Empowering School of Accounting Websites through Quality Assurance : Development and Implementation of a User-perceived Instrument

Today’s institutions of higher education are facing an increased number of significant challenges taking place in the political, economic, social and technological environment. Accordingly, the issues of performance, accountability, and marketing strategies have become ever more important. It has been suggested that universities that are more market or customer orientated can perform better. In dealing with customers, most universities have utilised Web technologies for both informational and promotional purposes. The question is whether they have designed their Web sites well enough in order to gain the benefits from Web utilisation. Previous studies on quality of Web sites are not lacking, but most of them have focussed mainly on business Web sites. Empirical research that focuses on the Web site quality of institutions of higher education has been scarce. This study focuses on the websites of Schools of Accounting, and selects the websites of Schools of Accounting in New Zealand’s universities as the research objects. The main objectives of this study are to develop and validate an instrument for measuring School of Accounting website quality from the perspectives of the users, and to implement the proposed instrument to measure and rank the quality of websites of Schools of Accounting in New Zealand. The results from this initial application substantiated the validity and reliability of the instrument.


Introduction
In recent years, universities globally have experienced a number of significant challenges and changes taking place in the political, economic, social and technological environment.One of the major challenges is the limited amounts of government funding that push universities to act more like business ventures.As a result, universities have to behave as rival firms and compete intensely for market share and revenue (Caruana, Ramaseshan, & Ewing, 1998;Scott, 2003;Soutar & Turner, 2002).With this in mind, it is obvious that the issues of performance, accountability, and service quality have become ever more important.It has been suggested that universities that are more market or customer orientated can perform better (Caruana et al., 1998).
The Web has been broadly considered as an effective tool for commercial and communicational purposes (Huizingh, 2000;Lin & Lu, 2000;Liu & Arnett, 2000;Ranganathan & Ganapathy, 2002).Web technologies enable organisations to reach out for broader customers and create more networking opportunities.Most today's universities have taken advantage of the commercialisations of the Internet, Web technologies and their applications, as parts of their customer or internal business process strategies.The purposes of a university website are both informational and promotional -it tells students, academic and administrative staffs about courses, timetables, and other relevant information, and it tells prospective students and prospective employees about the university and its programs.However, simply using websites does not guarantee a success.The benefits of website utilisation are rather hard to pin down.Almost always there are many alternating websites that users or customers can go to.University websites rely on their quality and ability to attract and keep both existing and potential customers, and hence an understanding of the online users' mindset is critical to designing a usable website.The chances of gaining the benefits from the implementation of Web technologies are higher if a cautious emphasis is put in place on the user-perceived quality of the websites (Aladwani & Palvia, 2001;Cao, Zhang, & Seydel, 2005;Liu & Arnett, 2000;McMurdo, 1998;Wan, 2000).
A website should be designed to suit relevant target users.Previous studies on quality of websites are not lacking, but most of them have focussed mainly on business websites (Cao et al., 2005;González & Palacios, 2004;Liu & Arnett, 2000;Ranganathan & Ganapathy, 2002).Empirical research that focuses on the website quality of universities has been scarce.This leads to a need for evaluating this issue and developing a relevant instrument to critically measure the quality of university websites.This study focuses on the websites of Schools of Accounting, and selects the websites of Schools of Accounting in New Zealand's universities as the research objects.The main objectives of this study are to develop and validate an instrument for measuring School of Accounting website quality from the perspectives of the users, and to implement the proposed instrument to measure and rank the quality of websites of Schools of Accounting in New Zealand.The results of the study are expected to provide some valuable input that can assist School of Accounting policy makers in designing and developing suitable and effective key performance indicators within their school's customer or internal business process strategies.
The remainder of the paper is structured as follows.First, the conceptual background of website quality is reviewed.Second, the process in developing and validating the instrument is explained.This includes various steps involved in the measurement and scale development.Third, the proposed instrument is used to measure.From this initial implementation, the quality scores and the ranking of websites of Schools of Accounting in New Zealand are described.Finally, concluding remarks are provided.

