being advised when the hotel had a special event or limited availability. MEASUREMENTS OF CUSTOMER LOYALTY In 2003 Shoemaker and Bowen evaluated the evolution of customer loyalty as practiced by organizations both inside and outside of the hospitality industry. When loyalty efforts were first realized to add value to the organization, most service organizations relied upon the sales and targeted promotion approach (Dube & Shoemaker, 2000). This concept offered discounts and promotions with little to no effort being made to track the overall effectiveness of the process. As time and research efforts progressed, the origin of the frequency programs emerged. As outlined in a previous section, although frequency programs are beneficial to some service organizations they were not as well received in the hospitality industry. 55 Today, many service organizations focus their loyalty efforts on building and strengthening relationships between the service organization and the customer (Shoemaker & Bowen, 2003). When building relationships between the customer and the service organization, knowledge of the customer comes from asking them to provide information on their expectations and perception of the services and products (Shoemaker & Bowen, 2003). This in turn sheds light on the numerous antecedents of customer loyalty. For the purpose of this study, the appropriate antecedents of customer loyalty are service quality, satisfaction and future behavioral intentions. Research suggests that service quality, satisfaction and customer loyalty, together, are invaluable to service organizations; thus, the three must be managed and measured as a cause and effect operating strategy. In fact, Olsen (2002) presented a model, featured in Figure 9, to signify the equality of quality performance and customer satisfaction with customer loyalty. Although the relationship between quality and loyalty and satisfaction and loyalty were positive, results varied across the various stages of the research. Also, results of the study did not support customer satisfaction as a mediating variable between quality performance and customer loyalty (Olsen, 2002). Figure 9. The Satisfaction-Loyalty Model Source: (Olsen, 2002). Quality Performance Customer Satisfaction Repurchase Loyalty 56 Although customer satisfaction may not serve as a mediating variable between quality performance and loyalty, service organizations strive to find a balance between the three functions. Service organizations still operate under the principle that satisfaction and its role in creating loyal customers is pivotal in long-term success. Combine this with the competitive nature of the services industry and the unique nature of intangible and tangible services and the relevancy of studying these attributes becomes clear (Martin, 2007). Edvardsson, Johnson, Gustafsson & Strandvik (2000) created the Loyalty Profit Chain, featured in Figure 10, to explain the relationship between satisfaction and loyalty. Their model suggests that satisfaction affects loyalty and retention. As expected, a combination of satisfaction, loyalty and retention spurs an increase in revenue, lowers operating costs, and increases profitability. Figure 10. Loyalty Profit Chain Source: (Edvardsson, Johnson, Gustafsson & Strandvik, 2000). The authors support the effectiveness of this model by stating, ?The overall result is a per customer profit stream that increased over time. The more loyal the customer and the Customer Satisfaction Revenue Growth Loyalty Profit 57 longer the customers are retained, the more sales and profits the customer generates? (Edvardsson, Johnson, Gustafsson & Strandvik, 2000, p. 919). To review the importance of service quality, Zeithaml et al. (1990) stated service quality has become a great differentiator and the most powerful competitive weapon which many leading service organization strive to possess. Bell, Auh and Smalley (2005) explored Gronroos?s (1983) quality model as it relates to customer loyalty. Their model, presented in Figure 11, proposed technical service quality and functional service quality were both antecedents to customer loyalty. In addition, Bell, Auh and Smalley (2005) explored the effects of investment expertise and perceived switching costs. Figure 11. The Quality-Loyalty Model 58 Source: (Bell, Auh & Smalley, 2005). For the purpose of their study, investment expertise has been defined as the customer?s cumulative knowledge about how a product or service should perform and a general understanding of the competition?s performance of similar services or products (Sharma & Patterson, 2000). Bell et al. (2005) defined switching costs as the financial or psychological costs that a customer expects to endure if he or she would choose to patronize a competitor. Switching costs form a significant challenge when trying to measure true loyalty. Dick and Basu (1994) explained that switching costs can lead service organizations into believing a customer is truly loyal to their organization, when in reality, the customer is less than satisfied with the selected brand but fears the costs endured in selecting a different service organization. The effect of switching costs has been reviewed by researchers and marketers over the last decade (Patterson & Smith, 2003). Financial switching costs can be viewed by the customer as a way to trap the customer; thus, giving the service organization the permission to practice poor service Technical Service Quality Perceived Switching Costs Investment Expertise Customer Loyalty Functional Service Quality 59 standards knowing that it is more difficult to lose their customer base (Jones, Mothersbaugh & Beatty, 2002). Many service organizations, maintaining a superior quality of service is their number one effort to gain customer loyalty (Zeithaml & Bitner, 2000); thus, a service organization?s long-term success in a market is essentially determined by its ability to expand and maintain a large and loyal customer base (Kandampully, 1998). To a service organization, customer loyalty is viewed as the customer?s commitment to patronize that particular provider over the competition; however, customers expect a valued commitment from the service provider in return. This statement is supported by Kandampully (1998) when he explained while service organizations aim to gain customer loyalty, customers on the other hand, seek an organization?s service loyalty. Service loyalty has been defined as the organizations assurance of a consistent and superior quality of service as proof of the organization?s commitment to offering superior service, for both the present and the long term (Berry, Wall & Carbone, 2006). Although at times the lines of customer satisfaction and customer loyalty become indistinct, service organizations should be mindful of their drastic differences. Customer satisfaction measures how well the customer?s expectations are met by a given transaction or service experience whereas customer loyalty measures how likely a customer is to return, recommend or help to improve the overall organization (Bowen & Shoemaker, 1998). Although many researchers (Oliver, 1997; Bowen & Shoemaker, 1998; Gustafsson & Johnson, 2004) testify that satisfaction is a necessary step toward customer loyalty; satisfaction does not always translate into loyalty (Oliver, 1999). 60 Bowen and Shoemaker (1998) outlined reasons that satisfaction does not always translate into loyalty. The reasons discussed were: 1) travelers who do not return to a specific area cannot be loyal to the organization, even if they were willing, 2) when exploring luxury hotels, some hotels customers enjoy a variety of choices and service offerings and do their best to experience as many different organizations as possible, 3) no matter how satisfied, some customers still remain price sensitive; thus will shop for the most reasonable price, and 4) some customers just expect to receive quality service and be satisfied every time; therefore, those customers do not view those attributes as reasons to earn their loyalty. Oliver?s (1999) contribution supported the above when his loyalty research revealed that for satisfaction to affect loyalty, satisfaction would have to be frequent and cumulative with the customer experiencing a deep commitment to consistently repurchase and recommend despite situational influences. Oliver (1999) continued by clarifying as the customer?s loyalty to a service organization strengthened, the steps necessary to form loyalty (i.e. quality, satisfaction) became less significant. One of the most sought after attributes of loyalty is the customer?s future behavioral intentions toward their preferred service organization. Perhaps the most common assessment of loyalty have been the behavioral measures of repurchase patterns and the likelihood to recommend expressed over time (Bloemer & Kasper, 1995). While the weakest link in the Service Profit Chain was the link between satisfaction and customer loyalty, Cronin and Taylor (1992) found that satisfaction had a stronger and more consistent effect on purchase intentions compared to service quality. However, the loyalty construct is multidimensional and includes not only repeat buying, but also considers other aspects such as decreased price sensitivity and word-of-mouth behavior 61 (Wangenheim & Bayon, 2004). Reichheld and Sasser (1990) found that a customer?s repurchase intentions and likelihood to recommend the service organization determines the lifetime value of the customer, and even small differences can have an impact on the profitability of the organization (Bolton, 1998; Reichheld & Teal, 1996). THE DIFFICULTY IN MEASURING CUSTOMER LOYALTY Just as the definitions of customer loyalty can be problematic, the ways in which to measure customer loyalty are just as difficult and complex. Mason, Tideswell and Roberts (2006) outline the complications of customer loyalty measurement by explaining that on one hand loyalty measurement is devised through the power of an object or a program (i.e. rewards program) that is purposely designed to influence or control the probability of a customer being loyal. On the other hand there is the customer?s propensity to be attracted to the service organization and loyalty is measured through an individual?s personal psychology. The attributes closely associated with customer loyalty are repeat purchasing, a customer?s willingness to recommend the service or product and the propensity to partner or be a positive influence on the growth and success of the organization (Jones & Sasser, 1995). However, as Jones and Sasser (1995) recognized, although repeat purchases are somewhat easy to observe and measure, repeat purchasing approaches to loyalty measurement are vulnerable and infrequent. Mason et al. (2006) clarified further by suggesting that there is a lack of structural evidence on how many times a behavior has to be carried out to qualify as frequent or how much time between service encounters is permissible. Dick and Basu?s (1994) research attempted to quantify the number of service encounters a customer must have to qualify as a loyal customer; however, their research 62 method resulted in categorizing low-frequency customers in the latent loyalty or no loyalty category whereas high-frequency customers were grouped in the loyalty category. Therefore, according to their research, Dick and Basu (1994) revealed that only high- frequency customers were truly loyal to a service organization. In the hospitality and tourism industry, Riley (2001) recognized that it could be years or decades between visits to a destination, hotel or restaurant; however, the customer may have been showing signs of true loyalty by harboring positive feelings and recommending the destination or property to business associates, friends and family. Thus, the frequency of visits or the time between service encounters is not a reliable form of measurement. EMOTION Supportive literature has contributed volumes of research to dissect the complex nature of customer satisfaction and the constructs that surround customer satisfaction. However, while the connection between emotions and satisfaction has been recognized by managers and academics, most forms of satisfaction research do not address emotional output (Liljander & Strandvik, 1997; Yu & Dean, 2001). To some degree the elimination of emotion as it relates to customer satisfaction, future behavioral intentions and the customer?s willingness to be loyal to an organization has been due in part to the complexity of the construct and the challenges presented when measuring customer emotion. Over the past decade researchers have made great strides with the development of various scales to scrutinize the importance of emotion in a customer satisfaction scenario; however, the most effective way to measure emotion is still debated (Westbrook & Oliver, 1991; Mattila & Wirtz, 2000; Smith & Bolton, 2002). 63 Current literature serves evidence to the theory that the relationship between satisfaction and loyalty is moderated by positive emotions in high involvement services; that is, the stronger the positive emotions experienced during the service encounter, the stronger the relationship between satisfaction and loyalty (Bloemer & de Ruyter, 1999; Scanlan & McPhail, 2000). In addition to the above, previous research has overwhelmingly focused on cognitive components of customer satisfaction and much of this has used some form of disconfirmation to compare perceived levels of performance with some form of benchmark standard (Liljander & Strandvik, 1997; Yu & Dean, 2001). While these studies have concluded that there is a significant relationship between service quality, customer satisfaction and future behavioral intention, the validity of the findings is more closely related to the cognitive components of the satisfaction construct opposed to the emotion construct (Liljander & Strandvik, 1997; Yu & Dean, 2001). Liljander and Strandvik (1997) explain that service quality and satisfaction are believed to contain an affective (emotional) component; thus, without the analysis of the customers? responses the emotion component cannot be fully appreciated. A growing body of literature clearly indicates that the positive and negative emotions that customers associate with the service experience play an important role in satisfaction and future behavioral intention (Allen, Machleit & Kleine, 1992; Oliver, 1993; Richins, 1997; Barsky & Nash, 2002). Without a doubt, it is widely accepted that customer satisfaction, loyalty and future behavioral intentions are to some degree influenced by pre-encounter emotions and post-consumption emotions (Oliver, 1997; Cronin, Brady & Holt, 2000; Barsky & Nash, 2002). EMOTION DEFINED 64 Related literature offers numerous definitions of emotions in the psychology and marketing literature. Oliver (1997) ascertained that ?Emotion includes arousal and various forms of affect and cognitive interpretations of affect that may be given a single description. One of the most predominant problems in defining emotion is to understand where emotion stops and where cognition begins. The greater the amount of cognitive interpretation required, the more cognitive the emotion becomes? (p. 294). To analyze the definition of emotion further, in 1991 Westbrook and Oliver stated that emotion is a primary source of human motivation which exerts substantial influence on satisfaction, loyalty and future behavior. In 1998, Bourne and Russo contributed to the definition of emotions by stating that emotions are based on 1) a combination of biological and cognitive factors, 2) are highly influenced by society and peers, and 3) emotions often change drastically due to physiological changes in the customer?s environment. In addition to the above, the authors also recognized that an individual?s emotional state can change based on their perception of a situation (i.e. a service failure or an effective service recovery experience). Although numerous definitions of emotion exist, Desmet and Hekkert (2003) noted difficulties in defining emotion and suggests that before one can measure emotions, one must first be able to characterize and distinguish them from other states. While it has been hard for researchers to define emotions, many specific definitions have been developed by researchers in regards to consumption emotions. CONSUMPTION EMOTION 65 For the scope of this study, it is most appropriate to review the significance of marketing related definitions of emotion or consumption emotions, along with the role that consumption emotions play in a service encounter. Numerous researchers (Westbrook & Oliver, 1991; Matilla & Wirtz, 2000; Smith & Bolton, 2002) have conducted research around consumption emotions in the service industry and specifically in the hospitality industry. As a result of their research efforts, consumption emotion has been defined as a set of emotional responses elicited specifically during product usage or consumption experiences. Each of their research efforts explores various elements of emotion as described either by the distinctive categories of emotional experience, expression (joy, anger and fear), or by the structural dimensions underlying emotional categories, such as pleasantness/unpleasantness, relaxation/action, or calmness/excitement emotions (Westbrook & Oliver, 1991). When looking at consumption emotion, recent literature clearly indicates that the customer associates both positive and negative emotions with the service encounter; which in turn, plays a significant role in the customer?s degree of satisfaction, loyalty and future behavioral intention (Allen, Machleit & Kleine, 1992; Oliver, 1993; Richins, 1997; Barsky & Nash, 2002). Research also lends evidence that customer satisfaction, loyalty and future behavioral intentions are, to some extent, influenced by the customer?s emotions prior to the service encounter and directly after the service encounter (Oliver, 1997; Cronin, Brady & Holt, 2000; Barsky & Nash, 2002). Mattila and Wirtz (2000) recognized that in order to truly understand a customer?s emotional state, research should be conducted in two phases; 1) the customer?s pre-consumption stimuli and emotion, and 2) the customer?s post-consumption stimuli and emotion. This would aid in an effort to 66 fully understand the cognitive process, on-going satisfaction, degree of loyalty and their future behavioral intentions. In addition, Price, Arnould and Deibler (1995), indicated that the customer is typically more engaged in a service encounter; therefore, the customer has a time and emotional investment in that interaction. The researchers suggest that the customer?s emotional attachment to a service encounter is greater when compared to their emotional attachment to advertisements or product purchases. In regards to the intangible nature of the service industry, it appears that a customer?s emotional attachment to an organization is greatly attributed to the people and intangible service elements in the service environment. In consideration of the above, the impact of the relationship between the service provider and the customer plays a vital role in analyzing emotion and the formation of satisfaction, loyalty and behavioral intentions. Clearly the relationship between emotion, satisfaction, loyalty and behavior is crucial when measuring the value of service quality. MEASUREMENTS OF EMOTION Mattila and Wirtz (2000) suggested that in order to truly understand the characteristics of service, the measurement of feelings and emotions experienced during the delivery process are necessary to adequately capture the experimental nature of the service encounter. Although the measurement of emotion is a critical key, Ekinei and Hosany (2006) criticize that there has been a lack of consistency when it comes to the measurement of emotions in the service encounter Three of the most popular scales to date that relate to this research include Izards? (1992) Differential Emotions Scales; Sui & Baloglu?s (2003) research on emotional 67 commitment in relationship marketing; and Smith and Bolton?s (2002) research on emotional responses to service failures. Izards? Differential Emotions Scales (1992) is an amended version of his 1972 research and consists of a combination of ten positive, neutral and negative emotions. The combinations of emotions are listed below: ? Interest-Excitement ? Happiness-Joy ? Surprise-Astonishment ? Sadness-Grief ? Anger-Rage ? Disgust-Revulsion ? Fear-Terror ? Contempt-Scorn ? Shame-Shyness ? Guilt-Remorse. To tailor the scale to a customer?s emotional responses, Izard (1997) suggests that the basic emotions can be experienced individually, or a customer may experience a combination of positive, neutral or negative emotions. Three specific emotions; anger, disgust and contempt are often considered to be the Hostility Triad (Oliver, 1997, Liljander & Strandvik, 1997, Liljander & Bergenwall, 2004). Izard?s (1997) scale is applied in current literature by asking the customer to attest to the various emotions they were/are feeling during the service encounter. 68 Current literature suggests that Izard?s scale is the most effective in customer satisfaction studies (Oliver, 1997; Liljander & Strandvik, 1997; Liljander & Bergenwall, 2004); however, it is not free from criticism. First and foremost, this scale has been criticized for the predominance of negative emotions. This is not an unrealistic criticism considering that 7 out of 10 emotions are anchored by negative emotions adjectives. Furthermore, because the outcome of the actual service encounter is either strongly positive (a victory) or strongly negative (a loss) the imbalance of Izard?s scale towards the negative leaves it unusable in this setting (Martin, 2007). From a different perspective, Sui and Baloglu (2003) considered that creating satisfied and loyal customers is based on the service organization?s relationship marketing philosophies. Their research contributed to emotions literature by exploring the role of emotion in customer loyalty and future behavioral intentions in the casino industry. Their study set out to 1) evaluate the role of emotional commitment in relationship marketing, 2) support the theory that a relationship exists between trust, switching costs and emotional attachment on behavioral outcomes, and 3) to determine if empirical linkages found in previous studies on service loyalty could be applied to the casino industry. Presented in Figure 12, Sui and Baloglu?s (2003) pointed out that a customer?s emotional attachment to the service organization was based on the degree of trust he or she felt toward the service organization and the emotional or financial switching costs associated with patronizing the competition. Their model signified that a customer?s emotional attachment toward the organization influenced five loyalty and behavioral 69 attributes; 1) word of mouth, 2) cooperation, 3) time spend, 4) revisit and 5) other product usage. Figure 12. Emotion in Relationship Marketing Source: (Sui & Baloglu, 2003). To analyze the importance of emotional attachment with relationship marketing and future behavioral intentions, Smith and Bolton (2002) explored the effectiveness of a customer?s service recovery satisfaction and the emotional effects in the up-scale food and beverage industry and several luxury hotels. The nature of the hospitality industry breeds service failures and reoccurring problems. When service failures occur, one of the most important elements is to understand and control a customer?s emotional response to the situation. In doing do, Smith and Bolton (2002) suggested that the organization should 1) attempt to ensure overall satisfaction, 2) engender customer loyalty, and 3) influence the customer?s future behavioral intentions. Abrams and Pease (1993) proposed that in the event of a service failure, if successfully resolved, customers may feel a stronger commitment and emotional Trust Switching Costs Emotional Attachment Use other Products Revisit Intentions Time Spent Cooper- ation Word of Mouth 70 attachment to the service organization than if no failure had occurred in the first place. The scope of service failure and service recovery is presented in detail in the pages that follow; however, for the significance of this section, Smith and Bolton (2002) believed that a service failure presents two areas of opportunity for a service organization. First and foremost, a service failure is an opportunity to truly test the provider?s service philosophy to satisfy and keep the customer, and second, the service recovery experience is a pivotal moment for the customer. Since customers tend to experience strong emotional reactions in response to service failures, many customers rely on their emotions as the deciding factor in whether or not to continue their relationship with the organization (Smith & Bolton, 2002). Presented in Figure 13, Smith and Bolton?s (2002) model is a result of their investigation into the emotional responses surrounding a service failures and how emotion influences satisfaction and behavioral intentions. The scope of their research explored three questions; 1) do emotional responses to service failures influence customer satisfaction with the service encounter?, 2) do emotional responses to service failures influence how the recovery efforts are perceived?, and 3) do emotional responses to service failures influence their overall degree of satisfaction with the service provider? Figure 13. The Role of Emotion is Customer Satisfaction 71 Source: (Smith & Bolton, 2002). As a result of their study, four common service recovery attributes emerged as the main drivers of satisfaction: 1) compensation, 2) quality of response, 3) apology, and 4) whether the service recovery process was controlled by the employee or by the customer. In addition to the above, five negative emotions were identified as being the strongest emotions experienced during the service failure; 1) anger, 2) discontent, 3) disappointment, 3) self-pity, and 4) anxiety. Although this research contributes to the understanding of emotion in the service failure/service recovery context it is not free of criticism. Smith and Bolton?s (2002) research found that when combining the responses collected from the selected restaurants and hotels there was a significant relationship between emotion and satisfaction. Consequently, when analyzing the two industries individually, there was no significant relationship in the restaurant segment and a minimal relationship between emotion and satisfaction in luxury hotels. This was attributed to two situations 1) the experimental Emotional Response Cognitive Antecedents Satisfaction Service Failure Service Recovery 72 design for the restaurant segment did not control for heterogeneity where the hotel segment did, and 2) more negative emotions and harsher comments were reported in the responses collected from the hotels. Thus, the researchers concluded that the more negative the emotion, the more important the recovery efforts. Due to the similarity in research settings, the negative emotion scale applied in Smith and Bolton?s (2002) research was applied for the purpose of this study; however, the scale was modified to include positive emotions. SERVICE FAILURE To create and maintain satisfied customers, the primary goal of a service organization in the hospitality industry should be to operate to a standard of quality and excellence; therefore, doing things right the first time. However, it only stands to reason that the intangible nature creates an environment where things can and frequently do go wrong; thus, service failures can occur. When a service failure occurs, the service organizations run the risk of jeopardizing the relationship between the organization and the customer. Kandampully (2002) cautioned ?Of all the challenges facing hospitality establishments today- including intense competition, globalization and technological innovation- the single most pervasive and pressing challenge is the ever-increasing demand of customers for service quality? (p. xi). Colgate and Norris (2001) recognized that while a considerable volume of research has focused on the one best way to define measure and operationalize the service quality construct, an increasing body of research is now being targeted at the issue of service failure; complaint management and the service recovery process. They continued that hospitality professionals now recognize 73 that a major factor that affects a customer?s perception of quality provision is the number and extent of the problems they encounter, and more importantly how these problems are handled when brought to the provider?s attention. SERVICE FAILURE DEFINED The concept of dissatisfaction in service has been researched extensively. Dissatisfactory experiences have been labeled as negative incidents (Bitner, Booms & Mohr, 1994; Bitner, Booms & Tetreault, 1990; Keaveney, 1995) or unsuccessful processes that trigger failed encounters (Roos & Strandvick, 1997; Smith, Bolton & Wagner, 1999; Coulter & Ligas, 2000); which in turn, cause the customer to realize that the service has not met their expectations (Zeithaml, Berry, & Parasuraman, 1993; Oliver, 1997). Related literature has contributed numerous definitions and has scrutinized the dimensions of a service failure. Bell and Zemke (1987) stated that a service failure occurs when the service delivery falls short of the customers? expectations; which in turn, necessitates the service provider?s response with recovery efforts. Mattila (2001) recognized that in the service industry, especially the hospitality industry, there is a high degree of human contact; thus, service failures are an inevitable element to service. Although service failures are viewed as a defect or problem area within the organization, Zemke and Bell (1990) believed that the true test of an organization?s commitment to service quality is the way the organization responds to the service failure. To enhance that concept, Mittal, Ross and Baldasare (1998) revealed that negative performances have greater influence on satisfaction and purchase intentions compared to positive performances. 