A Hierarchy Planning Model for TFT-LCD ... - Semantic Scholar

21 downloads 60560 Views 212KB Size Report
Keywords: Scale Development, Service Quality, Structural Equation Modeling, Website. Service Quality. 1. ... software houses [11], virtual communities [20], Web ... appliance enterprises Websites[49]. ..... and repair of air force facilities,” Project.
346

International Journal of Electronic Business Management, Vol. 9, No. 4, pp. 346-354 (2011)

HOW TO ESTABLISH A SCALE THAT BEST FITS YOUR AGENDA: THE GUIDELINES TO BUILD A WEB SERVICE QUALITY SCALE Chia-Chi Chang1*, Chia-Yi Chen2 and Yi-Hsuan Chiang1 1 Department of Management Science National Chiao Tung University Hsinchu (30010), Taiwan 2 Department of Business Administration National Pingtung University of Science and Technology Pingtung (91201), Taiwan

ABSTRACT As the Web-based commerce booms rapidly, it is crucial for companies to develop comprehensive yet customized measurement of Website service quality in order to formulate their winning strategies. However, no single exiting scale can be globally effective to capture Websites service quality because service quality dimensions can differ according to the various characteristics of the websites. Therefore, managers and engineers need to cultivate their own abilities to develop a scale that can be more effectively evaluating service quality of their Websites. Accordingly, this article aims to provide guidelines for managers and/or engineers to adjust current measures or build new scales for Website service quality that best fits their agenda. This paper firstly discusses how to use literature review and focus group interview to identify possible dimensions of Website service quality. Next, item generation and selection techniques are introduced. Last, statistical methods for refining items and accessing reliability and validity of the scale are described. An example from previously published paper is also provided for readers’ reference. Keywords: Scale Development, Service Quality, Structural Equation Modeling, Website Service Quality

1. INTRODUCTION *

The invention of the Internet is heavily changing the way a company manages its business. To date, Websites have become an important channel for organizations to make transaction and provide service to customers. In recent years, the applications of information technology have provided an even more powerful platform for new Website applications [4, 14-16, 35]. Most of the marketing implications and new services of the Internet are based on the Websites, such as an online map and social networking. As the Web-based commerce becomes increasingly important, managing the service quality of Websites is crucial to the viability of organizations. In order to deliver the most superior service quality through a Website, managers must first understand how exiting and potential consumers perceive and evaluate the service quality of their Websites. For this *

Corresponding author: [email protected]

purpose, companies should develop a comprehensive measurement system for the service quality of their Websites in order to identify its strengths and weaknesses and formulate strategies for improvement [8, 34]. Customer perception (s) of the service quality of corporations is/are based on a comparison of the level of service they “expected” versus the level of service they “perceived” [12, 22, 36]. Service quality can thus be defined as the difference between customer expectations of service and perceived service. If expectations are greater than performance, then perceived quality is less than satisfactory and hence customer dissatisfaction occurs [23, 31]. In order to maintain customer satisfaction, it is critical for firms to develop an appropriate approach for assessing the quality of a firm's service. In previous literature, the dimensions and measurement of service quality of psychical environments have been well established by Parasuraman et al. [31-33]. Based on cross industry analysis, they develop a multiple-item instrument naming “SERVQUAL” to assess the

C. C. Chang et al.: How to Establish a Scale That Best First Your Agenda service quality of a company. The SERVQUAL assessment is widely used to measure consumer perceptions of an organization’s service quality in various industries [19, 50]. The SERVQUAL scale consists of five dimensions: reliability, responsiveness, assurance, empathy, and tangibles. However, as the e-commerce market booms, customers transact and communicate with firms through Websites on the Internet instead of contacting companies directly through physical channels. Conceivably, factors determining service quality of e-commerce are different from those for traditional channels [45]. Thus, whether the items of SERVQUAL (e.g., tangibles) can be compatible to the e-commerce environment is in doubt. As a result, many scholars attempt to develop advanced scales to measure the quality of e-commerce or e-service on Websites. In the past, many scales have been developed in the literature to assess the service quality of Websites [24, 26, 30, 40, 42, 43, 46]. However, none of them is globally effective to capture the service quality of Websites due to several reasons. First, because there are various types of Websites, factors influencing customer perceptions of Website service quality differ due to the idiosyncrasy of the websites. For example, “security/privacy” plays an important role in determining the overall quality of E-bay but has less impact on the quality of Google’s web search service. Accordingly, scales assessing service quality of transaction-based Websites (e.g. e-bay, Best Buy.com) are conceivably unable to adequately measure the service quality of search engines (e.g. Google, Yahoo). Thus, an established scale is only valid in measuring service quality for Website providing a specific service. Consequently, lots of studies have proposed refined service quality scales for certain type of Websites, such as electronic commerce Websites [44], software houses [11], virtual communities [20], Web portals [45], airline Websites [38], technical support Website [6], government Websites [29] and home appliance enterprises Websites[49]. In addition, because information technology evolves so rapidly, new Website-based services on the Internet are more and more disruptive and revolutionary. The determinants of Website service quality have become increasingly diversified and complex. For instance, social networking and microblogging services (e.g. Facebook, Plurk, Twitter) are blooming since 2010. How customers evaluate the service quality of social networking remains unclear and needs further research. It is predicable that the composite of Website service quality will become increasingly complicated as more innovation launches Website service to unimagined levels. As a result, a universal measure of Website service quality is even difficult to develop.