Website quality
Website quality can be approached from many different viewpoints.No particular body of literature specifically addresses the concept of website quality (Cao et al., 2005).Computer and information specialists may focus on the technical aspects of the Web, such as how to develop tools to retrieve information from the Web or how to make a website works properly (Arasu, Cho, Garcia-Molina, Paepcke, & Raghavan, 2001;Kleinberg, 1999;Nel, van Niekerk, Berthon, & Davies, 1999).Behavioural scholars may be more interested in the issue of why and what users use the Web for or how Web technologies affect the behaviour of users (Burnett & Marshall, 2003;Rogers & Marres, 2000;Wellman, Quan-Haase, Witte, & Hampton, 2001).From a management point of view, it is interesting to see how organisations make the most of the emerging Web technologies and their applications (Huang, 2005;Liu & Arnett, 2000;Norton & McGovern, 2001;Wakefield, 2002;Wan, 2000).
The basic goal of a website is to provide information (Angehrn, 1997;Bhatti, Bough, & Kuchinsky, 2000).Information refers to processed data that is organised, meaningful, and useful to the users (Cushing & Romney, 1994).There are different dimensions of information quality according to the context it is being referred to (Fox, Levitin, & Redman, 1994).However, as a basis of any information quality initiative, the quality of information is commonly assessed via evaluation of its generic characteristics (Xu & Koronios, 2005).These characteristics have been consistently found to be similar across various previous studies, although some studies may make out more attributes than the others.For example, Ballou and Pazer (1982) defined four fundamental dimensions of information quality: accuracy, timeliness, completeness, and consistency.Burk and Horton (1988) identified three attributes of information quality: accuracy, comprehensiveness, and currency.Rai, Lang, and Welker (2002) described three attributes that can be used to measure information quality: accuracy, content and format.
Web operates in a cyber environment, and hence there is no direct human contact offered through a website.This makes website quality contextually different than information quality.Information quality in the virtual environment captures the issue of content, which was traditionally considered as the most important factor of a website.'Content is king' used to be a well-known slogan (McCarthy, 1995;Huizingh, 2000).Consequently, many previous studies have used mainly information quality criteria in evaluating and designing a website (McMurdo, 1988;Tate & Alexander, 1996;Huizingh, 2000).However, more recent studies have emphasised that website quality has to go beyond information quality and should focus on the users or customers as the major foundation in designing and evaluating website quality (van Iwaarden, van der Wiele, Ball, & Millen, 2004;Xu & Koronois, 2004).Norton and McGovern (2001) said that content is not king on the website, the reader is king.Saying 'content is king' is like saying the product is king, rather than the customer (McGovern, 2000).Katerattanakul (2002) proposed a definition of website quality as the website's fitness for use by users.All these indicate the importance of user-perceived quality.
There have been a few studies that try to accommodate both the 'content' and 'user' perspectives in the process of exploring the dimensions of website quality.These studies collected responses from participants that were selected to match a profile of a certain target Web user.Ranganathan and Ganapathy (2002) examined the dimensions of business-to-customer websites based on a survey of 214 online shoppers.They found four key website dimensions: information content, design, security, and privacy.Aladwani and Palvia (2002) developed a 25-item instrument for measuring user-perceived business website quality.Firstly, they collected responses from 101 Web users during their design process, and came up with four variables of website quality: content quality, specific content, technical adequacy, and Web appearance.In the normalisation process, the Web users were split into four different groups and asked to evaluate the websites of a bank (25 users), a bookshop (31 users), a car manufacturer (34 users), and an electronics retailer (37 users).Cao et al. (2005) also did a similar but smaller study as Aladwani and Palvia (2002).They asked 71 first and second year university students to express their perceptions of three online bookshops: amazon.com,biggerbooks.com,and half.com.They found four constructs that relate to website quality: information quality, service quality, playfulness, and system quality.These three studies, among others, confirm that in the perception of users, content or information quality is only one dimension of website quality.Therefore, the design and evaluation of a website should be assessed using a multi-item instrument that combines all of the users' or customers' values rather than mainly its informational content.

The design of an instrument for measuring user-perceived School of Accounting website quality
As a starting point, we used Aladwani and Palvia's (2002) 25-item instrument for measuring website quality.There were two main reasons for this.Firstly, this instrument has encompassed both the perspectives of users and the importance of 'content'.Secondly, this instrument has been developed and validated thoroughly in their study.They started with a conceptualisation process in delimiting the domain of the constructs and identifying indicators of each construct based on academic literature.It was then followed by a scale design step that analysed the validity and reliability of the constructs and their respective indicators.Finally a normalisation process was performed to verify and validate the constructs.
Nonetheless, while the development and validation processes applied to this instrument have been extensive and solid, the main focus of the instrument was measuring website quality of businesses.Therefore, in this study we made a few adjustments to suit the instrument for evaluating School of Accounting websites.Subsequently, we followed Churchill's (1979) recommendations by applying a reliability analysis and a normalisation process to ensure the validity and reliability of the adjusted instrument.
The first adjustment to Aladwani and Palvia's (2001) 25-item instrument was an exclusion of the security indicator.While one of the purposes of university websites is promotional (Brody, 1999), the university websites practically do not involve much the carrying out of transactions (e.g.used of secure sockets layer, digital certificates, etc.)The second adjustment to the instrument was fine-tuning the specific content of a few indicators to suit the academic environment (e.g.owners were changed to academic and administrative staffs, products and/or services were changed to courses and/or subjects, etc.)The result of the adjustment was a 24-item instrument consisting of four constructs: technical adequacy, information quality, service ability, and Web appearance (see Table 1).