74 When a service organization experiences a service failure, the organization is faced with a unique window of opportunity to rebuild trust and commitment and satisfy the customer. To reiterate the disconfirmation theory reviewed in a previous section of this chapter, negative disconfirmation occurs when the service delivery falls short of the customers? expectations (Bell & Zemke, 1987); which in turn necessitates the need for the service provider?s service recovery efforts. Zemke (1993) continued that the object of a service recovery effort is to move a customer from a state of dissatisfaction to a state of satisfaction in an effort to engender loyalty and influence the customer?s behavioral intentions. Not surprisingly, many companies consider the development of a complaint management system to be a vital component of their continuous quality improvement efforts. Put simply, service organizations realize that an effective system is a vital means to increasing customer satisfaction and establishing a long term relationship that leads to customer loyalty and positive behavioral intentions (Tax, Brown & Chandrashekaran, 1998). Lovelock, Patterson and Walker (2001) added to the above when they suggested that ?customer satisfaction evaluations are linked not only to fault-free service, but also to what transpires when something does go wrong? (p. 343). This emerging ideal is a significant advancement from the more traditional approach to handling a service failure. At one time, a service failure was more of an embarrassment to a service organization opposed to an opportunity. Prior to this paradigm, shift a service failure was something that a service organization would quickly brush under the carpet; never to be discussed. However today, service failures are regularly viewed as learning opportunities for service organizations; creating 75 opportunities for service organizations to achieve the competitive edge and understand how to better satisfy the customer. It is suggested that a majority of dissatisfied customers experience two different forms of service encounters. The first is described as a core service failure; which encompasses all actions that are involved in a failed delivery of service (Bitner, Booms & Tetreault, 1990; Bitner, Booms & Mohr, 1994; Keaveney, 1995). The second form of service failure includes negative and improper behaviors by the service provider or a lack of personalized service (Keaveney, 1995; Roos & Strandvik, 1997). Similar to disconfirmation, the amount of dissatisfaction will differ from customer to customer based on the form of the failure, severity, and the customer?s zone of tolerance. The challenge to service organizations today is undoubtly difficult: on one hand organizations must create a non-threatening environment to make complaining easier, while at the same time, design a recovery system around factors that 1) matter most to the customer, 2) build a relationship between the service provider and the customer, 3) engender loyalty, and 4) influence the customer?s future behavioral intentions. One of the most significant factors that effect a customer?s perception of quality is the number, and extent, of the problems they encounter, and more importantly how these problems are handled when brought to the attention of the provider (Colgate & Norris, 2001). To create and maintain satisfied customers, the primary goal of organizations in the service industry should be to operate to a standard of quality and excellence; therefore, doing things right the first time. However, in the service industry, as well as the hospitality industry, it only stands to reason that the intangible nature creates an 76 environment where things can and frequently do go wrong; thus, service failures can occur; potentially jeopardizing the service provider?s relationship with the customer. Although some organizations may view service failures as a devastating experience, Zemke and Bell (1990) believe that the true test of an organization?s commitment to service quality is not in the stylishness of the pledge it makes in the mission statement or the marketing literature but in the way the organization responds to the service failure. In the event of a service failure, measures and standards must be established for the service provider to take immediate action to recover the failed service attempt; thus the need for a quality systematic approach to service recovery is vital to preserving the service provider-customer relationship. SERVICE RECOVERY When service failures occur the negative effects on profitability and customer loyalty create drastic challenges for service organizations to overcome. Although at first a service failure can be a devastating embarrassment for the organization, if handled properly a service provider can counter act those negative effects and challenges by responding to a customer?s complaint in an effective manner. This task can be accomplished by the service organization?s approach to a quality service recovery process. A service organization can create the ideal strategic plan for handling customer complaints; however, a plan which is never implemented only looks pleasing on paper. Thus, one of the most important keys to providing excellent service recovery is convincing the customer to bring the failure to the provider?s attention and allow the organization to implement the service recovery process. 77 Service recovery related literature attributes the social exchange theory and the equity theory for providing the theoretical framework for studies exploring customer?s evaluation of service recovery efforts (Adams, 1965; Ok, Back, & Sharklin, 2005). According to Adams (1965), the two theories demonstrate that the exchange relationship should be balanced between the service organization and the customer. Oliver (1997) explained the significance of social exchange when he stated that the customer?s satisfaction is based on their perception of equal partners during the service recovery exchange. For example, the professionalism demonstrated between employees, managers and employees, and between the service organization and the customer. Oliver (1997) continued to explain that the equity theory; which is also a vital component, is the customer?s perception of fair or unfair interpersonal relations or the amount of compensation involved in correcting the problem. From this research, a distinction between distributive justice, procedural justice and interactional or interpersonal justice was developed, emphasizing differential effects of procedural elements on customer satisfaction outcomes (Greenberg, 1987). SERVICE RECOVERY AND THE JUSTICE THEORIES Understanding the significance behind the justice theories is one of the most essential elements in understanding service recovery satisfaction. Oliver (1997) explained that distributive justice is described as the allocation of benefits and costs; which relates to the degree of compensation associated with recovering the failed service offering. For example, distributive justice would relate to the way in which a customer evaluated the quality of the refunds, credits, charge corrections, replacements and apologies associated with the organization?s service recovery efforts. 78 Procedural justice is related to the customer?s perception of the service recovery process. For example, Wildes (2005) suggested that a uniform approach to service policies and practices generated the highest degree of customer satisfaction through employee knowledge and empowerment. Thus, procedural justice aims at resolving conflicts by use of a formal approach in order to be consistent and maintain the customer relationship- even if the outcome is unsatisfactory to one party. Finally, interactional or interpersonal justice is degree to which people are treated (Oliver, 1997). This typically refers to degree of politeness, dignity and respect displayed by the service organization toward the customers. Numerous studies have focused on the justice dimensions in service recovery satisfaction. One study in particular by Smith, Bolton and Wagner (1999) proposed a model, featured in Figure 14, which extended the notion that the service recovery attributes (compensation, response speed, apology, and initiation) influenced at least one of the three types of perceived justice in selected properties in the hotel and restaurant industries. 79 Figure 14. Justice Theories in Service Recovery Satisfaction Source: (Smith, Bolton & Wagner, 1999). The findings of their research revealed several elements to help academia and practitioners appreciate the relationship between the justice theories and service recovery satisfaction. First and foremost, their research found that the customer?s satisfaction with the service recovery process was positively related to their satisfaction with all three justice theories in both the hotel and restaurant industries. Second, both hotel and restaurant customers related the quality of compensation with distributive justice. Third, the quality of the apology offered and the organization correcting the problem with little effort or stress on behalf of the customer was related to interactional justice. Finally, respondents in both the hotel and restaurant industries attributed the speed of the recovery to procedural justice. Although their research made a significant contribution to service recovery literature, the researchers suggest that the scope of the study could be enhanced Type of Failure Magnitude of Failure Compensation Response Speed Apology Initiation Disconfir- mation Distributive Justice Procedural Justice Satisfaction w/ Service Interactional Justice 80 in two ways; 1) by reviewing the severity of the service failure, and 2) by classifying failures as they relate to each individual justice theory. SERVICE FAILURE AND SERVICE RECOVERY Researchers suggest (Parasuraman, Berry & Zeithaml, 1991; Singh, 1991) that to some degree, overall satisfaction in a service failure situation is determined by two factors: 1) the outcome of the original service encounter based on specific service attributes, and 2) attributes associated with the service recovery process. Although service organizations strive to offer quality service and get things right the first time, Mattila and Patterson (2004) explained that the hospitality industry is a high human contact industry; thus, failures are inevitable. In 1996, Reichheld stated that the most effective way to building a financially successful organization was to make efforts to retain current customers opposed to continuously obtaining new customers. In an effort to retain those customers, it appears that service organizations now recognize that long- term relationships do not just happen; they are grounded in the organization?s delivery of excellent service, value in the first instance, and complemented by an effective service- recovery system when things do go wrong. Blodgett, Wakefield and Barnes (1995) believed that customer?s selected their retailers or service organization based on service quality; thus, according to their research, getting things right is the best approach to service quality. But unfortunately, mistakes are inherent to the features of the service industry (Ruyter & Wetzels, 2000). Effective service recovery is a key ingredient to gaining an advantage over the competition; thus, it is imperative that service organizations take advantage of 81 opportunities presented to improve upon their service quality. An effective, reliable service recovery process is the best approach to that opportunity. Lovelock?s (2001) Service Failure Response Choice model, presented in Figure 14, demonstrated the customer?s choice in the event of a service failure. In the event of a service failure, customers are presented with numerous opportunities. To begin, customers have a choice whether to report the failure to a manager or employee of the organization or leave dissatisfied. Lovelock (2001) suggested that, at all costs, the organization should encourage the customer to complain. Although service organizations want the customer to lodge a complainant, for the organization to truly benefit, the complaint must be lodged to the organization. Some of the avenues a customer could explore after a service failure could be quite devastating to a service organization. For example, a customer lodging a complainant with the Department of Health for poor cleanliness could create different challenges for an organization to overcome. As outlined in the model, organizations are faced with various opportunities to right the wrong, preserve the relationship and, hopefully, engender loyalty and influence future behavioral intentions. The ultimate goal of any service organization should be for the customer to seek redress from the service organization. This in turn will give the organization an opportunity to correct the problem and avoid the other negative outcomes of a service failure. 82 Figure 15. Service Failure Response Choices Source: (Lovelock, 2001). In 1995, Keaveney found that the effects of switching behavior indicated that service failures and failed service recoveries were a leading cause of customer switching behavior in service organizations. Hart, Heskett and Sasser (2000) revealed that over fifty percent of the customers who do complain are typically dissatisfied with the service organization?s recovery efforts. Unfortunately, a large majority of customers choose not to complain; thus, leaving the organization with intentions to never return and practice negative word-of-mouth advertising. A study by Blodgett, Wakefield and Barnes (1995), found that 77% of all non complainers engaged in negative word of mouth, while conversely, only 48% of complainants engaged in negative word of mouth after attempting to seek redress from the service provider. Thus, research has supported the theory that customers who complain to the service organization are less likely to participate in negative word of mouth activates compared to those that do not complain. Encourage at all costs Dissatisfaction Occurs Take no action Take some action Private Action Public Action Negative word-of- mouth Refuse to revisit Complain to gov.t agency Take legal action Seek redress from organization 83 Service organizations need to recognize that every customer?s problem is an opportunity to prove its commitment to service- even if the provider is not to blame (Hart, Heskett & Sasser, 2000). Halstead and Page (1993) found that the most critical success factor to service recovery was to develop an environment where customers are encouraged to complain. Hart, Heskett and Sasser (2000) suggested that service organizations need to create an environment where customers feel confident that their problem will be solved. They continued by stating that managers and employees should actively evaluate customers to be sure their expectations are being met. Researchers have consistently found that dissatisfied customers who perceive a high likelihood of success are apt to seek redress (Day & Landon, 1976; Richins, 1997; Singh, 1991). SERVICE RECOVERY DEFINED Since the early 1990?s research surrounding satisfaction and service quality has slowly transformed to focus on the customer?s reaction toward service failures and their satisfaction or dissatisfaction with the organization?s approach to service recovery. As enforced in the preceding literature review, an organization?s approach to service recovery satisfaction should be equally important, if not more important, than their overall approach to service quality and doing things right the first time. Research has described service failure as one of the ?pushing determinates? that, if handled improperly, drives a customer?s switching behavior (Roos, 1999); thus effective service recovery satisfaction has become an emerging area of interest in an effort to minimize failures and retain customers. In 1988, Gronroos suggested that a service recovery was an organization?s response to poor quality service. Johnson (1995) enhanced that definition by stating that 84 a service organization must take some form of action and the action must be effectively carried out to reduce the damage in the relationship and to transfer the customer from dissatisfied to a satisfied state. In 2000, McCollough, Berry and Yadav recognized that a successful recovery was an action that returned a customer to a satisfied state with the service provider and the quality of the action could either destroy or enhance the customer?s loyalty and future behavioral intention. Although many definitions have been recognized in related literature, one of the most widely accepted definitions of service recovery is by Zemke and Bell (1990) when they defined service recovery as: ?a thought-out process for returning aggravated customers to a state of satisfaction with the firm after a service or product has failed to live up to the customer?s expectations? (p. 43). Mattila and Patterson?s (2004) research led them to define service recovery as a process which involved actions taken by a service provider to respond to a situation where a failure occurred in the organization?s core or supplementary offerings. Boshoff?s (1999) research acknowledged Zemke and Bell?s (1990) definition but contributed that an effective service recovery process was one that was proactive in nature opposed to reactive. The researcher stressed that a formal process for evaluating the effectiveness of the service recovery process was critical for a service organization. He also distinguished that positive service recovery tactics or strategies were more likely to return customers to a state of satisfaction (Boshoff, 1999). Nonetheless, for some service organizations, the service recovery process tends to be transaction specific; thus, the organization fails to follow a formal service recovery 85 process. Instead, some service organizations tend to respond to each situation in a unique and individual manner. While it is important to make each customer feel like the service recovery process is tailored to meet their needs, Wildes (2005) suggested that in order to achieve optimum levels of customer satisfaction employees must be well trained and empowered to deal with all service situations. Therefore, suggesting that a formal, documented approach to service recovery would aid in the training of employees and in the consistent delivery of service quality. Although an extensive body of literature exists on service recovery (Boshoff, 1997, 1999; Ruyter & Weltz, 2000; Strauss, 2002; Matilla, 2002 & 2003; Mueller, Palmer, Mack, & McMullan, 2003) there appears to be areas of opportunity for additional research and new contributions. For example, there is an emerging need for a systematic approach to track and measure service failures and the overall performance of the service recovery process within the hospitality industry; more specifically, within the luxury segment of the hotel industry. In addition, the mediating roles of emotion and loyalty should be explored in terms of recovery satisfaction, loyalty and the customer?s future behavioral intentions toward the organization. MEASUREMENTS OF SERVICE RECOVERY In 2000, Tax and Brown identified service recovery as one of the key ingredients to developing customer relationships, enhancing loyalty and influencing future behavioral intentions. This in turn, enhanced Abrams and Pease (1993) suggestion that when service failures are successfully resolved, customers may feel a stronger commitment to the organization compared to if no failure had occurred in the first place. Thus, it only stands to reason that in the event of a service failure, the service recovery experience is likely to 86 be the last experience the customer has with the organization. As a result, the effectiveness of the service recovery effort may have a greater impact on loyalty and repurchase intentions (Spreng, Harrell & Mackoy, 1995). Considerable evidence now indicates that the customer?s evaluation of the quality service hinders greatly on the organization?s approach to dealing with service failures. By this rationale effective complaint handling, has a dramatic effect on customer retention rates and employee satisfaction. This in turn directly affects customer?s retention, employee performance issues, negative word-of-mouth and the financial stability of the organization. (Fornell & Wernerfelt, 1987; Tax, Brown and Chandrashekaran, 1998; Smith, Bolton, & Wagner, J, 1999). For example, a TARP study (1996) found that the return on investment for handling complaints far exceeds the consistent return on investment for more traditional investments, ranging from 15 percent for packaged goods organizations to 100 percent for automotive and consumer durable organizations. The results even showed that proper handling of complaints in the banking industry yielded a 170 percent return on investment (Barsky, 1996). In 2001, Mattila evaluated the impact of relationship types on customer loyalty in the context of service failures. The scope of the research was to evaluate the relationship that developed when the customer experienced repeated contact with the same service organization. The study applied Gutek?s (1995) framework of social relationships to gain a better understanding of the optimal level of interaction between a service organization and its customers (Mattila, 2001). Gutek?s (1995) model demonstrated that service encounters can be divided into three conceptually distinct mechanisms for delivering service: true service relationships, pseudorelationships and service encounters. The study 87 found that in the case of a service failure, the customer?s degree of loyalty was significantly higher for true relationship-type customers compared to peudorelationship or service encounter customers. In addition, the study also revealed that the true relationship type received high ratings in terms of the amount of special attention and emotional bonding received from the service organization. The peudorelationship and service encounter customers did not receive the same quality treatment. In 1990, Bitner, Blooms and Tetreault studied the relationships between service failures and employee behavior and what caused customers to distinguish between satisfactory service encounters and dissatisfactory service encounters. Their study revealed that 42 percent of the customer?s dissatisfactory encounters were due to the employee?s response to the service failure. In addition, the study also revealed that when the service failure occurred, customers expected the employee to respond immediately in a friendly, efficient and professional manner (Bitner, Blooms and Tetreault, 1990). Thus, supporting the idea that in order to effectively respond to a service failure, an organization must have a standardized, consistent process in which employee?s are trained to effectively handle the service recovery experience. In addition, research suggests that service organizations should also rely on a measurement approach to ensure consistent service recovery quality. Craighead, Karwan and Miller (2004) explored effective service recovery activities in the food service, auto repair and retail industries. Their model, presented in Figure 15, shows that the antecedents of service recovery, in accordance with the severity of the service failure, directly affect the customer?s perception of the service recovery process. This ultimately led to the customer?s satisfaction or dissatisfaction with the 88 recovery process. Interestingly, their model reviewed the antecedents of service recovery by exploring several elements; 1) the customer?s degree of loyalty prior to the service failure, 2) their perception of quality 3) the severity of the failure, and 4) the service guarantee. In each situation, a fair fix was the most important service recovery factor across all three service organizations. This was followed by sincerity, apology and value. Their findings also revealed that the expectations of loyal customers were more demanding compared to dis-loyal customers. In addition, dis-loyal customers in severe situations only expected a fair fix and were not as demanding of additional compensation. The above study supports the theory that service profits are best bolstered by retaining loyal customers, and there are two primary ways of doing this; 1) through having zero defects, and 2) through an effective service recovery process (Reichheld & Sasser, 1990; Craighead, Karwan & Miller, 2004). 