347

Based on the foregoing discussion, this paper suggests that the dimensions of Website service quality highly depend on the types of service provided. To fully capture the Website service quality of an organization, it is critical for managers and engineers to understand the similarities and differences among exiting scales of Website service quality. More importantly, managers and engineers also need to cultivate their abilities to adjust exiting measurements or develop new scales to effectively assess the quality of Website services corresponding to their business and Website characteristics. Therefore, the purpose of this article is not to develop a new scale or criticize current measurements of Website service quality. We attempt to provide guidelines for managers and/or engineers to develop a more reliable and valid scale to evaluate the service quality of their Websites. In the following, a guidance of scale development process will be introduced with an example from Yang et al. [45].

2. SCALE DEVELOPMENT GUIDELINE 2.1 Indentify Possible Factors Influencing Website Service Quality The first step to developing a Website service quality measurement is to identify possible factors influencing Website quality. In general, two approaches are commonly used to investigate dimensions of Website service quality. The first approach is by referencing exiting scales of Website service quality. The other way is to conduct a focus group interview with Website users. In the following, we first review some well-known measurements of Website service quality for readers’ reference. In the past literature, scales of Website service quality either focus on the Website design quality or concentrate on the online-transaction quality [45]. The Website design quality approaches emphasize the interface design and hardware performance of a Website. For example, Lociacono et al. [25] developed a measure called “WEBQUAL” to evaluate the interaction between a customer and a company’s Website. This scale is composed of 12 dimensions which are: Information fit to task, Trust, Design, Visual appeal, Flow, Business process, Interaction, Response time, Intuitiveness, Innovativeness, Integrated communication, and Substitutability. The Website design quality approaches are more pertinent to interface design. Therefore, this approach neglected dimensions of a consumer’s experience and service process, such as order fulfillment. Thus, scales based on this approach (e.g. WEBQUAL) are criticized as being unable to fully capture the service quality delivered through the Website [48].

348

International Journal of Electronic Business Management, Vol. 9, No. 4 (2011)

The other research stream of Website service quality research focuses on measuring the customers’ perceptions of service quality during the transaction process deliver by online retailers. For example, Barnes and Vidgen [2] developed a WebQual scale which has five factors: usability, design, information, trust, and empathy. They also provide an index of an overall rating for an e-commerce Website. Another measurement focusing on the service quality of transaction-based Websites was proposed by Yoo and Donthu [47]. This scale called SITEQUAL has four dimensions: ease of use, aesthetic design, processing speed, and interactive responsiveness. The shortcoming of measurements based on transactional Websites is that they do not consider all aspects of the purchasing process, such as providing information and customer communication. Therefore, WebQual and SITEQUAL also do not fully constitute a comprehensive assessment of Website service quality [34]. To develop a more comprehensive scale of Website service quality, some scholars attempt to integrate these two streams of research. For example, Wolfinbarger and Gilly [43] developed a 14-item scale, called “eTailQ,” which has 4 dimensions: Web site design, Reliability/fulfillment, Privacy/security, and Customer service. Additionally, Parasuraman et al. [34] developed the “e-SERVQUAL” scale and further split it into two parts: “E-S-QUAL” (efficiency, reliability, fulfillment, privacy) and “E-RecS-QUAL” (responsiveness, compensation, contact). Apart from the academic research described above, some businesses have also conducted their own methods for measuring online retailers’ service quality, such as BizRate.com. A summery of detailed dimensions in some frequently cited Website quality scales is illustrated in Table 1. Existing Website service quality scales consist of dissimilar dimensions, and those dimensions are also named differently. To summarize the dimensions of the exiting scales of Website service quality, Zeithaml et al. [48] reviewed existing research and identified six underlining criteria (information availability and content, ease of use or usability, privacy/security, graphic style, fulfillment and others) that customers used in evaluating Website service quality. Dai et al. [7] proposed two additional dimensions – system availability and recovery service. Based on this framework, Table 1 provides a summary of dimensions in exiting Website quality scales. Although literature review is an effective way to identify possible dimensions of Website service quality, the shortcoming of this approach is that it can only provide “exiting” dimensions. When new implications of Website lunches, there may be some unique dimensions that are never been identified. To explore possible new dimensions of Website service

