Quality measurement and analysis of web based information systems

Introduction CHAPTER INTRODUCTION This research focuses on quality measurement and analysis of web-based information systems (WIS). In this introductory chapter, the evolutionary development of information systems and the service quality assessment in the previous research are described to give the research background of this current study. The objectives and scope of this dissertation are then highlighted. 1.1 An Overview of the Service Quality Assessment in Information Systems 1.1.1 Evolutionary Development of Information Systems Information systems (IS) used in an organization are very critical for its successful operation (Eriksson & Torn, 1991), which lead to higher productivity, better customer service, and better quality of working life (Iivari & Ervasti, 1994). Since their first introduction in the 1960s, the role of information systems has undergone a drastic change in organizations, as shown in the literature and practice. In the 1960s and the 1970s, IS were used as tools for the data processing. In the 1980s they evolved to support managers’ needs to make better decisions. In the early 1990s, IS were used to support business goals of organizations and help to create competitive advantages, which is a crucial role of the strategic planning in decision support systems (DSS). In line with the above discussion, Ward et al. in 1990 classified IS into three eras, Introduction namely, Era of Data Processing (1960s), Era of Management of Information Systems (1970s and 1980s), and Era of Strategic Information Systems (1990s). However, since 1993 with the birth of the World Wide Web (WWW), the role of IS has been widely broadened. A new era called Web-based Information Systems (WIS) has been paid much attention in the literature (for example, Lindroos, 1997; Lederer et al., 1998; Tenenbaum, 1998). WIS facilitate the intra- and inter- business process and system integration. WIS enable e-business and e-commerce to flourish. Companies are rushing to use the Web to enhance their competitiveness. Expenditures on WIS are at an ever-increasing rate. Despite the huge expenditure, WIS have often failed to support business effectively, in particular to meet web users’ needs. How to analyze and improve the quality of WIS comes to be a critical issue. WIS should be strategically managed. 1.1.2 Determinants of Information Systems Success In the literature, there are three important issues in any IS, that is, products, processes, and services. Information products delivered to users include the software and hardware systems, user documentations, training courses, etc. Information processes creating information systems refer to system analyses, technical designs, program codes, final handover tests, and so on. Information services deal with answering questions, solving problems, addressing concerns and aspirations of end users, and so forth (Whyte & Bytheway, 1996). Introduction Studies into the determinants of IS success have been conducted from the perspective of products, processes, and services. The measure of IS success is a multi-dimensional construct. In DeLone and McLean’s (1992) IS success model, six measures of IS success are identified, namely, system quality, information quality, use, user satisfaction, individual impact, and organizational impact. The basis of the D & M IS success model is product-oriented (Pitt et al., 1995). ‘System quality’ measures technical success; ‘information quality’ measures IS outputs; ‘use, user satisfaction, individual impacts’ and ‘organizational impacts’ measure system effectiveness. However, with the emergence of end user computing in the mid-1980s, especially with the development of WIS, the role of IS department within the organization has changed. It has been broadened from product developers and operations managers to information and service providers. Recent research has agreed that service quality measure is a part of IS success (For example, Kettinger & Lee, 1995; Pitt et al., 1995). The updated D & M IS success model (2003) has integrated service quality into their conceptual model. 1.1.3 Achieving High Quality in Customer Service High quality in customer service is the key factor leading to business success in today’s competitive world. The delivery of high service quality has been increasingly considered as a prerequisite for success in the marketplace. However, people’s awareness of the importance of quality and the exploration of formal methods for quality control and improvement is an evolutionary process (Montgomery, 1996), especially in the domain of service quality. Unlike product quality, which can be measured objectively by such indicators as the number of defects/nonconformists, service quality is an abstract and elusive construct due to the following three unique characteristics: Introduction (1) Intangibility, where a mismatch can occur between the service quality perceived by the information producers and consumers (Parasuraman et al., 1988). (2) Heterogeneity, where the performance of service delivered may vary from user to user, system to system, and from day to day (Kettinger & Lee, 1995). (3) Consumer participation, where service consumers will participate in the consumption of the service. In the absence of objective measures, developing an appropriate approach to assess service quality has raised a great deal of interest among researchers, from both theoretical and empirical perspectives. The SERVQUAL model is one of the most widely used models in the service quality literature. The SERVQUAL