89 Figure 16. Severity of Failure on Service Recovery Satisfaction Source: (Craighead, Karwan & Miller, 2004). In 2006, Kau and Loh explored the effects of justice on service recovery satisfaction and behavioral outcomes in the cellular phone industry. Their research investigated two dimensions. First their research reviewed the relationship between 1) the complainant?s perception of the service recovery process as it related to the justice theories, 2) their satisfaction or dissatisfaction with the recovery process, and 3) customer?s degree of trust, word-of-mouth intentions and loyalty. Second their research also explored the relationship between 1) non-complainants as satisfied or dissatisfied 90 customers, and 2) their degree of trust toward the organization, word-of-mouth intentions and loyalty. The model is presented in Figure 17. Figure 17. Complain vs. Non-Complaint Service Recovery Model Source: (Kau & Loh, 2006). Kau and Loh?s (2006) research is one of several which explored the justice theories in service recovery satisfaction. Kau and Loh (2006) suggested that even if the compensation is favorable to the customer, often times customers feel that they have been unfairly treated. The findings of their research revealed that procedural and distributive justice were the two most important elements in service recovery among cellular customers. Their research also found that, among their respondents, service recovery satisfaction led to high levels of trust and positive word of mouth endorsements; both which are critical in retaining customers. Although their research provided theoretical Cellular Service Providers Non- Complainants Distributive Justice Procedural Justice Interactional Justice Satisfied with Recovery Dissatisfied with Recovery Dissatisfied Non-Complaint Satisfied Customers Trust Word of Mouth Loyalty Trust Word of Mouth Loyalty Complainants 91 and empirical justification linking service recovery satisfaction with loyalty; the impact did not appear to be as strong as anticipated. RECOVSAT Christo Boshoff (1999) explored the relationship between service recovery attributes and the customer?s overall satisfaction with the service recovery process. Boshoff?s (1999) exploratory research began with an instrument of 83 items which were designed to measure the 15 original attributes of service recovery satisfaction. The 15 original attributes were identified based on related literature and the results of the researcher?s qualitative analysis. The 83 items were measured on a 7-point Likert type scale anchored by strongly agree (5) through to strongly disagree (1). A random sample of 2,000 households were selected from a New Zealand based marketing firm. Potential respondents were presented with hypothetical service failures and service recovery scenarios based on events that commonly occur in restaurants and financial institutions. Potential respondents were asked a series of questions regarding their service recovery expectations. Of the random sample, 416 usable questionnaires were returned; a response rate of 20.8% (Boshoff, 1999). The next step in Boshoff?s research was to perform an exploratory factor analysis. As a result, two factors were eliminated based on factor loadings and 13 attributes for service recovery satisfaction emerged. In addition, of the original 83 items, 61 loaded to a significant extent (.40) in the 13 factor solution; thus, based on suggestions by Nunnally (1978) a Cronbach?s alpha of .50 was used as the cutoff point for retentions of factors (Boshoff, 1999). The deletion of items with low correlations did not change the factor structure; thus, the scale was reduced to 54 items (Boshoff, 1999). From there, two 92 additional exploratory studies were conducted. As an end result, Boshoff?s confirmatory factor analysis confirmed a 6 factor structure measured by a 23 item scale. The 6 factor that emerged as drivers of service recovery satisfaction were: communication, feedback, explanation, atonement, empowerment, and tangibles. Boshoff referred to the 6 factor structure as RECOVSAT, which proved to be a valid and reliable instrument. ANTECEDENTS OF RECOVSAT Several antecedents of service recovery (i.e. quality, satisfaction, loyalty, and emotion) have been reviewed in great detail throughout this chapter; however, the antecedents of service recovery identified in Boshoff?s research are critical elements to this research. In consideration of the above, a detailed description will now be presented on the antecedent of RECOVSAT. The final factor structure, presented in Figure 18, was identified by Boshoff (1999) as the attributes of service recovery satisfaction. The attributes are: communication, feedback, explanation, atonement, empowerment, and tangibles. Boshoff (1999, p. 238-239) defined the attributes as follows: ? Communication: the style and manner in which the service employee or manager addresses the customer lodging a complaint. ? Feedback: the quality of the information the service organization provided about the cause of the problem and the resolution process. ? Explanation: whether the service provider explained to the customer why the problem occurred in a clear and concise manner. ? Atonement: value-added compensation to restore the ratio of inputs to outputs by making it up to the customer for the inconvenience caused by the service failure. 93 ? Empowerment: the authority to act and refers to the resources of which employees have access and he decision they are permitted to make. ? Tangibles: the appearance of the service environment (i.e. equipment and physical environment), as well as the appearance and standards of dress for employees. Figure 18. RECOVSAT Source: (Boshoff, 1999). The intention of Boshoff?s (1999) study was to identify several dimensions of transaction-specific service recovery satisfaction by analyzing customer expectations in various segments of the service industry. Additionally, his study sought to develop and validate an instrument to measure satisfaction with transaction-specific server recovery based on the dimensions proposed (Boshoff, 1999). Subsequent statistical analyses conducted during the development of RECOVSAT, led to the elimination of time as a driver of service recovery satisfaction. However, research suggests that a timely response Communication Feedback Atonement RECOVSAT Empowerment Explanation Tangibles 94 to a service failure is imperative to the customer?s overall service recovery satisfaction. Zemke and Bell (1990) found that a swift and effective service recovery enhanced the customers? perception of quality products or services purchased. Miller, Craighead and Karwan (2000) and Simmons (2004) contributed by stating that a successful service recovery depended on how quickly the service organization responded to the problem once they were made aware of the problem. While Boshoff?s (1999) research has contributed to the development of service recovery measurement literature, Strauss (2002) criticized that the measurement had yet to be tested in an empirical nature and had not been tested across industry specific organizations; such as airlines, restaurants or hotels. Significant studies have also contributed to service recovery research; however, a majority of the research revolves around hypothetical service scenarios, using convenience samples, with limited control for heterogeneity. Although hypothetical research is beneficial in academia as well as industry, measurements developed based on hypothetical data solely evaluate the customer?s expectations opposed to their perception. Therefore, there appears to be a gap in industry specific, empirical service recovery research in the hospitality industry. HYPOTHESES AND THE RELATED LITERATURE In addition to the findings in the qualitative analysis, the extensive literature review above has lent justification to the central research hypotheses surrounding this study. The significance of this final section is to introduce the hypotheses that will be analyzed and discussed later in this document. First and foremost, in consideration of the lack of empirical studies on Boshoff?s (1999) RECOVSAT, this study sought to evaluate 95 the psychometric performance of RECOVSAT to confirm or reject the factor structure in the luxury hotel industry. Thus, hypothesis one is stated below. ? H1 - The six factor structure proposed for the RECOVSAT instrument will not be held up in the luxury hotel sector. In response to the exploratory factor analysis conducted during the testing of hypothesis one, factors were identified as being driving factors in explaining customer on-going satisfaction in the luxury hotel segment. Thus, hypothesis two is stated below. ? H2 - Ongoing customer satisfaction will be driven by both ?process? and ?approach? components of the service recovery process. The degree of customer loyalty has been a key constituent in the exploration of service recovery satisfaction, on-going satisfaction and the customer?s attitudinal and behaviroal intentions toward the organization. Thus, the researcher deemed it necessary to explore the following hypotheses. ? H3 - There is a strong positive correlation between a customers expressed loyalty status at the time of the service failure and the degree of recovery satisfaction felt. ? H3a ? The degree of loyalty expressed toward an organization will serve as a mediator to recovery satisfaction. ? H4 - An individual?s level of overall satisfaction will be positively correlated to the degree of loyalty expressed for the organization post- service recovery. As supported in the literature above, a customer?s emotional connection to the organization before, during and after the service failure has a direct reflect on the 96 customer?s attitudinal loyalty and their behavioral intentions toward the organization. This study sought to explore the customer?s emotional state directly after the service failure compared to the customer?s emotional state directly after the service recovery efforts. Since the primary focus of this study are the issues surrounding the service recovery process, this decision was made in an effort to scrutinize the organization?s ability to influence the customer?s post-recovery emotions toward the organization. Thus, lending justification to the fifth and final hypothesis. ? H5: An individual?s degree of emotional satisfaction will be positively related to their overall satisfaction. SUMMARY In conclusion, this chapter has explored the surrounding literature that influences on-going satisfaction with the service recovery process; which in turn, encourages customer loyalty and influences future behavioral intentions. To begin, an extensive review of the literature surrounding satisfaction, quality, loyalty and emotion was presented. All of which are critical components when retaining customers; through both satisfying and dissatisfying service encounters. Next, this chapter presented a detailed analysis of the details surrounding the service failure and service recovery literature. Finally, this chapter concluded with the introduction of the researcher hypotheses that will be analyzed in the pages that follow. While many researchers have contributed to the body of the literature, it is obvious that few studies have examined actual customer perceptions of the service recovery process. This is vital to the development of service recovery literature in the hospitality industry, as well as the service industry. In consideration of the accelerating 97 competition, exceedingly high customer expectations and the duplication of service amenities, higher expectations and demands are being placed on hospitality organizations around the world; especially within the luxury segments. In consideration of the above, there appears to be need for exploratory and confirmatory research aimed at operating to a standard of excellence within the hospitality industry in good times and in bad. 98 CHAPTER III METHODS INTRODUCTION Recent studies (McCollough, Berry & Yadav, 2000; Tax & Brown, 2000; Mattila, 2001; Craighead, Karwan & Miller, 2004; Mattila & Patterson, 2004; Kau & Loh, 2006) have explored various forms of service failures, the organization?s approach to service recovery, and the influence service recovery satisfaction has on FBI. Although research has focused on various service recovery attributes, the effects of service recovery satisfaction on on-going satisfaction, loyalty and future behavioral intentions have yet to be explored in the luxury segment of the hotel industry. In addition, to this researcher?s knowledge, the effects of post-recovery emotion as it relate to on-going satisfaction, loyalty and future behavior has yet to be explored. Therefore, the purpose of this study was to 1) empirically test the performance of RECOVSAT in the luxury hotel segment, 2) explore the relationship between the customer?s degree of pre-failure loyalty and their on-going satisfaction with the organization, and 3) to explore the relationship between post-recovery emotion with on- going satisfaction and 4) to explore the relationship between on-going satisfaction with the customer?s post-recovery loyalty and FBI. RESEARCH CONSIDERATIONS In addition to profitability, the overall goal of most hospitality organizations is to deliver quality service that is designed to engender loyalty and influence future 99 behavioral intentions. Although the hospitality industry strives to meet the customer?s demands and expectations during every service encounter, the combination of the unique characteristics in the service industry (i.e. perishability, heterogeneity and intangibility, and inseparability) and the high human contact environment makes service failures unavoidable. Over the years, the importance of service quality has increased significantly. Accordingly, customers have become more experienced travelers, competition has become more plentiful, and the standards of excellence have amplified. In response to the rapidly changing environment, industry professionals are compelled to mull over the importance of developing a complaint management system as a vital component to their continuous quality improvement efforts. In an effort to improve upon the service offerings, service providers have experienced a paradigm shift from traditional attitudes toward complaint handling. In the past, it was not uncommon that customer complaints, at times, were viewed as antagonistic. Thus, service providers were more inclined to develop a negative opinion toward a customer who voiced their dissatisfaction. Companies were prone to viewing customer complaints in a negative light and some even believed that poor service situations were to be concealed immediately to avoid damaging ramifications. However, service organizations now realize that customer complaints present opportunities for the service provider to identify problem areas and prevent future failures. Tax, Brown and Chandrashekaran (1998) indicated that such systems are a vital component to increasing customer commitment and establishing a long-term loyalty proposition. In relation to this study, the goals of this research were to explore; 1) the selected hospitality properties? approach to service recovery and the overall effectiveness of their 100 efforts, 2) the relationship between the customers? degree of loyalty prior to the service failure and their on-going satisfaction, post-recovery loyalty and future behavioral intentions, 3) to explore the effects of post-recovery emotion on on-going satisfaction and 4) to evaluate the psychometric performance of the six factor structure, RECOVSAT, in the luxury segment of the hotel industry. RESEARCH HYPOTHESES In 1999, Boshoff?s research identified a six factor structure, RECOVSAT, as attributes to service recovery satisfaction in the airline, banking and health care industries. To date, the work of Christoff Boshoff (1997 & 1999) and the development of his RECOVSAT scale have gained wide acceptance and prominence in the academic press. However, Boshoff?s RECOVSAT instrument is a relatively new contribution to service recovery literature and has only been challenged empirically in the banking and finance industry. The result of the empirical research conducted by Boshoff (2003) supported the six factor structure of RECOVSAT. This was not an unexpected finding, considering that banking and finance customers of various institutions participated in the hypothetical service failure studies during the developmental stages of RECOVSAT. The supporting service recovery literature and the discussion during the qualitative section of this research suggested that the six factors presented in RECOVSAT might be mirrored in the luxury hotel industry. Such factors as communication, atonement, honesty, reliability, employee empowerment and professionalism were all topics of considerable discussion during the qualitative stage of this research. However, the addition of various time oriented variables and the numerous modifications required for this study bare reason to believe that Boshoff?s six factor 101 structure for RECOVSAT might not be supported. Thus, Hypothesis one (H1) can be positioned as follows: ? H1 ? The six factor structure proposed for the RECOVSAT instrument will not be supported in the luxury hotel sector. According to the results of Boshoff?s (1999) study, of the six factor structure identified, certain factors explained a larger percentage of the variance and made a unique contribution in explaining service recovery satisfaction. Thus, those factors were identified as the drivers of service recovery satisfaction for Boshoff?s study. This, in turn, led to the justification for Hypothesis two (H2). Therefore, H2 is proposed as follows. ? H2 ? Of the factors identified during the exploratory factor analysis, some factors will make a unique contribution to service recovery satisfaction in the selected luxury hotels. Tax and Brown (2000) identified service recovery as one of the key ingredients to developing customer relationships, enhancing loyalty and influencing future behavioral intentions. This in turn, enhanced Abrams and Pease?s (1993) suggestion that when service failures are successfully resolved, customers may feel a stronger commitment to the firm compared to if no failure had occurred in the first place. Thus, it only stands to reason that in the event of a service failure, the service recovery experience is likely to be the last experience the customer has with the company; as a result, the effectiveness of the service recovery effort may have a greater effect on repurchase intentions (Spreng, Harrell & Mackoy, 1995). Considerable evidence now indicates that recovering effectively from service failures contributes significantly to customer evaluations of the 102 organization?s products and services; therefore, a greater influence on attitudinal and behavioral intentions toward the service organization. As expressed in the literature review, supporting literature provides theoretical support to justify the evaluation of the customer?s degree loyalty and their satisfaction with the service recovery process. Recent studies (McCollough, Berry &Yadav, 2000; Tax and Brown, 2000; Craighead, Karwan and Miller, 2004; Kau and Loh, 2006) have explored the severity of failure on loyalty and future behavioral intentions across the food service, automotive and retail industries combine (Craighead, Karwan & Miller, 2004), while others have explored the comparison between complainants and non-complainants and justice on their degree of trust, loyalty and behavioral intentions in the cellular phone industry (Kau & Loh, 2006). However, to the researcher?s knowledge current literature has yet to explore the relationship between the pre-failure and post-recovery loyalty, emotion and the customer?s behavioral intentions and how the variables relate to on- going satisfaction. Therefore, in an effort to understand service recovery satisfaction from a different point of view the customer?s degree of loyalty prior to the service failure and the customer?s service recovery satisfaction, on-going satisfaction and their post- recovery loyalty and behavioral intentions will be explored. Based on supporting literature, three hypotheses were developed. First, H3 is positioned as follows: ? H3 - There is a strong positive correlation between a customers expressed loyalty status at the time of the service failure and the degree of recovery satisfaction felt. As an extension to H3, the customer?s degree of loyalty as a mediating variable with service recovery satisfaction will be explored. In other words, the customer?s state of loyal or disloyal at the time of the service failure will help to determine the customer?s 103 degree of satisfaction or dissatisfaction with the service recovery process. This hypothesis will be explored based on the findings in Mattila?s (2001) research. She revealed that foodservice industry customers who were truly loyal had significantly higher expectations of the service recovery process compared to those who were disloyal. Her research also revealed that truly loyal customers were not as satisfied after the service recovery process. Therefore, H3a is poised as follows: ? H3a ? The degree of loyalty expressed toward an organization will serve as a mediator to recovery satisfaction. In 1994, Heskett, Jones, Loveman, Sasser and Schlessinger suggested that the weakest link in the Service Profit Chain was the link between satisfaction and loyalty. Reichheld (1996) supported their finding and suggested that the services that satisfy customers may not always be the same services that engender loyalty to service organizations. However, Oliver (1997) stated that although satisfaction does not lead to loyalty, a customer can not be loyal without being satisfied with the overall services and products received. He continued by clarifying that as the customer?s loyalty to a service organization strengthened, the steps necessary to form loyalty (i.e. satisfaction) became less significant. Thus, H4 was proposed. ? H4 - An individual?s level of overall satisfaction will be positively correlated to the degree of loyalty expressed for the organization post-service recovery. Oliver (1981) suggested that emotion constitutes a primary source of human motivation and exerts substantial influence on memory, loyalty and customer behavior. Consequently, in the hospitality industry, where service failures often occur, understanding and controlling a customer?s emotional response to the situation may be 104 the most important element in ensuring overall satisfaction and influencing the customer?s willingness to return and recommend. Smith and Bolton?s (2002) research evaluated a customer?s emotional response to a service failure and recovery by comparing mid-scale hotels to mid-scale restaurants. Their research identified and categorized the most common emotional responses reported among their sample population and determined that certain emotional responses weight more heavily on behavioral intentions. Smith and Bolton?s findings provided the foundation of the emotion scale used in this study; however, in an effort to create a customized emotion scale for the luxury segment of the hotel industry the following emotions were used; anger, discontent, disappointment, self-pity, anxiety, gratitude, happiness and calmness. This led to the justification for the H5. ? H5: An individual?s degree of emotional satisfaction will be positively related to their overall satisfaction. THEORETICAL MODEL The model proposed consisted of eleven main variables. First and foremost, this research sought to test the performance of the six factors identified in Boshoff?s (1999) RECOVSAT as being attributed of service recovery satisfaction. In addition to the six factors, this research also hypothesized that the customer?s degree of loyalty felt toward the organization prior to the service failure was also an attribute of service recovery satisfaction. Next, the theoretical model demonstrates that there will be a relationship between the customer?s satisfaction with the service recovery process and on-going satisfaction. The final two paths on the model suggest that there will be a relationship 105 between on-going satisfaction, post-recovery loyalty and post-recovery emotion. Thus, the proposed model is presented below in Figure 19. Figure 19. Proposed Theoretical Model METHODS OVERVIEW There were several obstacles to overcome during the developmental phases of this research. The most prominent challenge to overcome was the need to administer the questionnaire to a representative sample of the population. In an effort to truly measure service recovery satisfaction and loyalty, in addition to empirically test Boshoff?s (1999) RECOVSAT measurement, the sample population had to consist of 1) customers who Communi -cation Recovery Satisfaction On-Going Satisfaction Post- Recovery Loyalty Tangibles Explanati -on Empower -ment