Reliability analysis
To evaluate the unidimensionality of the indicators and distinguish the appropriate indicators for each construct, a reliability analysis was conducted.The list of the 24 items was administered to third year accounting students who had experiences in using the Web.Students were selected since they represent one of the major users of School of Accounting websites.They fitted the profile of the target users for both service and marketing purposes.The students were asked to rate the importance of each item for measuring School of Accounting website quality.Each item was measured using a seven-point scale from 1 being 'extremely not important' to 7 being 'extremely important'.There were 126 participants, and 118 responses were usable.56% of the respondents were female and 44% were male.68% of the respondents accessed Web more frequently (more than three times a week), and 32% less frequently (less than three times a week).
From data that were collected, a reliability test was conducted using Cronbach's alpha.The alpha values for technical adequacy, information quality, service ability, and Web appearance were 0.68, 0.93, 0.72, and 0.87, respectively.Since the alpha value for technical adequacy did not achieve the common acceptable value of 0.7, an analysis of individual indicators was conducted.It was found that interactivity indicator had a low corrected item-total correlation of -0.28.Hence, the indicator was removed.This screening made up an alpha value of 0.89 for technical adequacy -hence, showed an improved level of reliability.At the end of this process, 23 indicators were kept.
The next step involved a factor analysis to discover whether the four constructs could be explained largely or entirely in terms of their respective indicators.We applied a factor analysis procedure with varimax rotation and used a cut-off point of 0.5, which was considered as a significant loading (Hair, Anderson, Tatham, & Black, 1995).The results of the factor analysis, which are shown in Table 2, were consistent with Aladwani and Palvia's (2002) findings.There were four website quality constructs explaining 65.22% of the variance in user-perceived website quality.Each of these four constructs matched all of its respective indicators, and therefore, all 23 indicators were retained.

Normalisation of constructs and indicators
Normalisation process was conducted to verify and validate all four constructs in the instrument.This process was similar to conducting a pilot test to ensure the validity of a questionnaire.Oppenheim (1966, p.29) said that "respondents in pilot studies should be as similar as possible to those in the main inquiry" A similar point was also mentioned by Sudman and Bradburn (1983, p.282), "pilot-test on small sample of respondents similar to the universe from which you are sampling".Therefore, a different group, consisted of 48 third year accounting students (27 females and 21 males), was asked to participate.Each of the participants was asked to evaluate a website of a School of Accounting in Australia using the 23-item instrument as shown in Table 3.Each item was measured on a seven-point scale with 1 being 'strongly disagree' and 7 being 'strongly agree'.To evaluate the validity of the instrument indicators and constructs, exploratory analyses were conducted.Two tests were conducted in these analyses: individual item reliability and composite reliability.The results of these analyses are presented in Table 4.
In general, for each indicator, a standardised loading of more than 0.7 is considered to be reliable, because it ensures that at least 50% of the variance of the indicator is explained by the construct that the indicator intends to measure (Bagozzi, 1994).In other words, a standardised loading of more than 0.7 ensures that each indicator distributes more variance with the construct score than with error variance.In studies where the indicators are not well developed, a loading of 0.5 or 0.6 is acceptable (Chin, 1998).The exploratory analyses showed that all 23 indicators had loadings above 0.6 on their particular constructs (as shown in Table 4).Additionally, it showed that in each block of indicators, each loading in the block related to its respective construct was higher than its loadings related to other constructs or other block of indicators.Therefore, all indicators were considered as reliable and were retained as indicators of their respective constructs.
To ensure the estimation of the composite reliabilities of the blocks of indicators, Cronbach's alpha was used to assess the internal consistency for each block of indicators.It was found that alpha values for technical adequacy, information quality, service ability, and Web appearance were 0.84, 0.81, 0.86, and 0.93, respectively.These values were above the acceptable value of 0.7, and hence, indicated that the composite reliabilities estimated by these measures were satisfactory.Having undergone thorough development and validation processes, we believe that the instrument is sound for measuring School of Accounting website quality.