model is based on the concept of service quality gaps (Parasuraman et al., 1988). In this dissertation, service quality is defined as a function of the gap between customers’ expectations of a service and their perceptions of the actual service delivered. Over the past two decades, the SERVQUAL model has been modified to suit the service nuances of specific industries, e.g., healthcare, tourism, information systems, festivals, and the automobile industry, etc. Since there are numerous differences between the traditional one-way communication channels and the new two-way communication via web-based exchanges, the SERVQUAL model developed for the traditional customer service should be modified to better suit the context of web-based customer service. Research on this modification should be useful for web designers and information service providers (ISP). It will also make some contributions to the information systems literature. Introduction 1.2 Objective The proposed research focuses on quality measurement and analysis of web-based information systems. Since the 1980s, the quality revolution in manufacturing had a profound impact on the competitiveness of companies. Nowadays the service industry is beginning to understand that quality does not improve unless it is measured. The main objective of this research is to develop a strategic and quantitative approach to identify elements of superior web-based service quality, to measure online customer satisfaction, and to analyze the determinants of high service quality for continuous and strategic improvement. Using the SERVQUAL model as a starting point, some conceptual and methodological issues involved in the management of web-based service quality are discussed. This research will provide guidance for practitioners on how to assess webbased service quality, how to implement service quality measures, and finally how to achieve online customer satisfaction. To achieve the stated objective, the research topics in this thesis include: (1) To explore the differences between web-based customer service and traditional communication channels, then identify the shifts in service quality dimensions in the context of this new information age. (2) To propose a general framework to assess web-based service quality, to implement measures, and finally to achieve service quality, from the perspective of web users instead of the system development related to information technology (IT). Introduction (3) To conduct a worldwide and online survey to develop a parsimonious measurement instrument for web-based service quality with sound psychometric properties in terms of reliability and validity. (4) To conduct a campus-wide survey to understand users’ perceptions of web-based service quality. A variation analysis is to be carried out in terms of web users’ experience. (5) To identify existing problems and limitations of the well-established service quality instruments, such as the SERVQUAL model that conceptualizes a linear and symmetric relationship between the service quality gaps and the overall service quality. This research is to investigate the asymmetric and nonlinear nature of this relationship. (6) To seek rigorous and data-driven approaches to develop robust constructs on which to categorize web users. In this dissertation, new approaches to user pattern identification and categorization are to be proposed. (7) To highlight some opportunities for further development and in-depth research on web-based service quality measurement, analysis, and improvement. Some suggestions on the strategic information systems planning (SISP) in WIS are to be proposed for continuous improvement. Introduction 1.3 Scope of the Research The proposed research focuses on quality measurement and analysis of WIS. The relationship among various concepts is configured in Figure 1.1 as below: Business Interface Web Users User Information Satisfaction Quality Measures Affect Interaction User Feedback Decide Information Technology Affect Affect Information Process Support Affect Decide Support Information Service Affect Behavioral Intentions Affect Web-based Information Systems Figure 1.1 Concepts and Relationships in this Dissertation The contents of the main body of this thesis and the relationships among chapters are described in Figure 1.2. Introduction Chapter Identification of Shifts in Web-based Service Quality Chapter A General Framework to Develop Quality Measures Chapter Case Study on Managing Web-based Service Quality Chapter Prioritization of Service Attributes Chapter Case Study on Variation Analysis of Web Users Chapter A Fuzzy Set-based Approach to Web User Categorization Chapter Using Alternating Least Squares Scaling in Web User Categorization Web-based Service Quality Improvement Web User Satisfaction Figure 1.2 Scope of this Research The whole dissertation is divided into three parts. Part I includes Chapter 2, which is a literature review of the service quality assessment in web-based information systems, from both theoretical and empirical perspectives. This brief overview on the existing literature Introduction provides the research background for the current study. This introductory chapter also pinpoints the limitations in the existing literature, which elucidates the motivation for the current study. Part II comprises five chapters (Chapters - 7). This part deals with both a conceptual and operational framework to manage web-based service quality. After an exploration of the differences between web-based and traditional customer service, Chapter notes that the SERVQUAL model, developed