Pre-Fail. Loyalty Atoneme- nt Feedback Post- Recovery Loyalty 106 stayed in one of the selected hotel properties during the previous year, 2) had experienced a failure in service, 3) reported the problem to an employee or member of the management team and 4) progressed through the hotel?s service recovery process and 5) had provided contact information as their willingness to participate in future research. The second most prominent challenge was to determine how to identify customers who experienced a service failure from the general customer data base provided by the hotel. To overcome these dilemmas, all registered guests who qualified though the hotels? system (i.e. guest of selected property, within the past year, and willing to participate in research) received a questionnaire. In addition, the first question on the questionnaire was ?While staying at (The Hotel) have you ever experienced a problem with the services or products you received?? Respondents were given a choice between (1) ?yes?, or (2) ?no? as a response to the question. Respondents who answered ?yes? were guided through a series of questions to evaluate the service failure situation, their perception of the service recovery process, post-failure and post-recovery emotions, pre-failure and post-recovery loyalty, future behavioral intentions and demographic information. Respondents who selected ?no? were guided through a series of questions surrounding their service recovery expectations in the event of a service failure. For the purpose of this study, the respondents who actually experienced a service failure will be analyzed and discussed. In order to effectively measure service recovery satisfaction, loyalty and future behavioral intentions, the hotel properties were selected based on their approach to service and the interrelationship shared with the primary research question. Thus, the properties were selected due to the following 1) their location was conducive to repeat 107 visitors, 2) their ability to attract a variety of business and leisure customers; placing an assortment of expectations on the service providers, 3) the properties? expectations for delivering service excellence as part of their commitment to continuous quality improvement, and 4) the properties? dedication to relationship marketing to engender loyalty and influence future behavioral intentions. In the sections that follow, a detailed analysis will be presented as to the design framework, research sample, the research instrument, and research procedures surrounding the study. QUALITATIVE AND QUANTITATIVE RESEARCH Although predominantly quantitative in nature, the study was preceded by a qualitative component. Leeds and Ormrod (2005) defines these two different research methods in the following way: ?quantitative research is used to answer questions about relationships among measured variables with the purpose of explaining, predicting and controlling phenomena. In contrast, qualitative research is typically used to answer questions about the complex nature of phenomena, often with the purpose of describing and understanding the phenomena from the participants? point of view? (p. 95). The qualitative component of this research was fulfilled in the form of conducting a focus group with several customers staying in one of the selected hotel properties. Although the final instrument administered for this study was modified and enhanced, the core of the instrument was developed based on Boshoff?s (1999) RECOVSAT measurement of service recovery satisfaction. In addition, various measures developed by Zeithaml, Parasuraman, and Berry (1985) along with Shoemaker and Bowen (2003) were modified to measure attitudinal and behavioral loyalty. Smith and Bolton?s (2000) 108 scale for emotion was enhanced in an effort to measure emotion directly after the service failure and directly after the service recovery process. Given that the scales were modified from preexisting measures, a focus group was not completely necessary; however, the qualitative experience and feedback satisfied the face validity requirement and aided in the development and modification of the instruments. Upon completion of the qualitative component and the finalization of the instrument, the next step was to begin with the quantitative analysis. The quantitative analysis began in the summer, 2007. The quantitative research consisted of a cross- sectional study of the effectiveness of service recovery performance within selected hotel properties. Five main constructs were used in the evaluation of service recovery satisfaction; loyalty, emotion, on-going satisfaction, future behavioral intentions and service recovery satisfaction. Although Boshoff (1999) conducted several exploratory and confirmatory factor analyses during the development of RECOVSAT, a factor analysis was conducted to explore the potential relationship between service recovery satisfaction and the added variable of pre-failure loyalty, in addition to testing that the original six variables of RECOVSAT in the luxury segment of the hotel industry. Further quantitative analysis was also applied in the form of reliability and validity testing, univariate analysis, multivariate regression analysis, and structured equation modeling. The above analyses were applied to either support or reject the hypotheses stated in this research. RESEARCH SETTING In an attempt to gather a sufficient sample size, data was collected from five luxury hotel properties across the United States. The hotel properties selected were 109 based on their affiliation with a hotel management company located in the Southeastern region of the United States. Hotels were selected based on three criterion 1) their commitment to service quality, 2) their commitment to service excellence, and 3) the location of the hotel. The questionnaire was administered as a Web-based survey through Survey Monkey; a survey design and management website. The questionnaire was administered to all five of the hotels? simultaneously. In addition to the information sheet required by the Auburn University, Office of Human Subjects, an invitational electronic mail message was delivered to all possible participants. The only specification stated in the electronic message was that all respondents must answer the first question stated. FOCUS GROUP For the qualitative component of this research, one focus group was conducted with 18 hotel customers and was held at one of the participating hotel properties. As stated in a previous section, due to the use of a valid and reliable instrument, supported by previous research, a focus group was not entirely necessary; however, the qualitative research satisfied the requirement for face validity and presented the opportunity to perfect the instrument This qualitative task was accomplished in the early spring, 2007 during the focus group stage of the research. Participants were drawn from one of the participating luxury hotels and were brought together to discuss and refine the instrument to be used. Hotel guests were randomly selected based on the length of their stay at the hotel and the time the guest checked-in to the hotel. Hotel executives were notified of the focus group approximately one week in advance. 110 The Director of Sales at the designated hotel property distributed an invitational letter to all eligible participants, explaining the scope of the focus group and inviting each customer to participate. Customers were asked to volunteer approximately one hour of their time to discuss topics and issues surrounding the quality of service delivered at the hotel. As compensation, The Hotel offered to treat each participant to a free entr?e at their fine dining restaurant located in the hotel as a gesture of appreciation for their time and participation. Focus group participants were brought together in a conference room at the designated hotel. All discussions were recorded, subsequently analyzed and cross- checked against independently recorded notes for accuracy. Although the participants of the focus group were previously informed of the purpose of the meeting, the scope of the study and the purpose of the focus group were discussed. To begin, each participant received an Introduction Letter, (See Appendix A) approved by the Institutional Review Board, and explaining the scope of the study. The letter outlined their responsibilities and rights as participants in this research project. In addition, the letter also ensured their confidentiality and the protection of the relationship between the customer and the selected hotel properties. The initial stage of the focus group was guided by questions surrounding service expectations, failure and recovery expectations and complaint behavior. Several open- ended questions were posed to facilitate discussion about dissatisfaction, complaint management and service recovery expectations. The objective of the open-ended questions was to identify similar expectations as what was outlined and measured in RECOVSAT. Specific indicators such as communication, apology, professionalism, 111 empathy and empowerment were identified. Participants were also asked to highlight those items in the recovery process that were of most importance to their on-going satisfaction. The recording identified several important features that guided the alterations of the instrument. The features are stated below. Upon completion of the open-ended question phase, participants received a copy of the questionnaire for review. As a result of the feedback from the focus group, several changes were made regarding the layout of the questionnaire, wording and clarification of several questions as well as the wording and clarification of the directions that accompanied. Some questions were reworded or combined to scale down the length of the questionnaire. Based on the review of current literature and the results of the open-ended questions, two questions were added in regards to the importance of time in service recovery satisfaction and the questions surrounding loyalty were reworded to measure attitudinal loyalty and behavioral loyalty. The initial intention of the study was to contact possible participants by phone if they failed to respond after a predetermined period of time. However, after much discussion, focus group participants did not agree with that approach. Most focus group participants claimed that a courtesy telephone call would be an invasion of privacy and would be counter-productive; thus, the only form of communication between the researcher and possible respondent was by electronic mail. RESEARCH SAMPLE The selection of the respondents for this study was based on three criteria. First the possible respondents had to be a customer at one of the selected hotel properties within the past 12 months. The second was that possible respondents had to provide the 112 selected hotel with a current electronic mail address to be a recipient of the questionnaire. Finally, possible respondents had to denote their willingness to participate in future research conducted by the hotel management company, the individual hotel property or by an affiliated partner. The survey was administered by electronic mail invitation (See Appendix D) through Survey Monkey, a web-based survey management website. Upon receiving the electronic mail invitation, respondents were asked to select a link embedded in the electronic message. Once selected, the electronic link directed the respondent to the questionnaire. One of the numerous features of Survey Monkey is the management and collection capabilities; thus, upon completion, all responses were automatically organized and stored in a designated folder. From Survey Monkey, the research transferred all responses in to a Statistical Package for the Social Sciences (SPSS) version 16.0. ADEQUACY OF SAMPLE SIZE The number of usable surveys collected for this study was critical to the overall success of the research design and performance of the statistical procedures used to analyze the data. In the design of this research, certain precautions were taken to control for Type I and Type II statistical error. Meyers, Gamst and Guarino (2006) described Type I error, or alpha, as ?a statistical error that occurs when the researcher incorrectly rejects the null, thus, incorrectly claiming that there is a significant group difference or a significant correlation? (p. 38). In order to control for Type I error the researcher sets the alpha level or the acceptable limits for error at .05. Type II error, or beta, is ?a statistical error that occurs when the researcher fails to find an effect that truly exists; thus, the researcher may incorrectly conclude that the group means do not differ significantly or 113 that the correlation is not significantly different from zero? (Meyers, Gamst & Guarino, 2006, p.39). ?A larger sample size or greater statistical power reduces the chances of Type II error? (Meyers, Gamst & Guarino, 2006, p.39). Hair, Anderson, Tatham and Black (1998), explained that Type I and Type II are inversely related; thus, as Type I error moves closer to zero, or becomes more restrictive, chances for Type II error increases. In addition to alpha, there are two additional factors that contribute to the level of statistical power; the effect size and the sample size. As stated above, alpha is the level at which the researcher is willing to be wrong in rejecting the null hypothesis. Meyers, Gamst and Guarino (2006) stated as alpha becomes more restrictive, power decreases; therefore, as the researcher reduces the chance of finding an incorrect significant effect, the probability of finding a correct effect also decreases. In terms of the effect size, Stevens (2002) defines effect size as ?the extent to which the groups differ in the population on the dependent variables? (p.6). The probability of achieving statistical significance is based on the difference of means between two groups or the correlation between variables in the population (Stevens, 2002). Mean differences are stated in terms of standard deviations, so that an effect size of .5 indicates that the mean difference is one-half standard deviation. In terms of the sample size, researchers typically achieve greater power by obtaining a large sample size for the study. Larger samples sizes typically result in increasingly more stable and precise estimates of the population parameters (Meyers, Gamst and Guarino 2006). RESEARCH INSTRUMENT The research instrument (See Appendix C) used to gather data for this study was a Web-based survey. The survey was designed to evaluate the customer?s perception of the 114 service recovery process after he or she had experienced a failure in service. The service failure questionnaire contained 47 questions that were presented in seven sections such as 1) Details Surrounding the Service Failure, 2) Pre-Failure Loyalty, 3) Post-Failure Emotion, 4) Post-Recovery Emotion, 5) Service Recovery Details, 6) Post-Recovery Loyalty and Future Behavioral Intentions, and 7) Demographics. The survey did not contain identifying items that would jeopardize the respondent?s right to confidentiality. A majority of the scale items used in this study were adopted and modified from preexisting measurements. The loyalty and future behavioral intentions sections of the survey were based on measures developed in the 1990 and 1993 studies by Zeithaml, Parasuraman, and Berry and a study in 2003 by Shoemaker and Bowen surrounding relationship marketing and loyalty in the hotel industry. To measure loyalty and future behavioral intentions a pre-failure and post-recovery assessment was administered. Both pre-failure and post-recovery were measured on a 5 point Likert type scale anchored by (1) strongly disagree, through to (5) strongly agree. Questions for pre-failure loyalty were 1) ?I considered myself to be loyalty or would have been willing to be loyalty to the hotel?, and 2) ?I believed the hotel had my best interest at heart?. Questions for post- recovery loyalty were 1) ?I consider myself to be loyalty or I am willing to be loyalty to the hotel?, and 2) I believe the hotel has my best interest at heart?. To measure emotion, measures were adopted from Smith and Bolton?s (2002) scale for emotions; however, their scale was enhanced to include both positive and negative scales of emotion. Emotions listed were Anger, Discontent, Disappointment, Self-Pity, Anxiety, Gratitude, Happiness and Calmness. The 5 point Likert type scale contained a drop-down box for emotions directly after the service failure and the 115 emotions felt directly after the service recovery process. Emotion scales were anchored in (1) extremely calm, through to (5) extremely intense. Items were stated as 1) ?the emotions and intensity felt directly following the service failure were?, and 2) ?the emotions and intensity felt directly following the service recovery efforts were?. Boshoff?s (1999) scale used during the development phases of RECOVSAT was applied to measure service recovery satisfaction. Each scale was measured by a 5 point Likert type scale ranging from (1) extremely dissatisfied to (5) extremely satisfied. SURVEY ADMINISTRATION The survey was introduced to possible respondents by an electronic message and accompanied by an informational letter (See Appendix B) approved by the Institutional Review Board of Auburn University. The selected hotel properties consisted of five luxury properties located around the United States. The survey was administered to possible participants during the summer, 2007. Participants were given 45 days to complete the survey. Approximately two weeks after the initial invitation, one friendly electronic mail reminder was sent. MEASUREMENT OF VARIABLES The measurements of the variables were based on valid and reliable measurements from previous research. RECOVSAT was selected and applied for several reasons. First, RECOVSAT had proven to be a valid and reliable instrument in hypothetical studies across the banking, retail and airline industries. Second, RECOVSAT had never been tested empirically in the hospitality industry. Finally, RECOVSAT was successfully tested empirically in the banking industry. Loyalty, future behavioral intentions, and emotion measurements were derived from their reliable and 116 valid performance in previous studies; which had been previously applied in several service organizations such as retail, airlines, hotels, restaurants, banks and retail service. The measurements surrounding emotion were selected due to the fact that they were specifically developed for the hotel and restaurant industries. In addition, the scale was tested empirically in both industries and also proved to be a valid and reliable in the hotel segment. ETHICAL CONSIDERATIONS The respondents? confidentiality and privacy was of the utmost importance to the researcher. First and foremost, all correspondence between the researcher and the possible respondents were approved by the Institutional Review Board (IRB) at Auburn University (see Appendix A). Policies and guidelines set forth by the IRB were strictly adhered to and all necessary written approval was granted to the researcher prior to the launch of the study. Neither of the surveys contained questions that would permit the researcher to identify the respondents. Upon the completion of this study, surveys were disposed of using the standard disposal methods of sensitive documents approved by Auburn University. SUMMARY In conclusion, this chapter has provided a detailed description of the methods applied during this study. Furthermore, this chapter contained a comprehensive explanation and justification for the creation of the questionnaires and the related literature that assisted in the developmental phases. Finally, this chapter concluded with a thorough description of the research setting, and research sample. 117 CHAPTER IV ANALYSIS OF RESULTS This chapter presents the results of the research in five sections. Section one provides a description of the returned questionnaires and a breakdown of the reporting hotels. Section two provides a detailed breakdown of the sample demographic characteristics and general information regarding the reporting and occurrence of the service failure. Next, a univariate description of each of the measurements used in this study will be presented. Section four will present statistics pertaining to the reliability and validity of all scales employed and section five portrays a sequential analysis of the results pertaining to each of the research hypotheses. All efforts shall be made to separate the reporting of the results to chapter four, while a detailed discussion of these results will be presented in chapter five. DISCUSSION OF RETURNED QUESTIONNAIRES In an attempt to gather a sufficient sample size, data was collected from five luxury hotel properties located across the United States. Presented in Table 1, the majority of respondents (46%) were guests of Hotel Property A with the second largest group of participants (17%) representing Hotel Property B. 118 Table 1. Hotel Participation Hotel Property f % Hotel A 527 46 Hotel B 162 14 Hotel C 182 16 Hotel D 191 17 Hotel E 80 7 The sample for this study consisted of guests who experienced a failure in service while staying at one of the selected hotel properties. A total of 3,519 guests received the Survey Monkey based invitation requesting their participation. Of the total guest database, 1,142 respondents or 32% responded to the questionnaire. Of the 1,142 questionnaires returned, 1,138 questionnaires were valid. Of the total, 553 respondents? or 48.5%, attested to experiencing a failure in service during their stay at their hotel of choice. For the purpose of this study, usable returns were defined as those respondents that experienced a failure in service during their stay at one of the selected hotels. In order to evaluate both the loyalty and emotion variables, questionnaires collected from respondents who experienced a service failure were analyzed and discussed. SAMPLE DEMOGRAPHIC CHARACTERISTICS Presented in Table 2, of the 1,138 respondents to the survey, 553 or 48.5% claimed to experience a service problem during their stay at one of the selected hotels. Over 56% of respondents were male with the remaining 40% being female. The largest age classification was ages 25-34, with almost 34% of respondents falling into this classification. The second largest or 23% fell into the 16-24 age classification. Of the 553 respondents, 30% of respondents were traveling to visit friends and relatives while 24.6% were traveling for leisure. Only 16.3% of respondents claimed to be traveling for 119 business. This was not surprising considering that several of the selected hotels were located in college towns or resort areas. Finally, respondents were asked about the most important factor applied when selecting the hotel of choice. Thirty-three percent claimed they selected the hotel due to a recommendation from a business associate, friend or family member while 22% made their selection based on convenience. Table 2. Demographic Profile of Respondents Gender f % Male 222 40.1 Female 313 56.6 Missing 18 3.3 Age 16-24 129 23.3 25-24 187 33.8 35-44 89 16.1 45-54 65 11.8 55-64 19 3.4 Missing 64 11.6 Purpose of Visit Business 90 16.3 Leisure 136 24.6 Visiting Friends/Family 167 30.2 Other 159 28.8 Missing 1 .2 Important Factor in Selecting Hotel Convenience 122 22.1 Recommendation 201 36.3 Previous Experience 48 8.7 Promotion/Advertisement 56 10.1 Missing 126 22.8 In addition to basic demographics, data was also collected on details surrounding the service failure. In an effort to identify a pattern in communication breakdowns or general problem areas within individual departments, respondents were asked: 1) ?Who was the problem was reported to?, 2) ?the type of problem experienced?, and 3) ?the 120 department within the hotel primarily responsible for the service failure?. As expressed in Table 3, of the total sample size, 45.6% reported the service failure to a front office employee with the second largest category of respondents, or almost 32% reporting the problem to a manager. Over 37% of respondents experienced a problem with the service delivery while over 27% filed a complaint with the quality control issues in the hotel. Finally, respondents were asked to identify the department primarily responsible for the service failure. Over 25% claimed they experienced a service failure with the check- in/check-out process while the second largest category or 24% claimed that the problem existed within the housekeeping department. Table 3. Service Failure Details Reported Problem f % Management 176 31.8 Front Office Employee 252 45.6 Food & Beverage Employee 15 2.7 Housekeeping 14 2.5 Missing 96 17.4 Type of Problem Communication 66 11.9 Quality Control 153 27.7 Service 207 37.4 Technical 70 12.7 Missing 57 10.3 Area Responsible for Problem Check-In/Check-Out 141 25.5 Grounds/Parking 48 8.7 Housekeeping 133 24.1 Personnel 114 20.6 Restaurant 53 9.6 Missing 64 11.6 121 DESCRIPTION OF LOYALTY AND FUTURE BEHAVIORAL DATA Section three entails a univariate analysis of the measurements used in this study. The segment pertains to customer loyalty and the respondent?s future behavioral intentions (FBI) toward the organization prior to the service failure and directly after the service recovery experience. Thus, two sections of the questionnaire pertained to loyalty; presented as pre-failure loyalty and post-recovery loyalty and two sections of the questionnaire pertained to FBI; presented as pre-failure FBI and post-recovery FBI. Both pre-failure and post-recovery loyalty and FBI were measured on a 5-point Likert type scale anchored at (1) strongly disagree; through to (5) strongly agree. Each segment comprised of four variables based on Shoemaker and Lewis (1999) research on relationship marketing in the luxury segment of the hotel industry. Pre-failure FBI consisted of two behavioral questions; 1) ?I would have been happy to recommend the hotel to family, friends and business associates?, and 2) ?If in the area, I would have been happy to revisit the hotel?. In addition, pre-failure loyalty consisted of two attitudinal variables; 1) ?I believed firmly that the hotel had my best interest at heart?, and 2) ?I considered myself to be loyal or would have been willing to be loyal to the hotel?. Post-recovery FBI consisted of two behavioral variables as well; 1) ?I would be happy to recommend the hotel to family, friends and business associates?, and 2) ?If in the area again, I would be happy to revisit the hotel?. Finally, post-recovery loyalty measured the respondents attitude in two variables; 1)?I believe firmly that the hotel has my best interests at heart?, and 2) ?The hotel remains my first choice when visiting the area?. 122 Table 4 presents the mean and standard deviations of each of the pre-failure loyalty and FBI scale items. Table 4. Descriptive Statistics for Pre-Failure Loyalty and FBI Variable Mean Std. Deviation Skewness I believed the hotel had my best interest at heart. I considered myself to be loyal to the hotel. I would have been happy to recommend. I would have been happy to revisit. 