Measuring and ranking the quality of websites of Schools of Accounting in New Zealand
There were eight universities in New Zealand.All of them had a "School of Accounting", either as a standalone department or as a part of a collective department.These Schools of Accounting were as follows: ( 1 To increase confidence that values of website quality found from this study were due to website quality constructs rather than user effects, two groups of students were used as respondents.The first group consisted of 33 third year accounting students (19 females and 14 males), and the second group consisted of 41 first year accounting students (21 females and 20 males).Each respondent was asked to evaluate websites of all eight Schools of Accounting using the 23-item instrument.Each item was measured on a seven-point scale with 1 being 'strongly disagree' and 7 being 'strongly agree'.
The results based on the responses from all respondents are reported in Tables 5 and 6.It can be seen in Table 5 that the websites of Schools of Accounting in the University of Auckland, the University of Otago, and Victoria University of Wellington are the top-three websites overall.From Table 6, it can be seen that in terms of all constructs (the technical adequacy, information quality, service ability, and Web appearance), the website of School of Accounting in the University of Auckland is consistently in the first position, and the website of School of Accounting in the University of Waikato is consistently in the last position.There are slight variations of ranks in the remaining six websites, but in general the websites of Schools of Accounting in the University of Otago, Victoria University of Wellington, and Massey University are consistently in the second to fourth positions, except for the Web appearance construct where the website of School of Accounting in Massey University is in the fifth position.Concurrently, the websites of Schools of Accounting in the University of Canterbury, Lincoln University, and Auckland University of Technology are consistently in the fifth to seventh positions, except for the Web appearance construct where the website of School of Accounting in Lincoln University is in the fourth position.
We also provide the ranking and descriptive statistics based upon each group of respondents in Table 7.There are four groups of respondents, which include (1) third year students; (2) first year students; (3) female students; and (4) male students.It can be seen that the ranking in each group is generally consistent with the overall ranking based upon responses from all respondents.The website of School of Accounting in the University of Auckland is consistently in the first position.Concurrently, the websites of Schools of Accounting in the University of Otago, Victoria University of Wellington, and Massey University are in the second to fourth positions.The websites of Schools of Accounting in the University of Canterbury, Lincoln University, and Auckland University of Technology are in the fifth to seventh positions.Finally, the website of School of Accounting in the University of Waikato is consistently in the last position.
The results of this implementation show that the 23-item instrument is able to differentiate the qualitative value of each School of Accounting website.Hence, it is a valid instrument.Simultaneously, the instrument is also been able to provide consistencies of the results across a diverse range of respondents.Hence, the instrument is also reliable.Nonetheless, chances for further improvement to the instrument are always open and any future change for the better is certainly enviable.

Conclusion
Websites have been widely used commercially throughout industry, government, education, and practically in any other types of institutions (Liu & Arnett, 2000).It has been acknowledged that in order to gain the benefits from utilisation of Web technologies, a high-quality website is needed.Previous studies on quality of websites are not lacking.However, most of the previous studies have been focussed mainly on business websites.In this study, an instrument for measuring School of Accounting website quality was developed and validated thoroughly by taking into account both the perspectives of the users and the importance of its informational content.The generated instrument has four main constructs with a total of 23 indicators.It has been subsequently put to the test by implementing it for measuring and ranking the quality of websites of Schools of Accounting in New Zealand.The results from this application have corroborated the validity and reliability of the instrument.
The results of the study provide some important guidelines for the design and evaluation of School of Accounting websites, and university website in general.Hence, not only they contribute to the existing empirical literature, but may also assist the policy makers in universities.The 23 School of Accounting website quality indicators can be used as a foundation for the design and development of a more effective School of Accounting websites.The quality scores and the ranking of websites of Schools of Accounting in New Zealand provide an illustration about the value and attractiveness of each website in the eyes of the users.The quality scores from the respondents give an indication of which website dimensions should be maintained and which dimensions should be improved.The best-performing website can also be used as a benchmarking model.It should be noted though that changes constantly take place in the economic, social and technological environment.Therefore, it is important to continuously monitor these changes and keep updating the instrument as necessary.Xu, H., & Koronios, A. (2005).Understanding information quality in E-Business.The Journal of Computer Information Systems, 45, pp.73-82.
) Department of Accounting and Finance, Business School, University of Auckland; (2) Accounting Discipline, Faculty of Business, Auckland University of Technology; (3) Accountancy, Finance and Information Systems Department, College of Business and Economics, University of Canterbury; (4) Centre of Accounting Education and Research, Commerce Division, Lincoln University; (5) School of Accountancy, College of Business, Massey University; (6) Accountancy and Business Law Department, School of Business, University of Otago; (7) School of Accounting and Commercial Law, Faculty of Commerce and Administration, Victoria University of Wellington; and (8) Department of Accounting, Waikato Management School, University of Waikato.All eight websites of these Schools of Accounting were evaluated.

Table 1 .
School of Accounting website quality constructs and indicators

Table 3 .
The user-perceived School of Accounting website quality instrument

Table 5 .
Ranking of the websites of Schools of Accounting in New Zealand

Table 6 .
Ranking and descriptive statistics of the websites of Schools of Accounting in New Zealand based on each website quality construct

Table 7 .
Ranking and descriptive statistics of the websites of Schools of Accounting in New Zealand based on each group of respondents