for the traditional customer service, might require adaptation for the use in this information age. Confirmatory Factory Analysis (CFA) is implemented in Chapter to serve the purpose of identifying the shifts in service quality dimensions. Chapter proposes a general framework as an effective and operational procedure to assess web-based service quality, to implement measures, and finally to improve service quality. Field studies are reported in Chapters and 6. Chapter describes an extensive empirical field study conducted online and worldwide. It focuses on developing a parsimonious measurement instrument with sound psychometric properties. Chapter intends to understand users’ perceptions of web-based service quality, regarding a local university as a campus-wide information system. This study not only develops a measurement scale which gives a good indication of users’ perceptions of web-based service quality, but also further conducts a variation analysis in terms of web users’ experience. Chapter is to develop an operational procedure that would prioritize customer service attributes in a simple, inexpensive, and accurate manner. This study discusses problems with the well-established instruments such as the SERVQUAL model in the prioritization of service attributes for improvement. By addressing the existing problems, Chapter investigates the asymmetric and nonlinear nature of the relationship Introduction between the service quality gaps and the overall service quality. A model is developed to advance utility theory into the prioritization at the service attribute level as well as at the service dimensional level. In the service literature, this is a first attempt at integrating utility theory into the prioritization of service attributes to help achieve quality in customer service. Part III includes Chapters and 9. The focus of this part is to propose two new approaches to web user identification and categorization. User-centered information systems represent the recent effort of delivering information more effectively in the modern electronic age. It is important for ISP to accommodate individual differences in web users and to tailor the web content according to the needs of a particular user or a set of users. This is in order to achieve system applicability, efficiency, and effectiveness. Although user profiling is a principal element of web personalization, surprisingly, not enough attention has been paid to it either in intranet design methodology (IDM) or in Internet commerce development methodology (ICDM). Most of the earlier approaches in the IS literature tend to model web users in terms of broad linguistic terms. This has resulted in the use of inconsistent theoretical constructs and limited applicability across studies. Chapter presents a data-driven approach to web user categorization with the help of fuzzy set theory. A multi-attribute structure is developed to enable fine distinctions among web users. Chapter focuses on developing a user categorization process model by using the alternating least squares scaling (ALSCAL) method for categorical data. In this model, numerical quantifications can be assigned to each category of the categorical variable. The predefined measurement levels are adjusted accordingly. Hence the optimal scale can be defined for each categorical variable. Both methods discussed in Chapters 10 Using Alternating Least Squares Scaling in Web User Categorization that when choosing the number of dimensions, a useful guideline is to keep the numbers small enough so that meaningful interpretation is possible. In this example, the twodimensional solution accounts for 87.2% of the variance as shown in Table 9.3. A third dimension probably would not add much more information. 9.5 Discussion and Managerial Implications The use of ALSCAL with optimal scaling features in the categorization of web users has minimized the problem in dealing with categorical data with mixed measurement scales. We now present some findings from the illustrative example and summarize some strategic managerial implications. 9.5.1 Predefined Obvious Measurement Scale and Adjusted Optimal Scale Categorical data are pervasive in opinion studies such as market segmentation, political parties, or social groups. In many studies, the method of specifying an optimal scale is essentially ‘ad hoc’. As mentioned previously, there is no intrinsic property on which a variable should be automatically predefined as an optimal scale. This chapter developed a process model for the categorization of web users. The obvious measurement scale might not be the best. The present research has identified a set of effective variables of the mixed measurement scales that can serve as principle components to clustering web users into different groups. Based on transformation plots, the predefined measurement scale for each variable is checked, and adjusted to obtain the optimal scaling for each variable if necessary. This makes interpretations easier and more meaningful. 