3.25 3.81 3.27 3.42 1.20 1.09 1.35 1.25 -.279 -1.03 -.404 -.499 Table 4 illustrates mean pre-loyalty values ranging from m=3.25 for respondents who ?believed that the hotel had their best interest at heart? to m=3.81 for respondents who ?considered themselves to be loyal to the hotel?. Pre-failure FBI values on the other hand range from m=3.27 for respondents who ?would have been happy to recommend? to m=3.42 for respondents who ?would have been happy to revisit?. This highlights a positive picture for the respondents pre-failure loyalty and FBI toward the hotel considering that a score of 2.4 or below on the satisfaction scale denotes below average performance, while 2.5 or above denotes above average performance. Featured in Table 5, the analysis will turn to the respondents? post-recovery loyalty and FBI toward the organization. Table 5. Descriptive Statistics for Post-Recovery Loyalty and FBI Variable Mean Std. Deviation Skewness I believe the hotel has my best interest at heart. The hotel remains my first choice. I would be happy to recommend. I would be happy to revisit. 2.73 2.42 2.69 2.7 1.28 1.29 1.31 1.30 .038 .517 .111 -.028 123 Table 5 illustrates a clear reduction in mean values across all variables. Looking firstly at the attitudinal variables, values range from m=2.69 for ?I would be happy to recommend? through to m=2.79 for ?I would be happy to revisit?. Similarly mean values across both attitudinal variables range from m=2.42 for ?The hotel remains my first choice? through to m=2.73 for ?I believe the hotel has my best interest at heart?. Table 6 indicates that this decrease could not entirely be accounted for by random errors, with a series of paired samples t tests revealing that differences in all three items were found to be significant at the 1% level (p<0.001). Table 6. Paired-Samples t Test for Loyalty Variables Variable Pre-Loyalty Mean SD Post-Loyalty Mean SD Mean Diff. Sig. (2-tailed) I believe the hotel has my best interest at heart. I would be happy to recommend. I would be happy to revisit. 3.30 1.33 3.38 1.26 3.21 1.24 2.69 1.31 2.80 1.30 2.74 1.30 -0.61 -0.58 -0.47 0.001 0.001 0.001 It should be noted that mean values for pre- and post- loyalty may differ somewhat from those reported in Tables 4 and 5, due to the fact that difference scores can only be computed for those cases were respondents have reported both pre- and post- loyalty variables (n=397). Although considerably lower when compared to pre-failure loyalty and FBI, it appears that respondents are still favourably pre-disposed toward the selected hotel group. In an effort to shed further light on this key issue, further analysis was completed comparing respondents? pre-failure and post-recovery loyalty and pre-failure and post- recovery behavioral intentions. As stated above, the instrument presented questions about pre- failure loyalty such as: ?I believed the hotel had my best interest at heart? and ?I considered 124 myself to be loyalty or was willing to be loyal to hotel?. Post-recovery attitudinal loyalty questions were presented as: ?I believe the hotel has my best interest at heart? and ?The hotel remains my first choice?. Table 7 features the pre and post loyalty analysis. Prior to the service failure, over 40% of respondents either agreed or strongly agreed that the hotel had their best interest at heart. To enhance that one step further, over 65% of respondents considered themselves to be loyal or were willing to be loyal to the hotel. When the analysis turned to post-recovery attitudinal loyalty, over 30% of respondents either disagreed or strongly disagreed that the hotel had their best interest at heart. In addition, to the 30%, another 17.5% claimed that they were unsure. Shifting to the question of the hotel remaining the respondent?s first choice, over 40% of respondent?s claimed to either disagree or strongly disagree with an additional 16% unsure. Table 7. Pre and Post Customer Loyalty Variable Pre-Failure Loyalty Post-Recovery Loyalty I believed/believe that the hotel has/ had my best interest at heart. Strongly Agree 13.9% 5.6% Agree 26.8% 17.7% Unsure 22.1% 17.5% Disagree 15.7% 13.0% Strongly Disagree 8.5% 17.5% I considered myself to be loyal or able to be loyal. / The hotel remains my first choice Strongly Agree 25.3% 6.3% Agree 39.8% 9.4% Unsure 13.4% 16.1% Disagree 6.0% 17.7% Strongly Disagree 5.4% 23.1% The analysis then turned to pre-failure and post-recovery behavioral intentions (FBI). The instrument presented questions about pre-failure FBI such as: ?I would have been happy to 125 recommend? and ?I would have been happy to revisit?. Post-recovery FBI questions were presented as: ?I am happy to recommend? and ?If in the area again, I am happy to revisit?. Table 8 features the pre and post FBI analysis. Prior to the service failure, over 40% of respondents either agreed or strongly agreed that they would have been happy to recommend the hotel to family, friends and/or business associates. To enhance that one step further, over 50% of respondents claimed that if in the area again, they would have been happy to revisit the hotel. When the analysis turned to post-recovery FBI, over 32% of respondents either disagreed or strongly disagreed when asked if they were still willing to recommend the hotel to family, friends and/or business associates; while 17.5% said they were unsure if they would recommend. Shifting to the question of if in the area again, I would be happy to revisit, over 30% of respondent?s claimed to either disagree or strongly disagree that they would revisit the hotel; while an additional 17.4% were unsure. However, on a more positive note, over 26% claimed that they would revisit hotel. Table 8. Pre and Post Future Behavioral Intentions Variable Pre-Failure FBI Post-Recovery FBI I would have been happy to recommend. I am happy to recommend. Strongly Agree 20.1% 6.5% Agree 31.6% 16.5% Unsure 18.1% 17.5% Disagree 13.2% 13.0% Strongly Disagree 15.6% 19.3% I would have been happy to revisit. I am happy to revisit. Strongly Agree 19.7% 6.5% Agree 30.9% 20.4% Unsure 19.0% 17.4% Disagree 12.5% 13.2% Strongly Disagree 9.6% 17.7% 126 DESCRIPTION OF EMOTION DATA A continuation of section three presents a detailed analysis of the emotional scales employed. In addition to loyalty, two segments of the questionnaire pertained to the issue of customer emotion in a post-failure and post-recovery sense. The expectation being that there should be a significant correlation between a customer?s emotional status and actual recovery satisfaction. In other words, both scales were designed to attest to the degree to which the hotel group has been able to return a customer to a calm and satisfied state as a result of their recovery process. Presented as post-failure emotion and post-recovery emotion, scales were adopted from Smith and Bolton?s (2002) scale for emotions and modified to account for both positive and negative emotional extremes. All variable scale items were anchored at (1) extremely calm through to (5) extremely agitated. Emotions listed were: Anger, Discontent, Disappointment, Self-Pity, Anxiety, Gratitude, Happiness, and Calmness. Table 9 summarizes the mean and standard deviations of each of the post-failure emotion scale items. By way of interpretation, it is anticipated that where a problem was experienced, customers would rate closer to five across all variables. For example, m=4.00 for ?anger? would indicate a good degree of agitation. Similarly, m=4.00 for ?calm? would indicate that a respondent is anything but calm. On reflection a ?LO (1) through to High (5)? anchored scale may have worked better for this analysis. 127 Table 9. Descriptive Statistics for Post-Failure Emotional Data Variable Mean Std. Deviation Skewness 1. Anger 2. Discontent 3. Disappointment 4. Self-Pity 5. Anxiety 6. Gratitude 7. Happiness 8. Calmness 2.98 2.36 2.11 3.85 3.06 3.35 3.42 3.42 1.13 1.07 1.05 .99 1.23 .93 .97 .98 -.164 .590 .939 -.587 -.012 -.132 -.374 -.490 Table 9 illustrates that a good degree of negative emotion ranging from m=2.11 for ?disappointment? through to m=3.06 for ?anxiety? and m=3.85 for ?self pity? may hint at the existence of a dissonant state of mind or arrest. These results are supported by a low degree of calmness m=3.42, happiness m=3.42 and gratitude m=3.35. The respondent?s post-recovery emotion data is listed below in Table 10. Table 10. Descriptive Statistics for Post-Recovery Emotional Data Variable Mean Std. Deviation Skewness 1. Anger 2. Discontent 3. Disappointment 4. Self-Pity 5. Anxiety 6. Gratitude 7. Happiness 8. Calmness 3.14 2.77 2.59 3.78 3.35 3.28 3.37 3.49 1.14 1.17 1.21 1.02 1.11 1.00 1.00 .93 -.301 -.031 .215 -.613 -.045 -.311 -.452 -.524 Mean values range from m=2.59 for ?Disappointment? to m=3.78 for ?Self-pity? on the negative emotional scales variables and m=3.28 for ?Gratitude? through to m=3.49 for ?Calmness? on the positive emotional scales. Interestingly, while the results indicate a clear increase in four out of the five negative emotions (anxiety experienced a slight decrease) post recovery, they also point to a decrease in two out of the three positive emotions (gratitude and 128 happiness). Table 11 indicates that this change could not entirely be accounted for by random errors, with a series of paired samples t tests again revealing that differences in all eight emotional scale items were found to be significant at the 1% level (p<0.001). It should be noted that while happiness actually increased between failure and post-recovery, this result does not bear out in Table 11. Once again this is due to the number of respondents (n=241) used to calculate this difference score. Table 11 also gives a clear indication of the overall picture concerning respondent emotion in a post-failure and post-recovery sense. Overall respondent emotion is seen to increase from m=2.98 ?post-failure? to m=3.23 ?post-recovery. Once again this difference was found to be significant at the 1 percent level, which is indicative of a rather poor recovery effort on the part of hotel group under investigation. Table 11. Paired-Samples t Test for Emotional Variables Variable Failure Mean SD Post-Recovery Mean SD Mean Diff. Sig. (2-tailed) 1. Anger 2. Discontent 3. Disappointment 4. Self-pity 5. Anxiety 6. Gratitude 7. Happiness 8. Calmness 9. Overall emotional status 2.93 1.14 2.27 1.03 2.01 1.03 3.83 0.99 3.04 1.22 3.28 0.93 3.33 0.99 3.30 0.99 2.98 0.79 3.15 1.14 2.77 1.16 2.58 1.21 3.78 1.02 3.35 1.11 3.27 1.00 3.37 1.00 3.45 0.95 3.23 0.81 -0.22 -0.50 -0.57 0.05 -0.31 0.01 -0.04 -0.15 -0.25 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 Further comparative analysis was conducted on the intensity of emotions and the actual respondent breakdown for each possible response. As stated above, the instrument presented five negative emotions and three positive emotions. Expressed in Table 12, as anticipated, during the service failure, two negative emotions rated extremely high or high with 32% feeling anger and 22% feeling self-pity. As expected the positive emotions rated either low or 129 extremely low intensity. Approximately 8% of respondents felt gratitude and happiness while 7% claimed they were calm during the service failure. Table 12. Emotion Intensity During Failure Variable Extremely Low Low Unsure High Extremely High Anger 7% 22% 11% 28% 4% Discontent 15% 35% 7% 15% 1% Disappointment 22% 33% 7% 9% 2% Self-Pity 1% 3% 18% 2% 20% Anxiety 6% 21% 10% 20% 9% Gratitude 8% 16% 31% 4% 3% Happiness 8% 19% 27% 3% 3% Calmness 7% 25% 19% 8% 3% Table 13 turns to the intensity of emotion felt directly after the service recovery experience, it appears that a larger percentage of respondents felt gratitude (14%), happiness (16%) and calmness (20%). However, these data also shows a large percentage, 21%, 15% and 17% were uncertain about the intensity of gratitude, happiness and calmness, respectably. Even though the positive emotion percentages do not represent a large majority of the sample size, there was a significant change between positive failure emotion and positive post- recovery emotion. However, the above is evidence that respondents? feelings were not overwhelming negative or positive; thus, drawing attention to the quality of the service recovery experience. When evaluating the negative emotions, it appears that the respondents? level of anger and discontent remained relatively consistent across low, unsure and high levels of intensity, while 28% of respondents? claim to feel a high or extremely high degree of disappointment. 130 Table 13. Emotion Intensity after the Service Recovery Process Variable Extremely Low Low Unsure High Extremely High Anger 5% 20% 11% 12% 5% Discontent 3% 14% 11% 15% 8% Disappointment 2% 13% 9% 16% 12% Self-Pity 13% 16% 13% 3% 1% Anxiety 7% 19% 12% 8% 3% Gratitude 3% 5% 21% 14% 5% Happiness 3% 5% 17% 16% 5% Calmness 1% 5% 15% 20% 5% DECRIPTION OF SERVICE RECOVERY DATA The next stage of the analysis was to analyze the service recovery data. Table 14 presents the mean and standard deviation for each variable measured. Respondents were asked to rate 18 variables measured on a 5 point Likert type scale anchored by (1) very dissatisfied; through to (5) very satisfied. Two additional items (19 and 20) were included to measure the customer?s ?overall satisfaction with the products and services of the hotel? and ?overall satisfaction with the hotel?s ability to solve my problem?. Table 14 also includes a mean score (item 21) for overall satisfaction with the recovery effort. This is a mean score calculated from all 18 recovery variables. 131 Table 14. ? Descriptive Statistics for Service Recovery Data Variable Mean Std. Deviation Skewness 1. Quality of communication during recovery. 2. Employee empathy and responsiveness. 3. Solicit relevant information. 4. Employee reliability during recovery. 5. Employee honesty and integrity. 6. Employee knowledge and ability. 7. Time it took to respond. 8. Time it took to resolve the problem. 9. Employee professionalism and discreet. 10. Informed about the problem resolution. 11. Hotel's apology. 12. Consideration of inconvenience. 13. Compensation package. 14. Degree of politeness and respect. 15. Explanation received as cause of problem. 16. Appearance of employee. 17. Professional environment. 18. Hotel did a great job keeping me informed. 19. Overall satisfaction with hotels products and services. 20. Overall satisfaction with the hotels ability to solve my problem 21. Overall recovery satisfaction 2.35 2.45 2.58 2.52 2.58 2.53 2.26 2.10 2.65 2.46 2.34 2.04 2.02 3.09 2.17 3.66 2.67 2.13 2.57 2.35 2.48 1.25 1.28 1.27 1.25 1.22 1.22 1.22 1.22 1.31 1.28 1.35 1.09 1.13 1.25 1.14 1.00 1.22 1.20 1.60 1.21 0.92 .580 .504 .392 .334 .217 .340 .628 .874 .085 .368 .569 .885 .738 -.313 .723 -1.139 .137 .773 .249 .465 .365 Table 14 illustrates mean values ranging from m=2.02 for ?Compensation Package? through to m=3.66 for ?Appearance of Employees?. A total of eight (8) items scored above the average benchmark of m=2.50, which denotes an above average performance in terms of the employee?s professionalism and style in which he/she responded to the situation. At the opposite end of the spectrum, several variables performed below average, or below the benchmark of m=2.40. Variables in that category encompassed quality, timeliness and communication; these variables appear to be strongly related to the process of the complaint management systems in place. Further analysis reveals that respondents were barely satisfied with the hotel?s products and 132 services (m=2.57); that they were less than satisfied with the hotels ability to resolve their problems (m=2.37) and less than satisfied with the hotels recovery process (m.2.48). Further analysis reveals a high degree of correlation between the two later variables; a correlation of 0.575 was found to be significant at the 1% level (n=392; p=0.001). EVALUATION OF VALIDITY AND RELIABILTY Leedy (1993) stated that validity would raise the following questions. What does the test measure? Does it, in fact, measure what it is supposed to measure? How well, how comprehensively and how accurately does the test measure? Thus, for the purpose of this study, the questions are best posed as follows. Does this measure truly identify the most critical attributes of service recovery in the luxury hotel segment? Does this measure truly assess the respondent?s attitudinal and behavioral loyalty to the organization after progressing though the hotel?s service recovery experience? In an attempt to answer these questions, the following section presents an overview of the data available to assess the instrument?s validity. Both content or face validity and construct validity will be reported in this section. CONTENT VALIDITY According to DeVellis (1991), the basic conceptual criterion a measurement scale must meet is content or face validity. Simply stated, an instrument could be considered to be high in face validity if the readability of the measurement appears to measure what it is intended to measure. During the development of Boshoff?s (1999) RECOVSAT, the service recovery scale had undergone the accepted process for scale development; however, modifications were necessary for three reasons: 1) the measurement had to be tailored to the luxury segment of the hotel industry, 2) scales were added to measure pre- 133 failure and post-recovery loyalty, and 3) scales were added to measure post-failure and post-recovery emotion. In addition, the instrument underwent construct or face validity by hotel professionals and guests. This process was applied in reference to Allen?s (1995) view that since the criterion for face validity is the adequacy of items in terms of content domain; review must be by appropriate experts. In summary, the hotel executives and guests made an expert, qualitative judgment that the instrument appeared to be a valid instrument. This qualitative task was accomplished in the early spring, 2007 during the focus group stage of the research. Participants were drawn from the designated hotel property and were brought together to discuss and refine the instrument to be used. Hotel guests were randomly selected based on the length of their stay at the hotel and the time the guest checked-in to the hotel. Hotel executives were notified of the focus group approximately one week in advance. Focus group participants were brought together in a conference room at the designated hotel. All discussions were recorded, subsequently analyzed and cross-checked against independently recorded notes for accuracy. Although the participants of the focus group were previously informed of the purpose of the meeting, the scope of the study and the purpose of the focus group were discussed. To begin, participants were asked to share and discuss their expectations and experiences of service failures and complaint management systems. Participants were also asked to highlight those items in the recovery process that were of most importance to their on- going satisfaction. The recording identified several important features that guided the alterations of the instrument. The features are stated below. 134 ? First and foremost, the majority of respondents felt that, at times, it was too cumbersome to take the necessary steps to register a complaint. They also felt that the service failure had to be a significant problem in order to take the time out of their travel experience to complain. From this a discussion formed around various ways to make it more convenient for a guest to register a complaint. ? Second, as a side bar to the above, a majority of respondents felt that it was too time consuming to register a complaint and wait for a manager or employee to take the necessary steps to correct the failure. There was an overwhelming feeling that the service recovery process should not inconvenience the guest further by having to explain the situation to several employees or managers or by having to wait for the hotel to figure out how to solve the problem. ? Thirdly, emphasis was placed on the fact that employees should be knowledgeable and confident in handling the problem on their own. Although it was recognized that certain problems require a manager?s attention, the majority of respondents felt that employees should be well trained and capable of solving the problem. ? Finally, various attributes previously identified in RECOVSAT were also identified during this session. The general consensus among the group was that managers and/or employees should be genuinely interested in solving the problem. Characteristics such as respectful, empathetic, understanding, and honesty were all mentioned on numerous occasions. The majority also agreed that an apologetic, respectful, friendly approach to solving the problem was more important than compensation of a free room, free meal or discounted service. 135 Upon conclusion of the discussion, all participants were presented with an initial draft of the survey instrument. Participants were given a brief overview of the aims and objectives of the research project. Participants were also given an initial draft of the electronic invitation prepared for distribution with the instrument. Discussion about the instrument included the readability of the instrument, organization of the scale items, and appropriateness of the scale items. Once again, the discussion was recorded and the instrument was revised in accordance with the feedback received. The following items resulted from the discussion: ? First, a majority of respondents recommended placing the loyalty and service recovery scale items at the top of the instrument. Participants felt that more respondents would be willing to complete the survey if the critical information caught the respondent?s attention in the beginning. This lead to a reorganization of the instrument, placing the pre-failure loyalty scales at the top, followed by emotion, service recovery and post-loyalty scales. The instrument concluded with the demographic questions. ? Second, participants were concerned about the complicated item wording, particularly with respect to the use of industry jargon and the detailed description of the scale items. As a result, the item wording and scale instructions and descriptions were restated and simplified. ? Finally, participants were concerned about the number of variables on the instrument and found that several items were measuring similar constructs. Participants felt that this gave the impression that the same question was being stated more than once. The notion was made that this may cause respondents to 136 become aggravated; thus, suggesting that the length of the survey may contribute to a high abandon rate. As a result, several scale items were reworded and clarified, three demographic questions were removed as deemed irrelevant to the study and repetitive loyalty and future behavioral intention variables were eliminated. In summary, several improvements were made to the instrument based on the qualitative analysis gathered during the focus group. In conclusion of the focus group, the agreement was reached that the items included on the instrument were relevant and useful to the domain of service recovery, on-going satisfaction, loyalty and future behavioral intentions in the luxury segment of the hotel industry. The event was concluded approximately two hours later and the participants were thanked for their time and participation. CONSTRUCT VALIDITY According to Cohen, Swerdlik and Smith (1992), construct validity refers to a judgment about the appropriateness of inference drawn from test scores regarding individual standings on a certain kind of variable called a construct, where a construct is described as an informed scientific idea constructed to describe or explain behavior. Principally, the researcher investigating a test?s construct validity must formulate hypotheses about the expected behaviour of high scorers and low scorers on the test. In short, if the test is a valid measure of the construct, the high scorers and low scorers will behave as predicted by the hypotheses. While a number of procedures may be used to provide different kinds of evidence that a test has construct validity, the two principal procedures relate to the provision of convergent and discriminant evidence. In turn, both 137 issues are addressed below in the context of the adapted RECOVSAT instrument utilized in the present study. CONVERGENT VALIDITY According to Leedy (1993), convergence is a means of testing for construct validity, which looks to the focal effect of various methods of measuring a construct and is assessed, in part, when other measures used to measure like-constructs converge (Rubin, 1993). Presented in Table 15, convergence was investigated by calculating the mean score for the post-recovery loyalty scale items and correlating (Pearson?s product moment correlation) this with the mean score for the post-recovery behavioral intention variables. In other words, it is hypothesized that post-recovery attitude towards the organization should correlate with post-recovery behavioral intention. A correlation of 0.908 was found which was significant at the 1% level. Absolute colinearity was found between both post-recovery loyalty and post-recovery future behavioral intentions which failed to meet the requirements for convergent validity of the instrument. Table 15. ? Correlation Matrix of Post-Recovery Loyalty and Post-Recovery FBI Mean Post-Recovery Loyalty Score Mean Post-Recovery FBI Score Post-Recovery Loyalty Pearson Correlation Sig. (2-tailed) N 1.00 387 .908** .000 383 Post-Recovery FBI Pearson Correlation Sig. (2-tailed) N .908** .000 383 1.00 399 **Correlation is significant at the 0.001 level (2-tailed). 