quality, focus group interview is a commonly used technique. Researchers generally conduct a focus group interviewing the Website’s heavy users or experts to collect the first-handed information and to determine the dimension of Website service quality. For example, while Yang et al. [45] attempted to build a scale for IP Web portals, they first review exiting studies of Website service quality to identify possible dimensions. Then, they conducted a focus group of IT and marketing managers from a company with IP Web portal. By using both literature review and focus group interview, the major service quality dimensions of IP Web portals were thus formulated for further investigations. 2.2 Item Generation and Selection The objective of this step is to generate specific items for the proposed dimensions of Website service quality and to select the items that can have a good validity in terms of describing relevant Website service quality. To generate the items, one should conduct an extensive literature review focused on concepts related to the dimensions of Website service quality and used the selection process proposed by Tian et al. [41]. First, from the literature review, one would select or as many relevant items as possible, and some of the selected items need to be reworded in order to be appropriate. Second, at least five experts should be asked to allocate each item to one of the dedicated dimensions [21]. After this process, items can only remain if more than half of those experts classified them into the appropriate category. Third, another set of five experts should be asked to evaluate each of the remaining items as clearly representative, somewhat representative, or not representative of the dimension. In Saxe & Weitz’s [37] study, they suggested that all retained items should be rated as “clearly representative” by at least 50% of the judges or the item should be deleted. The targeted sample will then be asked to respond to the Website service quality scale by using a seven-point Likert-type scale (1 = “strongly disagree,” and 7 =“strongly agree”). In Yang et al. [45], the authors drafted a survey questionnaire and asked the managers and the users of IP Web portals to screen the items to refine them. 2.3 Item Refinement and Confirmation of the Dimensions It was necessary to at least collect two sets of samples in the process of developing the scale. First survey of sample. An exploratory factor analysis will be performed to reassign items and restructure dimensions in this survey. Using data obtained from the first sample, items with a loading value below 0.5 on any factor, or high cross-loadings on two or more factors, should be eliminated through the principal

C. C. Chang et al.: How to Establish a Scale That Best First Your Agenda component factor analysis with a rotation of the first sample. The factor eigenvalues of Website service quality should be greater than 1. After Varimax

349

rotation, a clean factor structure which explains most of all variance will emerge. A preliminary scale will be produced.

Table 1: Summary of dimensions in different scales to measure website service quality

In Yang et al. [45], they conducted a survey and collected a sample with 1992 effective respondents. They randomly select 996 respondents and conducted a principal component factor analysis with a Varimax rotation. They following the rules described above to eliminate items that do not load strongly on any factor (values below 0.5) or had cross-loadings. Second survey of sample. Then, in the second survey, a confirmatory factor analysis is needed to construct the final Website service quality scale. To build a strict factor structure, confirmatory factor analysis is performed by using data from the second survey and items are deleted when the loading is below 0.7. For example, Yang et al. [45] also conducted confirmatory factor analyses using the remaining 996 cases in the dataset. They delete items with unacceptably small loading (below 0.7) on its designated factor.

3. DATA ANALYSIS 3.1 Measurement of the Model Fit To determine which measurement model fit the data best, structural equation modeling should be used to conduct confirmatory factor analyses. After

the second survey, confirmatory factor analyses needs to be performed by using statistic software, LISREL, in order to reveal model is the best model. If the major fit indices suggest a reasonable fit (see Table 2) [1, 17, 18]. At this stage, other analyses will also need to be performed in order to validate the scale, including reliability validity test. A brief summary of how it should be done will be included as follow. Table 2 Indices of SEM Indices of SEM χ2 χ2/df Goodness-of-Fit Index (GFI) Adjusted GFI (AGFI) Normed Fit Index (NFI) Non-Normed Fit Index (NNFI) Comparative Fit Index (CFI) Root Mean Square Residual (RMR) Root Mean Square Error of Approximation (RMSEA) Standardized Root Mean Square Residual (SRMR)

Suggested standard P > 0.05 0.90 > 0.90 > 0.90 > 0.90 > 0.95 < 0.05 < 0.05