201 Using Alternating Least Squares Scaling in Web User Categorization 9.5.2 Numerical Quantifications Assigned to the Categories of Each Variable With the development of ALSCAL for categorical data, numerical quantification can be performed on the categories of each variable. Categorical principal component analysis using ALSCAL is a major step in the process model of categorization. In this step, categorical variables are quantified, while the dimensionality of the data set is reduced. The goal of this step is to reduce an original set of variables to a smaller set of uncorrelated components that represent most of the information found in the original variables. Interpretations of fewer variables would be easier than doing the same on a larger number of variables. After assigning quantification to the categories of each variable, the relationship among categories can be analyzed numerically. Nonlinear relationships between categories might be modeled. In early works, the relationship is assumed be linear. In practice, this may not hold true. In this chapter, the relationship can be visualized using ALSCAL by transformation plots. The transformation plot suggests adjustments to the predefined measurement scale. The optimal scale is then found for each categorical data. 9.5.3 Some Managerial Implications Some strategic implications of this study are summarized below: (1) Previous research (Gnaho, 2001; Graja & McManis, 2001) has found that the most difficult part of web user categorization lies in dealing with categorical data and representing them on a numerical scale. It becomes more complex when the available information includes categorical data of mixed measurement scales. With the help of ALSCAL in the categorization model, the present research has 202 Using Alternating Least Squares Scaling in Web User Categorization developed a method to classify web users into categories. This method can also be applied in other types of categorical data studies (e.g., opinion surveys). (2) User-centered information systems represent a recent effort to deliver information more effectively in the modern electronic age. Due to the vast and rapid growing breed of web users, it is very critical for information providers to identify typical web users, and typical user characteristics. User profiling is a principal element in web personalization. Web personalization can be defined as the process of customizing the content and structure of a web site to the specific and individual needs of a particular user or a set of users. It takes advantage of the knowledge gained from the users’ navigational behavior and individual interests (Mulvenna et al., 2000, Eirinaki & Vazirgiannis, 2003). This chapter has demonstrated a procedure to investigate the relationship between individual differences in web users and user online behavior. This information can be used in identifying typical user categories. Information providers then can tune the presentation of WIS contents and structures according to the specification of the user categories, and define guidelines to achieve user information satisfaction. This is one objective of web personalization, which is currently known as the key to success in business today and in the future. 9.6 Conclusion Modeling user profiles is one way to investigate the relationship between individual differences in web users and user online behavior. This study developed a process model 203 Using Alternating Least Squares Scaling in Web User Categorization of web user categorization by using ALSCAL with optimal scaling features. In this model, numerical quantifications can be assigned to the categories of each variable. Thus the relationship between categories can be analyzed without being presumed to be linear as in previous studies. If non-linearity exists, transformation plots can be used. Adjustment to a pre-defined measurement scale could be made. This allows for a more accurate examination of the relationship between individual differences in user characteristics and their online behavior. The final goal is to capture user profiles, to identify user patterns, and to categorize web users. In early work on user profiling, the determination of a measurement scale given a particular data type, is essential ‘ad hoc’. The method developed in this chapter has made a step in the direction of data-driven approaches to web user profiling. This makes it possible for web designers and information service providers to adopt appropriate priorities in dealing with a variety of web users, with the final purpose of achieving quality in customer service. 204 Conclusion CHAPTER 10 CONCLUSION In this concluding chapter, discussion and conclusion of this dissertation are given. Specifically, compared with early work in the literature, some major contributions of this current research will be highlighted. Due to the limitations of this study, some suggestions and recommendations for future research are also covered in this chapter. 10.1 Concluding Remarks As mentioned in the introductory Chapter 1, the main objective of this dissertation is to develop a strategic and quantitative approach to identify elements of superior web-based service quality, to measure online customer satisfaction, and to analyze the determinants of high service quality for continuous and strategic improvement. This dissertation has addressed the following two key questions: (1) How to identify elements of superior web-based service quality and measure online customer satisfaction? The background of this current research and a literature review on the service quality assessment of WIS were presented in Part I. Part II focused on developing a conceptual and operational framework to manage the web-based service quality. Service quality dimension shifts in the information age were identified in Chapter 3. A general framework to develop a better service measure scale was proposed in Chapter 4. Finally two empirical case studies were reported in Chapters and 6, respectively. 