138 DISCRIMINANT VALIDITY Discriminability, on the other hand, means that the researcher should be able to discriminate, or differentiate, the construct being studied from other similar constructs (Leedy, 1993). According to Cohen et al., (1992, p.181), a validity coefficient ?showing a statistically insignificant relationship between test scores and/or other variables, with which scores on the test being construct validated, should not theoretically be correlated provides discriminant evidence of construct validity?. The question of discriminant validity necessitated the computation of a further correlation coefficient (Pearson product moment) between respondents? ?Pre-failure Loyalty? to the hotel and the mean overall ?recovery satisfaction? score; calculated post failure. Both variables worked well for this analysis as neither could be said to be related in any significant way given that pre- failure, respondents had no idea that a failure was about to occur and therefore the loyalty expressed should have been independent of any ensuing service recovery effort. The results attest to the discriminant validity of the research measure in that they reveal a statistically insignificant (0.05; p=0.230) correlation between respondents pre-failure loyalty and their mean perception of the hotels recovery effort in a post-failure sense (Cohen et al., 1992). RELIABILITY SCORES OF THE MEASUREMENT INSTRUMENTS The evaluation of reliability of a measurement procedure consists of estimating how much of the variation in scores of different variables is due to chance or random error. According to Billinton and Allen (1996), such reliability measures are necessary in order to test the stability of any measure taken. Of the three traditional methods used to estimate reliability (for example test-retest, alternative form and internal consistency), 139 only the internal consistency method shall be reported on within this analysis. Allen (1995) stated that internal consistency reliability is established by correlating parts of the test with each other. In other words, it will be used to test for the strength of linear relationships between scale items for each of the key constructs included as part of the research instrument. This approach relies upon Cronbach?s alpha; which calculates the average of all possible split-half reliability coefficients. Cronbach?s alpha is viewed as a more expedient indicator of internal consistency and, according to Nunnally (1978) the benchmark that applies should be 0.6 or above. The higher the correlation coefficient between scale items (closer to 1.0), the stronger the linear relationship of the items being correlated and the higher the internal consistency of the research instrument. Table 16 illustrates that internal reliability statistics were calculated for all scales included in the measurement instrument including: pre-failure loyalty attitude, pre-failure loyalty behavioral intention, post-failure emotion, post-recovery emotion, post-recovery loyalty attitude and behavioral intention and the adapted RECOVSAT instrument. Table 16. Internal Reliability Statistics for Measurement Scales Measure ? Statistic # Variable # Cases Pre-failure loyalty Attitude .61 2 470 Pre-failure loyalty Behavioral Intention .90 2 506 Post-failure Emotion .82 8 308 Post-recovery Emotion .84 8 348 Post-recovery loyalty Attitude .96 2 399 Post-Recovery Behavioral Intention .82 2 387 Adapted RECOVSAT .95 18 268 Analysis reveals that each of the scales performed well with co-efficient alpha ranging from ? = 0.61 for ?Pre-failure loyalty attitude? through to ? = 0.96 for ?Post-recovery loyalty 140 attitude?. Scale items measuring the respondent?s loyalty in a behavioral sense ranged from ? = 0.90 ?Pre-failure? to ? = 0.82 ?Post-failure?. Post-failure emotion was recorded at ? = 0.82 , while post-recovery emotion was recorded at ? = 0.84. The adapted RECOVSAT instrument also performed well with ? = 0.95. DIMENSIONALITY OF THE LIKERT-TYPE SCALES While the overriding goal of the research was to explore the relationship between service recovery attributes, on-going satisfaction, customer loyalty and future behavioral intentions and what if any influence loyalty plays as a mediator in the recovery process, it was also deemed essential to test the psychometric performance of each of the scales employed in the study; namely the emotional, loyalty and recovery satisfaction scales. DIMENSIONALITY OF EMOTIONAL SCALES As stated the measurement instrument sought to measure customer emotion across two scales; the first scale was intent on measuring intensity of felt emotion post-failure and the other measuring intensity of felt emotion post-recovery. Scales were adopted from Smith and Bolton?s (2000) scale for emotions and modified to account for both positive and negative emotional extremes. Respondents were asked to rate the intensity of emotion felt on a five (5) point Likert scale anchored at (1) Extremely Calm through to (5) Extremely Agitated. Actual emotions included: anger, discontent, disappointment, self-pity, anxiety, gratitude, happiness and calmness. Previous studies of emotion have proposed a three factor solution encompassing positive, negative and neutral states (Lilijander & Strandvik, 1997; Yu & Dean, 2001) others have uncovered a more easily explained two factor solution encompassing both positive and negative aspects of emotion (Evrard & Aurier, 1996; Hausknecht, 1988; 141 Westbrook & Oliver, 1991). Given the very highly charged nature of the event under investigation and the fact that it relates to a service failure, the researcher believes that the possibility of experiencing a neutral state of emotion is highly unlikely. An exploratory factor analysis using the principal component extraction technique was performed on both emotional scales. The analysis made use of the VARIMAX factor rotation procedure in SPSS version16. A component matrix was initially generated to ensure that the analyzed variables had reasonable correlations (greater than or equal to 0.4) with other variables. Unrotated and rotated component matrices were inspected and all variables were found to correlate well. Looking firstly at the post-failure emotional scale; the result of the corresponding KMO of ?sampling adequacy? was 0.790 and Bartlett?s test for sphericity was 854.920; significant at the level of 1 percent (sig.=0.001). The results of these tests rendered the data factorable and consequently the factor analysis was generated. Table 17 illustrates strong factor loadings (item to total correlations) along two dimensions with coefficient alpha (?) scores ranging from 0.73 (POSITIVE EMOTION) to 0.78 (NEGATIVE EMOTION); explaining approximately 62% of the variance. 142 Table 17. Factor Analysis ? Emotional Scales Emotional Variables Post-failure DIM 1 DIM 2 NEG POS Post-recovery DIM 1 DIM 2 NEG POS 1. Anger 2. Discontent 3. Disappointment 4. Self-pity 5. Anxiety 6. Gratitude 7. Happiness 8. Calmness .737 .845 .801 .514 .634 .864 .861 .552 .887 .917 .888 .437 .672 .878 .912 .559 Eigenvalue % of variation alpha (?) 3.557 1.364 44.460 17.051 0.78 0.73 3.845 1.621 48.063 20.256 0.87 0.71 A further two factor solution (positive/negative) was uncovered for the post- recovery scale. The corresponding KMO of ?sampling adequacy? was 0.782 and Bartlett?s test for sphericity was 884.613; significant at the level of 1 percent (sig.=0.001). The results of these tests rendered the post-recovery emotional data factorable and consequently the factor analysis was generated. Table 17 illustrates a similar two factor structure with coefficient alpha scores ranging from 0.71 (POS) to 0.87 (NEG), which combined explain approximately 69% of the variance. These findings are in keeping with other studies where a similar two factor (positive/negative) structure was uncovered (Liljander & Strandvik, 1997; Yu & Dean, 2001). DIMENSIONALITY OF LOYALTY SCALES Both loyalty scales (pre-failure and post-recovery) were subject to a similar degree of analysis, the results of which are shared in Table 18. Previous research (Dick & Basu, 1994; Oliver, 1997; McMullan & Gilmore, 2003; McMullan, 2005) suggested that the loyalty construct is have two key components: attitude and behavior. It is 143 anticipated, therefore, that this two factor structure should be replicated within the current study. Table 18. Factor Analysis ? Loyalty Scales Loyalty Variables Pre-failure DIM 1 DIM 2 FBI ATT Post-recovery DIM 1 LOYALTY 1. Recommend 2. Revisit 3. Best interests 4. Loyalty attitude .944 .945 .827 .840 .961 .962 .874 .922 Eigenvalue % of variation alpha (?) 1.840 1.433 45.995 35.823 0.90 0.61 3.464 85.597 0.94 Looking firstly at the pre-failure loyalty scale; the result of the corresponding KMO of ?sampling adequacy? was 0.496 and Bartlett?s test for sphericity was 594.170; significant at the level of 1 percent (sig.=0.001). The results of these tests rendered the data factorable and consequently the factor analysis was generated. Table 18 illustrates strong factor loadings (item to total correlations) along two dimensions with coefficient alpha (?) scores ranging from 0.61 (ATTITUDE) to 0.90 (FBI); explaining approximately 82% of the variance. Interestingly the post-recovery factor rotation revealed a one factor solution (LOYALTY) accounting for approximately 87% of the variance. The corresponding KMO of ?sampling adequacy? was 0.843 and Bartlett?s test for sphericity was 1701.339; significant at the level of 1 percent (sig.=0.001). The results of these tests rendered the post-recovery loyalty data factorable and consequently the factor analysis was generated. The resulting one factor solution recorded a coefficient alpha (?) score of 0.94 which does point to the reliability of the scale. 144 DIMENSIONALITY OF THE RECOVERY SATISFACTION SCALE Attention next turned to the dimensionality of the recovery satisfaction measure or adapted RECOVSAT. As previously stated, service recovery literature suggests that understanding the dimensions of service recovery satisfaction is paramount in helping service professionals satisfy customers when things go wrong and building relationships that engender loyalty. The search for a reliable service recovery measure has generated a healthy academic debate over the years, with a variety of measures being put forward. That said, there is little consensus on an effective service recovery measure and even less consensus as it relates to the hotel industry. To date, the work of Christoff Boshoff (1997 & 1999) and the development of his RECOVSAT scale have gained wide acceptance and prominence in the academic press. His research produced an instrument that identified six factors related to service recovery satisfaction in the health care, banking and finance industries. Boshoff?s RECOVSAT instrument is a relatively new contribution to service recovery literature and has only been challenged empirically in the banking and finance industry. The result of the empirical research conducted by Boshoff (1999) suggested a six factor structure called RECOVSAT. This was not an unexpected finding, considering that banking and finance customers of various institutions participated in the developmental stages of RECOVSAT and in the hypothetical service failure studies. The supporting service recovery literature and the discussion during the qualitative section of this research suggest that the six factors presented in RECOVSAT might be mirrored in the luxury hotel industry. Such factors as communication, atonement, honesty, reliability, employee empowerment and professionalism were all topics of considerable discussion during the qualitative stage of the research. That said, 145 the addition of various time oriented variables and the numerous modifications required for this study bare reason to believe that Boshoff?s six factor structure for RECOVSAT might not be supported. Hypothesis one (H1) can thus be posited as follows: ? H1 ? The six factor structure proposed for the RECOVSAT instrument will not be supported in the luxury hotel sector. In order to test and validate this hypothesis, it was deemed necessary to conduct: 1. An exploratory factor analysis of the data across a random selection of respondents to uncover the actual factor structure, and 2. In order to validate this finding, to expose the revealed factor structure to a confirmatory factor analysis for all other responses. Stage one of this analysis exposed 50% (randomly selected) of the respondent data set to a VARIMAX factor rotation procedure in SPSS version 16. A component matrix was initially generated to ensure that the analyzed variables had reasonable correlations (greater than or equal to 0.4) with other variables. Unrotated and rotated component matrices were inspected and variables that did not correlate or correlated weakly with others were excluded (De Vaus, 1996). 146 Table 19. Exploratory Factor Analysis of Service Recovery Satisfaction Attributes Variable Factor 1 APPROACH Factor 2 PROCESS Employee efforts to solicit information Employee reliability during problem resolve Employee knowledge and ability to deal with the situation Employee empathy and responsiveness Hotel?s efforts to keep me informed Honesty and integrity of approach Time it took to respond to the problem The hotel?s apology Employee?s ability to be professional and discreet Quality of communication from employees The appearance of hotel employees Consideration of inconvenience caused The compensation package Hotel did an excellent job of keeping me advised The explanation received as to cause of problem The professional environment Time it took to resolve the problem The degree of politeness and respect The appearance of hotel employees .860 .859 .846 .814 .808 .791 .737 .686 .667 .654 .502 .834 .805 .793 .739 .748 .693 .645 .625 .419 Eigenvalue % of variation Co-efficient alpha 11.445 63.58% 0.95 1.373 7.62% 0.90 Note ? all absolute values less than 0.4 have been suppressed for the purpose of analysis. The result of the corresponding KMO of ?sampling adequacy? was 0.948 and Bartlett?s test for sphericity was 2296.134, which is considered a high Chi-Square, yet significant at the level of 1 percent (sig. =0.001). The results of these tests rendered the data factorable and consequently the factor analysis was generated. Results of the factor analysis are presented in Table 19. This table expresses the item number loading on each factor as well as the extent to which it correlates with each factor. The results illustrate strong factor loadings along two dimensions with coefficient alpha scores 147 ranging from 0.95 (APPROACH) to 0.90 (PROCESS), which combined account for almost 69% of the explained variance. From the analysis, extracted component one is reflective of what might best be described as the more human and relational oriented elements of the service recovery process (APPROACH), while component two is reflective of the more process (PROCESS) oriented aspects of the recovery process. These results demonstrate clearly that the six factor component structure of RECOVSAT proposed by Boshoff (1999) was not supported in the selected luxury hotel properties. Further, this finding also suggested that the issue of recovery satisfaction might be a lot less complicated than first thought. When looked at in greater detail, the two factor structure presented seems to support the notion that customer satisfaction is achieved through a combination of what is done (PROCESS) as well as how it is done (APPROACH). This in clearly in line with the service quality literature and the earlier work of Gummesson (2000) and Gronroos (1988 &1989) amongst others who define service quality in both technical and relational terms. A detailed discussion of the findings and cross-loadings will be featured in chapter five; however, at this time, it should be recognized that items 1, 7, 8, 10, 11, 14, 16 and 17 cross-loaded on two different factors. The degree of difference between the correlation coefficients registered for each of the items was significantly different; therefore, based on supporting literature and factor loadings, factors registering the highest correlation coefficient were selected. Stage two of this section of the analysis required that the two factor structure found as a result of the preceding exploratory factor analysis be exposed to a confirmatory factor analysis in order to confirm the exposed factor structure and whether it held up with the remaining respondent group. This procedure was completed in AMOS 7 and the results, as shown in 148 Figure 19, confirm the existence of the two factor structure uncovered in the preceding exploratory factor analysis. Figure 19. ? Confirmatory Factor Analysis for Service Recovery The model was evaluated by four fit measures (1) the chi square (CMIN), (2) the Comparative Fit Index (CFI), (3) the Root Mean Square Error of Approximation (RMSEA) and (4) the Normed Fit Index (NFI). The CFI and NFI is a measure of relative fit comparing the hypothesized model to the null model with an acceptable value of .95 and .94, respectively (Hu & Bentler, 1999), while the RMSEA, measures the discrepancy between the sample coefficients and the population coefficients with acceptable values below .10 to reflect a good or moderate fit (Loehlin, 2004). While the chi square was found to be statistically significant F1 v10e1 .71 v9e2 .70 v8e3 .61v7e4 .77 v6e5 .81 v5e6 .77 v4e7 .86 v3e8 .84 v2e9 .91 v1e10 .87 F2 v11 e11 .81 v12 e12 .67 v13 e13 .72 v14 e14 .74 v15 e15 .83 v17 e16 .57 v18 e17 .71 v16e18 .43 .74 Chi-square = 377.931; df = 81; p =.001 RMSEA = 0.081; CFI = 0.95; NFI = .94 149 (377.931; p< 0.05), indicating a lack of fit, the analysis revealed a very credible and supportive fit in the CFI = .946, NFI = .94, and RMSEA = .081. All three are indicative of a good fit between the proposed model and the data analyzed. TESTING THE CENTRAL RESEARCH HYPOTHESES Attention shall now turn to section five of the analysis and the testing of the key research hypotheses as set within the original research framework. In short, the data gathered during the quantitative stage of the research process will be scrutinized in order to test for consistency with each of the research hypotheses. Figure 20 presents a revised pictorial path representation of the key research hypotheses and the order in which they will be addressed. This model has been revised from that originally proposed within the theoretical framework section of the study due to the results uncovered in the preceding analysis of instrument factor structure. The model as originally proposed suggested that Boshoff?s six service recovery factors would best explain recovery satisfaction and that this would in some way be mediated by the degree of expressed loyalty customers held towards the hotel group under investigation. The preceding factor analysis however, has proven that in this instance Boshoff?s six factor structure did not hold up; uncovering instead a much less complicated two factor structure encompassing the more technical and/or relational aspects of the recovery process. It is with this in mind that the following revised theoretical model is presented. 150 Figure 20. Modified Theoretical Model APPROACH PROCESS PRELOY OSAT POSTEMOTION POSTLOY 0, e1 1 0, e2 1 0, e3 1 In each case, a short description of each hypothesis will first be offered: HYPOTHESIS H2 The model proposes that a variety of technical (PROCESS) and relational (APPROACH) variables will work to explain recovery satisfaction in the customers mind and that both components of the recovery satisfaction construct will drive ongoing satisfaction with the service organization. Hypothesis H2 is proposed as follows: ? H2 - Ongoing customer satisfaction will be driven by both ?process? and ?approach? components in addition to ?pre-loyalty? during the service recovery process. This hypothesis was investigated by first calculating the mean scores for each of the scale items loading on each of the two factors and correlating (Pearson?s product moment correlation) these with the overall satisfaction scores for all respondents. This correlation H2 H2 H3 & H3a H4 H5 151 was used to give an indication of the direction and strength of the linear association between both variables. The closer the correlation efficient (r) is to 1 or ?1, the stronger the association (+/-) between the variables. Table 20. Correlation Index ? Recovery Satisfaction Components and Ongoing Satisfaction **Correlation is significant at the 0.01 level (2-tailed). Results of this test (Table 20) indicate significant correlations between ongoing satisfaction and all three components of the recovery satisfaction process. All components were found to be moderate or strongly correlated and significant at the level of 1% (p=0.001). Regression analysis was then employed to assess the degree to which recovery satisfaction is explained by all three recovery components and which, if any, is the most important driver of ongoing satisfaction with the company. Ongoing satisfaction was to serve as the sole dependent variable, while the two factors (APPROACH and PROCESS) uncovered during the preceding exploratory factor analysis, in addition to Preloyalty served as the independent variables. The results point to an adjusted R 2 = .51, and F (3, 367) = 141.848, sig. = .000. With just over 51% of the total variance explained, the standardized beta coefficients (APPROACH = -.042, sig. 729; PROCESS = O-SAT PROCESS APPROACH O-SAT Pearson Correlation Sig. (2 tailed) N 1.00 398 PROCESS Pearson Correlation Sig. (2 tailed) N .488** .000 388 1.000 448 APPROACH Pearson Correlation Sig. (2 tailed) N .417** .000 389 .655** .001 425 1.000 434 PRELOY Pearson Correlation Sig. (2 tailed) N .192** .003 243 .209** .001 260 .270** .000 260 152 .526, sig. 000; and PRELOY = .079, sig.275) indicate that the more relational (PROCESS) oriented component of recovery satisfaction is most important in explaining ongoing customer satisfaction. HYPOTHESIS H3 The literature provides strong theoretical support for the idea that the degree of recovery satisfaction experienced will in some way relate to the degree of expressed loyalty felt toward the failing organization at the time the failure took place. Hypothesis H3 may thus be posited as follows: ? H3 - There is a strong positive correlation between a customers expressed loyalty status at the time of the service failure and the degree of recovery satisfaction felt. This hypothesis was investigated by calculating the mean score for each of the two recovery satisfaction components confirmed in the preceding analysis (APPROACH and PROCESS) and correlating (Pearson?s product moment correlation) these this with the mean scores for each of the two pre-loyalty components identified previously (ATTITUDE and BEHAVIOR). Results of this test (Table 21) reveal a rather weak correlation between both recovery components and loyalty as expressed as an attitude. While weak, both correlations were nonetheless significant at the 1% level (p=0.001). Results also reveal that neither component correlated well with the issue of loyalty as expressed in behavioral terms. 153 Table 21. Correlation Index ? Recovery Satisfaction Components and Expressed Loyalty APPROACH PROCESS LOY - ATT APPROACH Pearson Correlation Sig. (2 tailed) N 1.00 460 PROCESS Pearson Correlation Sig. (2 tailed) N .655** .001 425 1.000 458 LOY - ATT Pearson Correlation Sig. (2 tailed) N .241** .001 433 .263** .001 432 1.00 508 LOY - BEH Pearson Correlation Sig. (2 tailed) N .016 .728 454 .052 .268 453 .030 .502 504 **Correlation is significant at the 0.001 level (2-tailed). HYPOTHESIS H3a The literature supports the view that loyalty may act as a mediating variable when it comes to recovery satisfaction. In other words the more loyal or disloyal a customer is to an organization, the more forgiving or unforgiving they may likely be. Hypothesis H3a is thus put forward as follows: ? H3a ? The degree of loyalty expressed toward an organization will serve as a mediator to recovery satisfaction. This hypothesis was investigated by first calculating the mean score for all 18 recovery satisfaction variables and comparing this mean value across two groups assigned as being either ?loyal? or ?disloyal? to the organisation. This loyalty assignment was generated through the use of a single item attitudinal loyalty measure, where respondents were asked to rate the degree to which they considered themselves ?loyal to the hotel?. Response choices were anchored at (1) Strongly disagree through to (5) Strongly agree. For ease of interpretation, all those responding as 1-3 were recoded as ?disloyal?, while those responding as 4 or 5 were recoded as ?loyal? to the hotel. This 154 recoded variable was to serve as the grouping variable for the ensuing Independent Samples t-test. Diekhoff (1992) advocates the use of this test to compare the means of two samples to determine if those means differ significantly. In short, the phrase ?independent-samples? refers to the fact that ?the cases that comprise one sample (disloyal respondents) are independent from, unrelated to, and in no way influenced by the cases that complete the other sample (loyal respondents)?. In this situation the t is used as the test statistic to measure and evaluate the statistical significance of any difference between the means of both sample groups. If the value for two-tail significance is less than 0.05 (p<0.05), then the difference between the means is significant. According to Coakes and Steed (1999), such a level of output would be indicative of the fact that a significant difference exists between the assigned groups. The results of this test as shown in Table 22 indicate that a statistically significant difference exists between the two groups and the degree of recovery satisfaction experienced. Table 22. Independent-samples t Test for loyalty and recovery satisfaction Variable N Mean Std. Dev Sig (2-tailed) Disloyal Respondents Loyal Respondents 124 326 2.34 2.55 0.88 0.95 0.03 0.02 *At p < 0.05, a statistically significant difference was found between both groups These results lend clear support to the view expressed in Hypothesis H3a that the degree of loyalty felt for an organisation will in some way affect the degree to which the 155 customer is willing to forgive the organization following a service recovery process. Both recovery satisfaction means were significantly different between the groups and those customers classifying themselves as loyal appear much more forgiving. HYPOTHESIS H4 The literature suggests clear and direct links between a customer?s ongoing satisfaction with an organization and the degree of loyalty held toward the organisation, whether expressed as an attitude or in the form of future behavioral intention. Similarly, the literature also suggests that the degree of satisfaction held should correlate positively with the customer?s emotional predisposition toward the organization. Hypothesis H4 may thus be stated as follows ? H4 - An individual?s level of overall satisfaction will be positively correlated to the degree of loyalty expressed for the organization post- service recovery. This hypothesis was investigated by calculating the mean loyalty score for each of the four post-recovery loyalty scale items and correlating (Pearson?s product moment correlation) these with a single item measure of overall satisfaction. All four loyalty variables were used in the calculation of this mean due to the results of the preceding factor analysis which found loyalty to be a uni-dimensional construct in a post-recovery sense. Results of this test show a significant positive correlation with the post-recovery loyalty construct (0.875; Sig. = 0.001), thereby confirming the view taken in hypothesis H4. 