205 Conclusion (2) How to analyze the determinants of high service quality for continuous and strategic improvement? Chapter advanced the utility theory into the prioritization of service attributes and resulted in an operational procedure that would prioritize customer service attributes in a simple, inexpensive, and accurate manner within the constraint of limited resources. To further the whole scope of this study, Part III focused on developing two new approaches to user pattern identification and categorization for personalizing WIS. User profiling is crucial to tailor the content and structure of WIS to the needs of a particular user or a set of users. Therefore, usability and IS effectiveness can be achieved. A multi-attribute structure was developed in Chapter to enable fine distinction among web users with the help of fuzzy set theory. Chapter presented a process model of web user categorization in which numerical quantifications could be assigned to the categories of each variable and predefined optimal scaling levels could be adjusted accordingly. 10.2 Contributions of this Study Various topics discussed in this dissertation are organized in an integrated structure, with the goal to develop a strategic and operational framework to measure and improve webbased service quality. Based on those discussions, some major contributions of this study, compared with early work in the literature, are summarized below: This dissertation has explored differences between web-based and traditional customer services. It then proposed shifts in service quality dimensions accordingly. CFA was used 206 Conclusion to test alternative models to determine which has the best fit. One important finding of this is that there is no single solution to the question of what dimensions must be managed in order to achieve user satisfaction of web-based customer service. This research has demonstrated that it is possible to analyze these differences. This makes it possible for web designers and service providers to adopt appropriate policies and priorities in dealing with web-based customer service. Instead of simply applying the well-developed SERVQUAL model to the new context, we used it as a starting point and made modifications accordingly. The results from CFA have proved that the shifts exist in quality dimensions largely due to the differences between web-based and traditional communication channels. After exploring the shifts in service quality dimensions for web-based service using CFA, there is a need to develop a general framework for web-based service quality measurement, which possesses good psychometric properties in terms of reliability and validity. This information will provide guidance for practitioners to adjust this general framework according to the specific questions that need to be addressed in the practice. Following the developed general framework and using the conceptual e-SERVQUAL model as the basic structure of web-based service quality domain, modification of those dimensions should be taken in account after the exploration of differences between webbased and traditional customer services. An extensive world wide empirical study was carried out online. A six-dimensional service quality measure was generated. Especially, this case study found out that the dimension Quality of Information was important in the measurement scale in terms of high item-to-total correlation for each item in this 207 Conclusion dimension and high Cronbach’s alpha of this dimension. Today web sites, which are useful, timely, accurate, and rich in detail, become an integral part of our lives. Those attributes in turn determine whether and how frequent surfers will return to the web sites. Since the purpose of web-based service is to meet the needs and requirements of web users, determinants of web-based service quality should be discussed from the perspective of web users. Studies on how to develop quality measures from web users’ perspectives and how to implement those measures to support business effectively are desirable and would benefit both practitioners and researchers. This dissertation has addressed those critical conceptual and measurement issues. Due to the heterogeneity of web users, this dissertation has conducted a variation analysis in terms of user experiences along three characteristic attributes. A campus-wide case study was carried out. The developed scale and analysis are helpful for ISP to set management policies and guidelines to achieve user perceptions of web-based service Another major contribution of this current study is on the development of an operational procedure that would prioritize customer service attributes in a simple, inexpensive, and accurate manner. The ability of a company to achieve excellence in service quality depends on the determination of service attributes and their desired levels. It also depends on the prioritization of service attributes, using appropriate quality improvement indices, in a consistent manner within the constraints of limited resources. By addressing the problems resultant from using the SERVQUAL gaps in quality improvement, this study has investigated the asymmetric and nonlinear nature of relationship between service quality gaps and the overall service quality. Instead of point 208 Conclusion estimation widely discussed in the service literature as the standard practice, customers form expectation distributions based on their cumulative experiences with service attributes. The single-level hierarchical Logit model has been applied to advance the Sshaped utility function in the strategic prioritization of service quality both at the attribute level and the dimensional level. In the service literature, this is a first attempt at integrating utility theory into the prioritization of service attributes to help achieve quality in customer service. Another important contribution relating to variation analysis of user heterogeneity is on two new approaches to user pattern identification and categorization for WIS personalization. Focusing on user profiling, a multi-attribute structure is developed with the help of fuzzy set theory. The model offers the means to assess the categorization process methodically, and to understand explicitly the linkage between attribute components and overall dimensions. This approach provides diagnostic information at the attribute level as well as the dimensional level. This study has demonstrated how to analyze and accommodate those individual differences in WIS design and personalization. Note that user profiling often means to distinguish users in terms of broad linguistic terms, which are presented in categorical data. A research problem occurs that how to select the optimal measurement level for the categorical data. It is because there is no intrinsic property that can automatically predefine what measurement scale should be specified for a categorical variable. The obvious measurement scale might not be the best. One can explore the data in any way that makes sense and makes interpretation easier using the ALSCAL method with optimal scaling features. Chapter presents a process model using the ALSCAL method for categorical data, in which numerical quantifications can be 209 Conclusion assigned to the categories of each variable. Both of the new approaches have made a step to analyze individual differences in web users, capture user profiles, and finally cluster web users into different categories of each posited variable, so that ISP can accommodate individual differences in web users to achieve system applicability, efficiency, and effectiveness in WIS personalization. 10.3 Limitations of this Study WIS are new with a short history. Previous studies have found varying results with respect to some issues. However, other issues have not yet been addressed. The complexity inherent in this new information system makes the theoretical studies more difficult. Arguments are more than questions to be put down. It has resulted in the use of inconsistent theoretical constructs and an inability to interpret across studies. This also could build barriers for the applications of those approaches proposed in this study in a real world. The emphasis of this dissertation is on the identification of superior elements of webbased service quality, and how this should be managed by ISP. Domain knowledge on Web and IP technology has a major influence on the measurement and analysis of WIS, but is treated superficially in this dissertation. Some efforts are definitely desired on this aspect. The dimensions and attributes in service quality are generic in nature. Those parameters can be translated into technical parameters according to definition approaches derived by those standardizing bodies, such as ITU, IETF, and ETSI. However, practitioners with domain technologies such as SOAP protocol for XML, WSDL, and 210 Conclusion UDDI, can use the framework developed in this dissertation as a guideline to design the system configuration and to improve service quality according to customers’ requirements. The methods proposed in this current study could provide an indication of potential important factors emerging to be related to web-based customer service. However, this could be ineffective in particular circumstances, due to variations from one situation to the next. It is important to be able to identify variations that can occur. It is the responsibility of practitioners to adjust the framework according to the specific questions that need to be addressed. Specific guidance for practitioners will play an important role in the applications of those approaches to achieve web-based service quality. The dynamic nature of service quality evaluation in WIS should be paid enough attention. Customers’ requirements and the levels of preferred service performance would change with time. With the advances in Web and IP technology and changes in circumstances of the company or the individual, the preferred service quality may change with time. For effective management of WIS, it is necessary to identify when these change take place and be able to monitor WIS. An integrated static and dynamic model should be built for strategic service quality implementation. 10.4 Suggestions for Future Research This dissertation has attempted to discuss several issues involved in the quality measurement and analysis of WIS for strategic and continuous improvement. Major 211 Conclusion findings and contributions have been summarized as above. Nevertheless, due to the limitations involved in the current study, much more research needs to be carried out conceptually and empirically. Many new issues should be well thought-out as WIS keep changing due to dramatic innovations in IT. From the conceptual perspective, the following three points are certainly worth of further research. First, as mentioned that the existence of various measures using different definitions has caused the use of inconsistent theoretical constructs and an inability to interpret across studies. Clarification of concepts is desirable and worthwhile. The border between these concepts, such as IS success, IS effectiveness, and IS service quality, is very diffusing. As a result, attempts to develop measure instruments from web users’ point of view seem to measure quite different