156 HYPOTHESIS H5 While it is widely accepted by hospitality professionals that emotion plays a crucial role in determining customer satisfaction levels, there has been an almost complete dearth of studies concerning the role of emotion in hospitality satisfaction research (Barsky & Nash, 2002). Further, while many studies have concluded that there is a significant relationship between satisfaction and future behavioral intention, the validity of their findings is now being questioned in that they relate solely to measures of the more cognitive component of the satisfaction construct (Liljander & Strandvik, 1997; Yu & Dean, 2001). Much of this research has been focused around the disconfirmation of some comparison standard or perceived service performance. Satisfaction however, is also believed to contain an affective (emotional) component without which customers responses cannot be fully accounted for (Oliver, 1997; Liljander and Strandvik, 1997). It is therefore hypothesized that: ? H5: An individual?s degree of emotional satisfaction will be positively related to their overall satisfaction. This hypothesis was investigated by calculating the mean score for each of the eight post-recovery emotional scale items and correlating (Pearson?s product moment correlation) these with a single item measure of overall satisfaction. Results of this test show a significant positive correlation with the post-recovery loyalty construct (0.388; Sig. = 0.001), thereby confirming the view taken in hypothesis H5. PATH ANALYSIS OF CENTRAL RESEARCH HYPOTHESES Figure 21 shows the inter-correlations for all constructs at stage two of the research. The proposed model was assessed using AMOS version 16. The results lend 157 further support to certain of the key research hypotheses associated with the model. As with the preceding confirmatory factor analysis, the model was evaluated by three fit measures (1) the chi square (CMIN), (2) the Comparative Fit Index (CFI) and (3) the Root Mean Square Error of Approximation (RMSEA). While the chi square was found to be statistically significant (36.70; p< 0.05), indicating a lack of fit, the CFI was .971 and the RMSEA = .088, indicating a good fit for data and model. Figure 21. Path Analysis of Proposed Relationship The initial results highlighted a number of questionable paths, which led to a re- specification of the model shown above. This re-specification involved the removal of the Post-Emotion variable, in addition to 1) the removal of the path between approach and overall satisfaction, 2) the addition of the path between approach and post-loyalty, and 3) the addition of a path between process and post-loyalty. The respecification of the model lead to numerous favourable paths. The results led to a significant improvement in the fit Indices. Upon making the changes, chi-square (CMIN = 8.68, p=0.01), which was Process Preloy Approach OSAT Postloy .46 .27 .21 .81 .38 e1 e2 .45 .04 .12 158 statistically significant and therefore indicative of a poor fit; however, a slight increase in the CFI to .99, NFI to .98 and a marginal improvement of the RSMEA = .080 all revealed that the model is a good fit of the data. Both sets of results point to the fact that service recovery certainly does drive ongoing customer satisfaction. Additionally, the model confirms the earlier regression analysis which points to the unique contribution played by the softer and more relationship oriented elements of the recovery approach in achieving ongoing satisfaction. Similarly, the results offer further support as to the key links between satisfaction, and the customers? post-recovery loyalty status. A question mark remains however over the more process and/or mechanical aspects of the recovery process, which may actually point to the fact that for most customers it is more about how they are treated. While univariate statistics were provided to support the mediating role of loyalty, the effects of service recovery satisfaction on post-customer emotion, and the relationship between APPROACH and POST-LOYALTY, the results of the path analysis raise a huge question mark over their exact role and influence. SUMMARY OF FINDINGS This chapter has addressed the research questions: What, if any, is the relationship between an organization?s service recovery efforts; subsequent customer satisfaction and customer loyalty? And what, if any, role does perceived loyalty play as a mediator in the recovery process? In so doing, it has addressed a number of central and secondary research hypotheses addressing the issue of recovery satisfaction, its constituent properties, the role of loyalty as a mediating variable and its relationship to ongoing satisfaction and loyalty. The study has shed light on how customers define recovery 159 satisfaction and through a series of statistical tests; it has demonstrated that recovery satisfaction is influenced by the degree of loyalty a customer holds towards an organization. The results also demonstrated that recovery satisfaction is a key driver of ongoing satisfaction and loyalty. In addition, the results also show the important role of the more relational aspect of the recovery process and the fact that above all it is about the manner in which guest problems are attended to and resolved. In all cases, the results were found to be statistically significant. The significance of these results shall now be discussed with respect to their contribution to the literature, their implications for practitioners and the wider academic community. 160 CHAPTER V DISCUSSION AND CONCLUSION RESEARCH OVERVIEW As outlined in the methods chapter, this research consisted of both qualitative and quantitative analysis. To begin, the qualitative research consisted of an organized focus group conducted with a selected panel of hotel guests and executives. The results of the focus group were used to establish a basic understanding of guest?s expectations and/or perception of service recovery; within organizations in the service industry. In addition, the researcher also sought to identify a link between the guest?s service recovery expectations and the expectations identified in previous service recovery measurement research. The focus group was also used to improve the instrument used in collecting data for the quantitative analysis of this research. The quantitative research consisted of a cross-sectional study of the effectiveness of service recovery performance within selected hotel properties. Five main constructs were used in the evaluation of the service recovery experience; loyalty, emotion, on- going satisfaction, future behavioral intentions (FBI) and service recovery satisfaction. Four of the main constructs were measured based on modified scales from previously developed, reliable and valid measurements. The service recovery satisfaction measurement applied Boshoff?s (1999) RECOVSAT measurement, while loyalty and FBI were measured based on Zeithaml, Parasuraman, and Berry (1995) and Shoemaker and 161 Bowen (2003). The emotion scales were based on Smith and Bolton?s (2002) research in hotels and restaurants. In order to support the cross-sectional design of this research, respondents were selected based on three criteria. First the possible respondents had to be a guest at one of the selected hotel properties within the past 12 months. Next, possible respondents had to provide the selected hotel with a current electronic mail address to be a recipient of the questionnaire. And finally, possible respondents had to denote their willingness to participate in future research conducted by the individual hotel property, the hotel management company or by an affiliated partner. Surveys were administered by electronic mail invitation through Survey Monkey, a web-based survey management website. Upon receiving the electronic mail invitation, respondents were asked to select a link embedded in the electronic message. Once selected, the electronic link directed the respondent to the questionnaire. Approximately two weeks after the initial electronic message, respondents who had not completed the survey received a friendly reminder. After the friendly reminder, possible respondents were given two additional weeks to complete the survey. One of the numerous features of Survey Monkey is the management and collection capabilities; thus, upon completion, all responses were automatically organized and stored in a designated folder. From Survey Monkey, the research transferred all responses in to a Statistical Package for the Social Sciences (SPSS) version 16.0 for analysis. This chapter will be divided into five sections. First, a restatement of the research hypotheses and the related findings will follow. Second, the service failure details, pre- 162 failure and post-recovery loyalty and FBI data, and post-failure and post-recovery emotion data will be summarized and discussed. Third, the implications to the academic and practitioner communities will be introduced. Next, the major contributions of the study will be outlined. Finally, this chapter will conclude with a restatement of the limitations and delimitations and recommendations for future research. H1 Christoff Boshoff (1997 & 1999) and the development of his RECOVSAT scale have gained wide acceptance and prominence in the academic press. As thoroughly discussed in the literature review, theoretical framework and analysis sections, his research produced an instrument that identified six factors related to service recovery satisfaction in the health care, banking and finance industries. Boshoff?s RECOVSAT instrument is a relatively new contribution to service recovery literature and has only been challenged empirically in the banking and finance industry. Strauss (2000) suggested, although RECOVSAT performed well in banking and finance, it had yet to be tested empirically in other segments of the service industry. Supporting service recovery literature and the results of the qualitative analysis suggested that the six factors presented in RECOVSAT might be mirrored in the luxury hotel industry. However, modifications were made during the preparation of the instrument for this study. For example, first and foremost, in order to evaluate the relationship between service recovery satisfaction and emotion, loyalty and FBI, numerous scale items were added to correspond with the scope of this study. The next significant modification was that pre-failure loyalty was evaluated as an attribute to 163 service recovery satisfaction. Consequently, it was apparent that Boshoff?s (1999) six factor structure might not be supported. Thus, Hypothesis one (H1) was posited as follows: ? H1 ? The six factor structure proposed for the RECOVSAT instrument will not be supported in the luxury hotel sector. To test the hypothesis, an exploratory factor analysis of the data across a random selection of respondents was conducted to uncover the actual factor structure. In order to validate the findings, analysis exposed the revealed factor structure to a confirmatory factor analysis for all other responses. The results of the factor analyses revealed strong factor loadings along two dimensions with coefficient alpha scores ranging from 0.95 (APPROACH) to 0.90 (PROCESS), which combined explained almost 69% of the variance. Interestingly, the emerging factors are reflective of what might best be described as the human or interpersonal elements of service recovery (APPROACH), while component two is reflective of the technical or formal elements to service recovery (PROCESS). These results demonstrate clearly that Boshoff?s (1999) six factor component structure, RECOVSAT, was not supported in the selected luxury hotel properties. One of the most significant finding also suggested that the issue of recovery satisfaction might not be as complex as expected. Based on the results, it appears that the two factor structure suggests that customer satisfaction is achieved through a combination of what is done (PROCESS) as well as how it is done (APPROACH). This in turn, relates to the findings of Gummesson (2000) and Gronroos (1983, 1988 & 2001) who defined service quality in both technical (what is done) and function (how it is done) terms. At this time, it is important to recognize that several factors cross-loaded as a result of the exploratory factor analysis. Items 1, 7, 8, 10, 11, 14, 16 and 17 cross-loaded on two 164 different factors. Interestingly, the first five items 1, 7, 8, 10, and 11 all directly related to the ?approach?, or in other words, the employee?s intangible role in the service recovery process. Items related to: the time it took to respond, the quality of communication, ability to ask questions, the quality of the apology, and the employee?s professionalism. In further consideration, the degree of difference between the correlation coefficients registered for each of the items was significantly different; therefore, based on supporting literature and factor loadings, factors registering the highest correlation coefficient were selected. The three additional items, 14, 16, and 17; all related to the ?process?, or in other words, that the guest was not further inconvenienced, received an explanation and was informed of the progress. In addition, the degree of difference between the correlation coefficients registered for each of the items was also significantly different; therefore, this decision was also made based on supporting literature and factor loadings registering the highest correlation coefficient. In an effort to confirm, the factor structure analysis then led to a confirmatory factor analysis to see if the findings remained consistent across the remaining respondent group. This procedure was completed in AMOS 7. While the chi square was found to be statistically significant, which denotes a lack of fit, other fit indices indicated by the CFI = 0.95, NFI = .94 and the RMSEA = .081 suggested a good fit between the proposed model and the data analyzed. In summary, H1 was supported, in that the six factor structure of RECOVSAT is not a valid measure of service recovery satisfaction in the selected luxury hotel properties. However, of significant importance, instead a less complicated, two factor structure has been confirmed. 165 H2 The preceding factor analysis revealed that Boshoff?s six factor structure did not apply in the selected luxury hotel properties. Instead, a factor structure emerged which was much less complicated encompassing two factors outlined as the approach or the functional aspect of service recovery and the process or technical aspect of service recovery. The proposed model outlined in the analysis chapter suggested that a variety of technical (PROCESS) and relational (APPROACH) variables will work to explain recovery satisfaction in the customers mind and, of those variables; one may make a unique contribution as the driver of ongoing satisfaction with the service organization. Hypothesis H2 was proposed as follows: ? H2 - Ongoing customer satisfaction will be driven by both ?process? and ?approach? components in addition to ?pre-loyalty? during the service recovery process. This research hypothesis was investigated using two forms of analyses. First, the mean scores for each of the scale items were calculated. Then a correlation matrix (Pearson?s product moment correlation) was conducted using both factors with the overall satisfaction scores for all respondents. Next a multiple regression analysis was employed to evaluate the degree in which each factor explains service recovery satisfaction and to determine if one factor made a strong contribution compared to the others. Just as anticipated, results showed that the Process, Approach and Pre-Loyalty during a service recovery experience were highly correlated or associated with on-going satisfaction. This in turn suggests that the method or the procedure used when correcting the service failure is related to the style and professionalism of the employee or 166 manager?s interpersonal skills and both are directly related to the degree of loyalty the customer felt prior to the service recovery process. All three correlated with on-going satisfaction; suggesting that customers expect a service recovery process that is accommodating, personalized, and pleasant but at the same time, organized and progressive. Although all factors correlated with on-going satisfaction, results from the multiple regression analysis revealed that the service organization?s process to correcting the service failure made a unique contribution; thus, the tangible elements were truly the determining factor in on-going satisfaction. This was a an unexpected finding. Typically, guests at a luxury property place emphasis on professional, accommodating, intangible elements of service; however, it is important to remain mindful that luxury hotel guests may have higher standards and expectations. Thus, as a result, it appears that a guest is not as willing to accept variations in the organization?s formal process and tend to be more demanding of the employee or manager?s approach to solving the problem. H3 The customer?s degree of loyalty prior to the service failure played a critical role in dissecting service recovery satisfaction in this study. Tax and Brown (2000) identified service recovery as one of the key ingredients to developing customer relationships, enhancing loyalty and influencing future behavioral intentions. Suggesting that a customer must be able to evaluate the service provider in good times and in bad. Therefore, the customer?s willingness to be loyalty is based on a combination of satisfying service and a service recovery process that diffuses the problem and restores the customer?s trust and satisfaction. Therefore H3 was proposed as follows: 167 ? H3 - There is a strong positive correlation between a customers expressed loyalty status at the time of the service failure and the degree of recovery satisfaction felt. The loyalty scales used in the instrument were not designed to measure loyalty as attitudinal and behaviour loyalty; however, a two factor structure did emerge during the analysis. Interestingly, previous research (Dick & Basu, 1994; Oliver, 1997; McMullan & Gilmore, 2000; McMullan, 2003) suggested that the loyalty construct comprised two key components: attitude and behavior. Therefore, this hypothesis was investigated by calculating the mean score for each of the two recovery satisfaction components confirmed in the preceding analysis (APPROACH and PROCESS) and correlating (Pearson?s product moment correlation) these this with the mean scores for each of the two pre-loyalty components identified previously (ATTITUDE and BEHAVIOR). Results of this test revealed a weak correlation between both recovery components and loyalty as expressed as an attitude. While weak, both correlations were nonetheless significant at the 1% level (p=0.001). Results also revealed that neither component correlated well with the issue of loyalty as expressed in behavioral terms. The data not supporting this hypothesis was an unexpected finding considering that over 65% of respondents claimed to either be loyal or were willing to be loyal to the organization and over 40% claimed to believe that the hotel had their best interest at heart. Also, over 50% were both happy to recommend and happy to revisit. These findings could be attributed to two fundamentals. First, Oliver (1999) stated that although customers enter into the service encounter with expectations, many customers tend to evaluate a service failure and service recovery process separate from previous service experiences; especially if it 168 was the first time the customer experienced a service failure. Second, this study did not explore the degree of the customer?s loyalty prior to the service failure in great detail. Thus, their loyalty at the time of the failure could be called into question. H3a As an extension to H3, the researcher explored the customer?s degree of loyalty as a mediating variable with service recovery satisfaction. In other words, suggesting that the customer?s state of loyal or disloyal at the time of the service failure would determine the customer?s degree of satisfaction or dissatisfaction with the service recovery process. This hypothesis was explored based on Mattila (2001). She revealed that customer?s in the foodservice industry that were truly loyal had significantly higher expectations of the service recovery process compared to those who were disloyal. Therefore, H3a was: ? H3a ? The degree of loyalty expressed toward an organization will serve as a mediator to recovery satisfaction. In order to differentiate between loyal and disloyalty a single item attitudinal loyalty measure, where respondents were asked to rate the degree to which they considered themselves ?loyal to the hotel?, was used. For ease of interpretation, all those responding as 1-3 were recoded as ?disloyal?, while those responding as 4 or 5 were recoded as ?loyal? to the hotel. Although ?3? was labelled as ?unsure?, all respondents claiming ?unsure? were grouped as disloyal. This decision was based on the ideal that customers who are uncertain of their feelings or attitude toward the organization are, most likely, not familiar with the organization enough to make a determination. The results of the Independent Sample t-test revealed a significant difference between loyal and disloyal state of service recovery satisfaction. Thus, H3a was supported. Results 169 suggested that customers? who considered themselves to be loyal to the organization were more forgiving compared to customers? who were labelled as disloyal. This could be justified in several ways. First, luxury hotel properties tend to place a large emphasis on personalized service; therefore, efforts are put in place to build a strong relationship between hotel guests and hotel employees or managers. In the event of a service failure, service providers that know and understand their customer?s expectations tend to be more successful in solving the problem. Lovelock, Patterson and Walker (2001) stated that ?customer satisfaction evaluations are linked not only to fault- free service, but also to what transpires when something does go wrong? (p. 343). Therefore, the better the organization understands the customer, the better able they are to serve the customer. Also, luxury properties typically spend large sums of money to produce well-trained, customer oriented employees who focus on building customer relationships. Second, McKenna (1991) as well as Shoemaker and Bowen (2003) and Shoemaker and Lewis (1999) revealed that a true emotional bond between the customer and the organization was built on trust and commitment; therefore, customers tend to be more forgiving of ?their? hotel. The third consideration is the customer?s zone of tolerance. Guests who prefer luxury accommodations tend to be seasoned travellers; and a higher degree of expectations may exist. H4 Hypothesis H4 was also related to customer loyalty. In 1994, Heskett, Jones, Loveman, Sasser and Schlessinger suggested that the weakest link in the Service Profit Chain was the link between satisfaction and loyalty. Reichheld and Teal (1996) supported their finding and suggested that the services that satisfy customers may not 170 always be the same services that engender loyalty to service organizations. However, Oliver (1997) stated that although satisfaction does not lead to loyalty, a customer can not be loyal without being satisfied with the overall services and products received. He continued by clarifying that as the customer?s loyalty to a service organization strengthened, the steps necessary to form loyalty (i.e. satisfaction) became less significant. Thus, H4 was proposed. ? H4 - An individual?s level of overall satisfaction will be positively correlated to the degree of loyalty expressed for the organization post- service recovery. This hypothesis was investigated by calculating the mean loyalty score for each of the four post-recovery loyalty scale items and correlating (Pearson?s product moment correlation) these with a single item measure of overall satisfaction. All four of the loyalty variables outlined in the methods section were used in the calculation of this mean. This was due to the results of the preceding factor analysis which found loyalty to be a multi-dimensional construct in a post-recovery sense. Results showed a significant positive relationship between overall satisfaction and the post-recovery loyalty; thus, H4 was supported. This certainly supports previous findings by suggesting that, although loyalty does not always lead to satisfaction, a customer must be satisfied with all elements of the organization in order to truly be loyal. Reichheld (2003) stated that a loyal customer was one who valued the relationship with the company enough to make the company a preferred supplier; regardless of small variations in service. 