aspects of users’ web experience and opinions. It would be worthwhile to address those critical conceptual issues and make a comparative study on those concepts. Second, among various surrogates as IS effectiveness, UIS and the SERVQUAL model are widely advocated as valid measures of IS effectiveness. However, criticisms on their theoretical backgrounds and empirical applications are evident in numerous papers and books. The relationship between service quality, UIS, and systems success need to be clarified. UIS is criticized on the theoretical relationship between user information satisfaction with IS effectiveness. Instruments of UIS were not designed or validated for measuring end-user satisfaction. Instead they focused on general satisfaction rather than empirical applications. It would be worthwhile to conduct a comparative study between 212 Conclusion UIS and the SERVQUAL model, thus to create a more comprehensive overall measure of user satisfaction. Third, another topic worthy of notice is the determinants of users’ expectations. Users’ expectations may have changed since the use of WIS varies from strict local information insider organizations to world reach. The role of information vendors and the user support systems should be paid enough attention. A new conceptual framework of determinants of users’ expectation is desired. In application, four directions are recommended. First, in Chapter and 4, shifts in service quality dimensions were identified and a general framework was proposed. The attempt to develop a generalized instrument to measure service quality across industries has proved to be fruitless. Variation analyses are needed in developing any measure instrument due to the complexity of WIS in terms of the type of WIS in different stages, the heterogeneity of web users, and the characteristics of the information systems themselves. In application, it should be of interest for future research to provide specific guidance for practitioners on how to adjust the general framework to the real industry or business environment undertaken. Second, using the scale developed in Chapter we can identify the strengths and weaknesses of web-based service quality and possible margin for improvement. In Chapter utility theory is integrated into the prioritization of service attributes. Moreover after prioritizing service attributes for improvement, the next step is to develop action plans to improve those service attributes. Future research efforts can attempt to broaden 213 Conclusion the implementation of the SERVQUAL model by integrating it into Kano’s model and QFD. Future research in this aspect will be fruitful to help organizations evaluate their web-based service quality, design improvement guidance, and finally embed them into future services to achieve customer satisfaction. Moreover, in Chapter the research is on applying utility theory into prioritization. An asymmetric relationship was assumed between positive and negative disconfirmation, by adopting an asymmetric S-shaped utility function. However, the magnitude of asymmetry of each attribute was not explored. It was simply assumed that they are identical. There is a need to develop a topology to classify those attributes into different categories by their nature. The possible integration of Kano’s model or the P-C-P model (Philip & Hazlett, 1997, 2001) into prioritization can help to further explain the dilemmas pointed out in the introduction. Research in this direction will make fruitful contributions to the quality literature. To investigate the asymmetric relationship between service attributes and the overall service quality, we have applied S-shape utility function to prioritization. Another alternative can be based on the notion of loss. Three levels of loss functions can be considered, that is, categorical, linear, and quadratic. Categorical loss is simple to count the number of the SERVQUAL gaps that fail to meet the respondents’ expectations. Linear and quadratic functions consider the linear and quadratic functions of the disconfirmation gaps. When the quadratic loss function is considered, reduction of variability is introduced as a key component of Taguchi’s philosophy. It is said that customers require each quality characteristic have a target or nominal value. The objective 214 Conclusion of quality control is to reduce variability around this target. Taguch’s quadratic loss function refers to the cost that is incurred by society when customers use a product whose quality characteristics differ from the nominal. Future research can focus on the determination of accurate significance levels for the three different loss statistics. Third, the economic dimension of service quality evaluation in WIS is usually treated subjectively. It is not easy to accurately quantify the economic dimensions, which are often soft and speculative in nature. However, the cost of providing a level of service quality, the revenue generated, and the efforts of benchmarking are definitely some of considerations that should be addressed by ISP. Quantitative measures developed to determine the balance of benefit and cost of improving service quality in WIS could be very fruitful for ISP. Finally, building a time-extended control system is desirable for web designers and ISP. The framework developed in this current study can be implemented to assess the overall web-based service quality. Another possible method is to translate those generic customer requirements of service dimensions/attributes into technical parameters, accordingly to definition approaches derived by those standardizing bodies, such as ITU, IETF, and ETSI. For example, under the dimension accessibility, there are technical parameters such as domain name system (DNS) lookup failure, transmission control protocol (TCP) timeout failure; under the dimension fulfillment, there are DNS lookup time, TCP connect time, time to first byte, content download time, and total response time, etc. In the quality literature, QFD has been approved as a good and systematical translation mechanism from voice of customers (VOC) to specified technical parameters. 215 Conclusion Those parameters of service quality in WIS can be used as system-performance and business excellence indicators. Practitioners can plot time series data to analyze trends, use quality control charts to analyze dynamic behavior of the observed overall service quality index, and also predict forecasting on future trends. 216 [...]... definition and classification, the evolutionary development and features, and the dissemination and applications of WIS Thereafter is a survey on the development of the service quality assessment and its applications in traditional information systems Since there are some similarities between web- based and traditional information systems, the existing literature on traditional information systems will... of communication 2.1.1 Definition and Classification of WIS Due to its short history and diverse applications, there is no standard definition of WIS in the literature Generally speaking, WIS are information systems based on web technologies (Isakowitz et al., 1998) A web presence is generally a part of WIS Web browsers serve as a common interface Web technologies, such as related protocols and standards,... development of new quality measures Finally, limitations of the existing literature are discussed Several directions for the further improvement and refinement are highlighted 2.1 An Overview of Web- based Information Systems The public-domain Internet Protocol (IP) network and the WWW provide a dynamic and distributed platform for interactive business applications Using the Internet and Web technologies,... the concept of service quality from the marketing research (Pitt et al., 1995) WIS are information systems with characteristics of web systems Thus, the measures or instruments developed to measure the service quality of traditional information systems will shed some insight on the way to develop new instruments to measure WIS quality A literature review on IS reveals that among various quality instruments... WIS are linked systems of entities made up of both humans and computers The human entities consist of: 1) users who interact with each other and the information device, 2) designers who create the device, and 3) information specialists who select designers and organize the data In general, WIS are information systems with characteristics of web systems (Terveen et al., 1999) WIS are classified according... number of measurement problems and result in even opposite impacts on the perceived service quality Fortunately, findings of the SERVQUAL developers’ follow-up studies have proved that the use and interpretation of the ‘expectations’ construct is appropriate in terms of diagnostic value (Parasuramn, et al, 1994a, b) 28 Literature Review In some industrial settings, such as web- based information systems, ... Information Systems In the literature, there are three important issues in any information system, that is, product, process and service Studies into the determinants of the success or failure of information systems have been conducted from the perspective of products, processes, and 33 Literature Review services, respectively Earlier research focuses more on the observable and tangible attributes of information. .. practitioners and researchers The purpose of this chapter is to: 1) provide a thorough review on the WIS literature, 2) highlight some deficiencies and limitations in 12 Literature Review the existing service management and information systems literature, and 3) hence stress motivations for the current study of managing web- based service quality This chapter is organized as follows First, an overview of the... their information requirements Kim (1990) regarded UIS as a function of organizational factors and the discrepancy between the expected information service quality and the perceived information service quality This definition paralleled UIS with the concept of service quality in the marketing research In the marketing research, service quality involves a comparison between customers’ expectations and. .. Review the definition of service quality has various contributions on the nature of service quality (Harvey & Green, 1993; Rust & Oliver, 1994; Russell & Miles, 1998) In view of different perspectives, quality can be defined as: (1) Transcendent perspective: Quality is distinctive and linked to notions of excellence that is unattainable by most (2) Product perspective: Quality is a set of attributes that . focuses on quality measurement and analysis of web- based information systems (WIS). In this introductory chapter, the evolutionary development of information systems and the service quality assessment. focuses on quality measurement and analysis of web- based information systems. Since the 1980s, the quality revolution in manufacturing had a profound impact on the competitiveness of companies namely, Era of Data Processing (1960s), Era of Management of Information Systems (1970s and 1980s), and Era of Strategic Information Systems (1990s). However, since 1993 with the birth of the World