171 H5 As outlined in the introduction, previous research has primarily focused on hypothetical service scenarios to evaluate the customer?s service recovery expectations opposed to their perception. Although that approach has contributed to the understanding of service recovery satisfaction, the critical element of emotion has not been applicable. Research suggests that without a doubt, customer satisfaction, loyalty and future behavioral intentions are, to some degree, influenced by pre-encounter emotions and post-consumption emotions (Oliver, 1997; Cronin, Brady & Holt, 2000; Barsky & Nash, 2002). Researchers? also recognized that satisfaction contains an affective (emotional) component without which customers responses cannot be fully accounted for (Oliver, 1997; Liljander & Strandvik, 1997). Therefore, H5 was stated as follows: ? H5: An individual?s degree of emotional satisfaction will be positively related to their overall satisfaction. To test this hypothesis, a mean score was calculated for each of the 8 post-recovery emotion scale items and a correlation matrix (Pearson?s product moment correlation) was conducted with a single item measure of overall satisfaction. Results of this test revealed a significant positive relationship between post-recovery emotion and post-recovery loyalty, thereby confirming the view taken in hypothesis H5. The emotion expressed is termed as consumption emotion; which was discussed in the literature review section. To reiterate, a customer associates both positive and negative emotions with the service encounter; which in turn, plays a significant role in the customer?s degree of satisfaction, loyalty and future behavioral intention (Allen, Machleit & Kleine, 1992; Oliver, 1993; Richins, 1997; Barsky & Nash, 2002). This finding suggests that a service organization?s 172 approach to service recovery should return the customer to a calm, satisfied state immediately. As a result, influencing the customer?s positive emotions and defusing negative emotions. SUMMARY In summary, four of the six research hypotheses presented in this research were supported by the resulting analysis. Inherent to these hypotheses is the idea that 1) the concept of service recovery satisfaction is not as complicated as anticipated in the selected luxury hotel properties, 2) in the selected luxury hotel properties, it appears that the customer?s degree of loyalty is a mediating variable to service recovery satisfaction, 3) the customer?s overall satisfaction directly influences the customer?s degree of loyalty after the service recovery process, and 4) the customer?s post-recovery emotions directly effect their overall satisfaction with the service organization. This does in fact lead to a more detailed understanding of the vital role the quality of the service recovery process plays in on-going satisfaction, loyalty and FBI. These findings support Johnson and Gustafsson (2000) findings that quality, customer satisfaction and loyalty form a chain of cause and effect that build on one another and can not be treated or managed successfully as individual segments of the business. PERFORMANCE OF THE MEASUREMENT INSTRUMENTS As described in detail in the analysis section, the construct validity and reliability of the instruments used in this study were all found to be well within the acceptable ranges as prescribed by modern statistical methods. Although all scales performed well, it is necessary to recognize that each scale made unique contributions to the study. The following section will comprise the discussion of these factors. 173 This section will begin with a discussion of the RECOVSAT scale. First, numerous modifications were made to the RECOVSAT scale. Primarily, scales to measure pre-failure loyalty and FBI, post-recovery loyalty and FBI, and emotion were added. Next, based on the qualitative research, scales to measure the importance of time were also added. Finally, the pre-loyalty construct was added as a possible driver of service recovery satisfaction. Consequently, as detailed in the previous section, the qualitative findings of this research, in accordance with the supporting literature, suggested that the six factor structure would be supported in the selected luxury hotel properties. However, as criticized by Strauss (2002), RECOVSAT had not, prior to this research, been tested empirically in service organizations other than banking and finance. Thus, alluding to the fact that the drastic differences among service organizations, paired with the modifications to meet the scope of this study, may alter the performance of the instrument. Interestingly enough, the two factors structure that emerged makes its own unique contribution in understanding service recovery satisfaction in the luxury segment of the hotel industry. The second set of scales applied in this study were the items designed to test loyalty and future behavioral intention. The loyalty and FBI sections of the survey were based on measures developed in 1993 by Zeithaml, Parasuraman, and Berry and a study in 2003 by Shoemaker and Bowen surrounding relationship marketing and loyalty in the hotel industry. Scales were positioned in two locations on the questionnaire in order to denote pre-failure loyalty and FBI and post-recovery loyalty and FBI. The only modification made to the scale items was the wording to reflect past-tense (i.e. pre- failure) and future-tense (i.e. post-recovery). Although the scale items were designed to 174 measure a multi-dimensional construct, a post-recovery uni-dimensional construct emerged. This contribution supports the findings of previous researchers (Dick & Basu, 1984; Oliver, 1997; McMullan & Gilmore, 2000; McMullan, 2003) in regards to pre- failure; however, it appears that the customer?s attitude and behavior intentions toward the organization are viewed as one construct after the service recovery experience. Interestingly, prior to the service failure, a large percentage of respondents claimed to be loyal or were willing to be loyal and felt that the hotel had their best interest at heart. In accordance, a large percentage of respondents also planned to revisit and/or recommend to business associates, family and friends. In reviewing the post- recovery results, loyalty and FBI were viewed as a uni-dimensional construct, as a result respondent?s willingness to be loyal, recommend, revisit and their perception of the hotel?s interest declined drastically. When it comes to the hospitality industry, customers tend to recommend and/or revisit those hotels, restaurants or tourism destinations that leave a positive, lasting impression. Especially when recommending a property to business associates, family or friends, customers tend to be selective in the type of hotel they endorse. Therefore, minimizing the risk of being embarrassed of recommending a property with a sub-standard service philosophy. The emotion scales were adopted from Smith and Bolton (2000) research. The five negative emotions identified in Smith and Bolton?s (2000) research applied; however, three positive emotions were added. In retrospect, a balance between positive and negative emotions should have been applied. While the positive and negative emotions were not balanced, the emotion scale performed similarly to other emotion 175 scales employed by Izard, 1992, Liljander and Strandvik, 1997; Yu and Dean, 2001; and Sui and Baloglu, 2003. Although in retrospect changes could be made, the emotion scales revealed a clear picture of the respondent?s emotional state directly after the service failure and directly after the service recovery process. As anticipated, respondents claimed to experience a vast amount of negative emotions directly after the service failure. While results showed an increase in positive emotions directly after the service recovery experience, a large percentage of respondents were not overwhelmingly pleased with the end result of the recovery. Although respondents no longer claimed to feel anger, discontent, or anxiety; a large percentage claimed to either be disappointed, or unsure of their feelings, while a moderate percentage claimed to be calm, happy or grateful. ACADEMIC IMPLICATIONS One of the goals of this research was to make a unique contribution to supporting literature that will lead to a greater understanding of service recovery satisfaction and the related constructs. In academic terms, this means one of two things. First, that constructs and hypotheses are re-tested and supported or second, a clarified, more elaborate interpretation has emerged. Either way, a better approach to explaining a specific relationship is revealed. The following section will highlight ways in which the results of this study will hopefully add to the current body of literature. This study makes several unique contributions. First and foremost, the research design employed for this study sought the customer?s evaluation of the actual service recovery process. In other words, customers who truly experienced a failure in service had the opportunity to evaluate the service recovery process in place within each of the 176 selected hotel properties. To the best of the researcher?s knowledge, previous studies in the luxury hotel industry have not explored the service recovery satisfaction construct as a response to an actual failure in service. This in turn led to a refined, yet less cumbersome, understanding of service recovery satisfaction and the relationship between loyalty, emotion and FBI post-recovery in luxury hotels. The role of loyalty in service recovery, on-going satisfaction and FBI made another distinctive contribution which will hopefully lead to a clearer understanding in related literature. Although the role of loyalty in a service recovery situation has been explored in the past, this study revealed that prior to the service failure; customers tend to view loyalty and FBI as a multi-dimensional construct and directly after the service recovery process, customers view loyalty as a uni-dimensional construct. To reassess, pre-failure loyalty findings showed that a customer viewed their attitude toward the organization separate from their behaviour intentions (i.e. willingness to revisit or recommend). This was an interesting yet expected finding, considering that previous research (Dick & Basu, 1994; Oliver, 1997) suggested that the customer passes through two distinct phases. First, in order for a customer to be loyal the customer had to pass through the attitude phase and then be willing to take action. One could conclude that although the customer had a positive attitude toward the organization, they may not be willing to revisit or recommend or the customer may choose not to be loyal. Bowen and Shoemaker (1998) outlined reasons why customers are not always willing to be loyal. The reasons discussed were: 1) travellers who do not return to a specific area cannot be loyal to the organization, even if they were willing, 2) when exploring luxury hotels, some hotels customers enjoy a variety of choices and service offerings and do their best 177 to experience as many different organizations as possible, 3) no matter how satisfied, some customers still remain price sensitive; thus will shop for the most reasonable price, and 4) some customers just expect to receive quality service and be satisfied every time. Therefore, some customers do not view quality and satisfaction as reasons to earn their loyalty. However, upon completion of the service recovery process, results showed that loyalty and FBI were uni-dimensional. This could lend to the fact that the customer has experienced the service provider in a negative light; thus, their perception of the service quality either meets their approval and earns their loyalty and FBI or fails all together. In regard to emotion, the findings correspond quite well with previous emotion research. It appears that post-recovery emotion plays a pivotal role in the customer?s on- going satisfaction and their attitude and behaviour intentions toward the organization. Therefore, the need to further explore the role of emotion in academia is vital to dissecting the service recovery construct within various service organizations. Price, Arnould and Deibler (1995), indicated that the customer is typically engaged in a service encounter; therefore, the customer has a time and emotional investment in that interaction. In retrospect, if given the opportunity to evaluate emotion, the research design should assess emotion directly after the event. Research lends evidence that customer satisfaction, loyalty and future behavioral intentions are, to some extent, influenced by the customer?s emotions prior to the service encounter and directly after the service encounter (Oliver, 1997; Cronin, Brady & Holt, 2000; Barsky & Nash, 2002). Unless the emotions felt during the failure and recovery experience were extremely positive or extremely negative, the lapse of time between experience and the evaluation, 178 could call the emotion results into question. Therefore, the researcher is of the opinion that emotion measurement should occur as close to the failure and recovery experience as possible. The final contribution to this study is the mediating role of loyalty in on-going satisfaction. Results showed that customers who considered themselves to be loyal at the time of the service failure were more forgiving and satisfied with the service recovery process when compared to disloyal customers. Although the researcher hypothesized that loyal customers were more likely to be satisfied, this was a pleasant, and somewhat unexpected, finding. This lends evidence to several of the researcher?s ideas. First, a strong relationship between the customer and the service provider does aid in the quality of service the customer receives. Second, customers who share an emotional attachment (i.e. trust and commitment) to the service organization are more willing to help the organization improve upon their service offerings (i.e. the customer may be willing to voice their service recovery expectations to help the organization solve the problem). Finally, Reichheld (2003) stated a loyal customer was one who valued the relationship with the company enough to make the company a preferred supplier; regardless of small variations in service. PRACTITIONER IMPLICATIONS In an effort to explore the importance of service recovery satisfaction, emotion, loyalty and FBI in the luxury hotel industry, several important fundamentals emerged as a result of this research. This section will highlight how the findings of this research can be applied in the industry. 179 First and foremost, the two factor structure for service recovery satisfaction clearly indicated that customers expect a balance between the processes employed as well as the approach taken. Additionally, between the two, process and approach, results clearly showed that the customer placed higher expectations on the employee or manager?s people skills. Therefore, suggesting that the intangible elements in the luxury industry are more critical in the service recovery process. Although at first glance this structure seems less cumbersome, it is important to recognize the hefty expectations that lie within each factor. Various elements must meet the customer?s approval for satisfaction to occur within the ?approach?. For example, the ?approach? in service recovery satisfaction related to: 1) the quality of communication and apology, 2) the time it took to respond and solve the problem, 3) the employee?s ability to a) solve the problem, b) be respectful and empathetic and c) be knowledgeable, honest and reliable. Lofty customer expectations were present in regards to the ?process? of service recovery as well. For example, the ?process? related to 1) the customer was no longer inconvenienced, 2) the compensation package was acceptable, 3) the environment, as well as the employees, had a professional appearance, and 4) the customer was informed of the progress being made. The results of this section provide concrete evidence on ways in which luxury hotels can improve upon their service recovery offerings. Primarily, standard operating procedures and training manuals can be structure as the intangible or ?approach? and tangible or ?process? elements to achieve service recovery satisfaction. Furthermore, the various elements within each factor require a commitment to continuous quality 180 improvement, on-going employee training, evaluation, coaching and constant feedback and consistent, timely customer evaluations. The role of loyalty explored in this research is significant in the luxury hotel segment. Clearly the loyalty results justify the importance of building relationships between the customer and the service provider. To reiterate, loyalty results showed that customers who considered themselves loyal were more satisfied with the service recovery process compared to those who were disloyal. It is the opinion of the researcher that the better the organization understands the customer, the better the organization can serve the customer. Therefore, relationships between the customer and service provider tend to influence personalized, more accommodating service; in both positive and negative encounters. Also, results showed the uni-dimensional evaluation of loyalty after the service recovery experience. By that rational, the quality of the service recovery must leave a positive, lasting impression on the customer for the he or she to be willing to revisit, recommend and feel that the organization is worthy of their loyalty. In industry, this issue could be addressed as another justification for 1) constantly improving upon the employee or manager?s approach to failed service; as well as the formal process employed, and 2) consistent evaluation of the customer?s perception of the recovery process. The evaluation of emotion is another area offering industry importance. As a result of this study, it is apparent that post-recovery emotion truly plays a role in on-going satisfaction with the organization. Thus, it only stands to reason that a customer feeling a positive post-recovery emotion would be more inclined to feel more trust and/or 181 commitment toward the organization. While the evaluation of emotion made a significant contribution, in an effort to truly understand the role emotion plays in service recovery, practitioners should evaluate the customer?s perception of the service recovery experience and related emotions immediately following the encounter. SUMMARY OF THE CONTRIBUTIONS In summary, this study made several important contributions in both the academic and practical areas. Although the contributions were discussed above, this section will present a brief summary. First, the testing of a previously development service recovery measurement was applied. This was based on a recommendation by Strauss (2002) where he suggested that RECOVSAT should be tested empirically in various organizations within the service industry. Although RECOVSAT did not apply in the selected luxury hotel properties; a structure emerged that was unique to the selected hotel properties. While the critical elements suggested in RECOVSAT were still present, it appears that guests in luxury hotels view the service recovery process as the intangible and tangible elements surrounding the event. This finding certainly clarifies the relationship between the service recovery attributes and satisfaction in the luxury hotel industry. Plus, it may also aid in the development of employee training programs and continuous quality improvement efforts. The need to understand loyalty and its role in service recovery satisfaction is exceedingly important in hospitality. As stressed in the previous pages, the hospitality industry in destined to be constantly challenged by competition, price wars, duplicated services, and service failures. Thus, one of the most effective ways to triumph over the 182 challenges is to create an environment which engenders loyalty through personalized, accommodating services that are tailored to meet the customer?s expectations. Finally, the role of emotion signified the importance of emotion in the customer?s evaluation of service recovery satisfaction. As outlined in the related literature section, there are numerous components to the measurements of loyalty and emotion that may be burdensome to an organization; however, as disclosed in this study, the loyalty and emotion constructs are necessary measures to truly understanding service recovery satisfaction. FUTURE RESEARCH This research opens the door to numerous areas of opportunity for future projects. First, this study can be replicated across other luxury hotel properties, as well as other segments of the hotel industry. It would also be a valuable contribution to academia and industry to explore service recovery satisfaction in a similar fashion within the restaurant, airline and cruise line industries. Next, as stated earlier, it is in the opinion of the researcher that the emotion construct could be more effectively measured if the evaluation was distributed immediately following the service encounter. This would enable the researcher to explore the customer?s emotional state prior to the service failure as well as gather a more accurate analysis of post-failure and post-recovery emotion. In addition, although the emotion scale performed admirably in this study, the emotion scale applied in future research should be balanced between positive and negative emotions in an effort to not influence the results in anyway. 183 Another area of opportunity for future research deals with one of the delimitations to the study. One of the most significant delimitations to this study was the role of the employee in the service recovery process. Employees and managers play a pivotal role in customer service, satisfaction, engendering loyalty and influencing future behavioral intentions toward the organization. In consideration, the employee and manager?s attitude and opinion toward complaint behavior, the attributes surrounding service recovery, and their perception of customer relationships could further clarify the service recovery satisfaction construct. Finally, research suggests that well-traveled or seasoned guests have a different service expectation compared to guests who are not well traveled. Therefore, it would be of interest to the researcher to explore the differences in service recovery expectations between seasoned travelers and undeveloped travelers. CONCLUSION In conclusion, this chapter provided a discussion of the findings, contributions, and future research considerations which materialized as a result of this study. Each hypothesis was highlighted, the associated results were revealed and a detailed discussion followed. From there, the performance of the instruments applied during the course of this research was revealed. Afterward, a discussion of the academic and practical implications from this research was presented. This chapter concluded with a discussion of areas of opportunity for future research. 184 REFERENCES Abrams, M., & Pease, M. (1993). Wining and dining the whiners. Sales and Marketing Management, February, 73-75. Adams, J.S. (1965). Inequity in social exchange, in Berkowitz, L. (Ed.), Advances in Experimental Social Psychology. (pp. 267-297). New York, NY: Academic Press. Albrecht, K., & Zemke, R. (1995). Service America!. New York: Warner Books, Inc. Allen, C., Machleit, T., & Kleine, S. S. (1992). A comparison of attitudes and emotions as predictors of behaviour at diverse levels of behavioral experience. Journal of Consumer Research,18, 493-504. Andreassen, T. W. (1999). What drives customer loyalty with complaint resolution? Journal of Service Research, 4, 39-49. Barksy, J. (1996). Building a program for world class service. Cornell Hotel and Restaurant Administration Quarterly, 37(1), 17-27. Barsky, J., & Nash, L. (2002). Evoking emotion: affective keys to hotel loyalty. Cornell Hotel and Restaurant Administration Quarterly, 43(2), 39-46. Bejou, D., & Palmer, A. (1998). Service failure and loyalty: An exploratory empirical study of airline customers. Journal of Services Marketing, 12(1), 7-22. Bell, S. J., Auh, S., & Smalley, K. (2005). Customer relationship dynamics: Service quality and customer loyalty in the context of varying levels of customer expertise and switching costs. Academy of Marketing Science Journal, 33(2), 169-183. Bell, C. R., & Zemke, R. E. (1987). Service breakdown: the road to recovery. Management Review, 10, 32-35. Berry, L. L., Wall, E. A., & Carbone, L. P. (2006). Service clues and customer assessment of the service experience: Lessons from marketing. Academy of Management Perspectives. 20, 43-57. Berry, L. L. (1995). Relationship marketing of services: Growing interests, emerging perspectives. Journal of the Academy of Marketing Sciences, 23(4), 236-245. 185 Berry, L. L., Zeithaml, V. A., & Parasuraman, A. (1985). Quality counts in services, too. Business Horizons, May-June, 44-52. Billinton, R. & Allen, R. N. (1996). Reliability evolution of power systems. (2 nd . Ed.). New York, NY: Plenum. Bitner, M. J., Booms, B. H., & Mohr, L. A. (1994). Critical service encounters: The employee?s viewpoint. Journal of Marketing, 58(10) 95-106. Bitner, M. J., Booms, B. H., & Tetreault, M. S. (1990). The service encounter: Diagnosing favorable and unfavorable incidents. Journal of Marketing, 54(1), 71- 84. Blodgett, J. G., Granbois, D. H., & Walters, R. G. (1993). The effects of perceived justice on complainants? negative word-of-mouth behavior and repatronage intentions. Journal of Consumer Research, 69(Winter), 399-428. Blodgett, J. G., Wakefield, K. L., & Barnes, J. H. (1995). The effects of customer service on customer complaining behavior. Journal of Services Marketing, 9(4), 31-42. Bloemer, J. M. M., & Kasper J. D. P. (1995). The complex relationship between consumer satisfaction and brand loyalty. Journal of Economic Psychology, 311- 329. Bloemer J. M. M., & de Ruyter K. (1999). Customer loyalty in high and low involvement settings: The moderating impact of positive emotions. Journal of Marketing Management, 315-330. Bolton, R. (1998). A dynamic model of duration of the customers relationship with a continuous service provider: The role of satisfaction. Marketing Science, 17, 45- 65. Boshoff, C. (1997). An experimental study of service recovery options. International Journal of Service Industry Management, 8(2), 110-130. Boshoff, C. (1999). RECOVSAT ? An instrument to measure satisfaction with transaction-specific service recovery. Journal of Services Marketing, 1(3), 236- 249. Boshoff, C., & Staude, G. (2003). Satisfaction with service recovery: Its measurement and its outcomes. South African Journal of Business Management, 34(3), 9-16. Bourne, L. E., & Russo, N. F. (1998). Psychology: behaviour in context. New York: W.W. Norton and Company, Inc. 186 Bowen, J. T., & Shoemaker, S. (1998). Loyalty: A strategic commitment. Cornell Hotel and Restaurant Administration Quarterly, (39)1, 12-25. Bureau of Labor Statistics (online) (cited 20 May 2006). Available from