Key usability factors of service-oriented web sites for students: an empirical study

Abstract

This paper aims to report on a web usability study and to identify and prioritise key web interface usability factors (WIUFs) for web sites of 36 student-related online services categorised into three groups: personal services, purchase services and study-related web sites. In this study, involving 400 student internet users (SIUs), 12,310 data points were collected and analysed using a multiple linear regression test. Seven WIUFs were tested: use of colour and font (UCF), use of graphics and multimedia (UGM), clarity of goals in web site (CGW), trustworthiness of web site (TOW), interactivity of web site (IOW), ease of web navigation (EWN), and download speed of web site (DSOW). The study results reveal that every online service category has a different set of crucial WIUFs. SIUs' web usability preferences were compared with those of general internet users. The participants were all Malaysians; therefore, generalising the findings to all SIUs will require a confirmatory study with SIUs from other parts of the world. Web developers can use the results to design usable web sites for specific online service categories. The research offers a simpler alternative to measure web usability and to determine which WIUFs are crucial for a specific online service category with consideration of the users' role. This study overcomes some weaknesses of previous studies, i.e. small sample size, no consideration of product-task relationship, no specific customer group and cumbersome procedures.

Introduction

Web sites have become an important medium connecting buyers and sellers in cyberspace. Hence the design of web sites, particularly their usability aspects, assumes great significance. ISO 9241 defines usability as the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use ([30] Karat, 1997).

Because of the importance of usability in web applications, there have been several studies of web site usability. For example, [48] Nielsen (2000) and [37] Lecerof and Paterno (1998) studied how web site usability reduces errors, enhances accuracy, increases usage and provides a better "look" for a web site. An empirical study by [41] Lohse and Spiller (1999) found that web site usability explains a high percentage of variance (61 per cent) in online sales. [44] Moraga et al. (2006) compared the different quality models for web portals and found that researchers paid special attention to visual aspects of interface design. [7] Calero et al. (2005) reviewed 60 papers on web quality (including web design) from 1992 to 2004 and classified them according to the web quality model. [3] Allen (2002) studied the interface design of the University of South Florida's virtual library employing a usability testing method. [14] Chowdhury et al. (2006) reviewed many usability studies (including research on interface design) and studied their impact on digital libraries. They suggested that digital libraries should be evaluated with respect to their target users. [68] Xie and Cool (2000) compared searcher experiences with web and non-web interfaces to online databases and found that some web interface designs outperformed non-web interfaces.

[37] Lecerof and Paterno (1998) highlighted the importance of specified users in determining the crucial usability aspects of a system. For example, ease of use may be the crucial usability aspect for web sites targeting children, whereas efficiency may be the crucial usability aspect for business-to-business e-procurement web sites. [19] Ginige and Murugesan (2001) recommended ten key steps for the development of a successful web site. Among them is the need for developers to identify clearly the system's main users. This is in agreement with the ergonomics rule of thumb which states that "one size does not fit all".

However, most web site design guidelines presently cater for general users, e.g. Ergonomic Guidelines for User-Interface Design ([23] Hix and Hartson, 1993) and Web Analysis and Measurement Inventory Factors ([34] Kirakowski et al. , 1998). Users vary in many ways, so different groups of internet users may have different needs for web interface design. Therefore, user-centric interface design is vital for web-based systems ([19] Ginige and Murugesan, 2001; [63] Tilson et al. , 1998).

As internet users vary in many ways, one means of clustering them is through the use of demographics. This is commonly done when the internet community profile is portrayed as different homogeneous groups ([36] Lebhar-Friedman, Inc., 2000). [12] Copas (2003) suggested that online users be described using personality traits; this researcher also commented that frequent internet users are considered part of the "internet community", and therefore their traits can be studied as a homogeneous group.

One group belonging to the internet community is student internet users (SIUs). This group is information and communications technology-savvy and generally has a high level of internet familiarity ([20] Greenspan, 2003). [51] Pastore (2000) found that university students use the internet for shopping and that many describe themselves as "internet dependent". [20] Greenspan (2003) also observed that tertiary students surf the internet extensively and participate actively in online spending. The high level of internet penetration by these students is believed to be due largely to the fact that many of them have grown up with computers and have incorporated the internet into their daily routines. [21] Greenspan (2004) also found that Asian students in particular lead in spending through electronic commerce. Thus, studying the perceptions of Asian university students (e.g. Malaysian students) of web site usability will assist this market segment to be understood better and consequently reached more effectively.

SIUs were found to be the most internet-connected population ([42] Looney and Lyman, 2000). In 2002, 86 per cent of students went online, compared with 59 per cent of the other population segments ([25] Hoffman et al. , 2004). SIUs' interest in the internet has created a huge market for online products and services. According to Greenfield Online (a US-based research company), 81 per cent of US tertiary students were active online shoppers ([51] Pastore, 2000). In addition, [57] Saliba (2000) found that SIUs spent about US$164 billion online per year. Furthermore, [60] Stark and Meier (2001) believed that internet purchasing among students would continue to grow tremendously. Their findings showed that the average expenditure per student had increased yearly - US$235 in 1998, 381 in 1999 and 471 in 2000.

There have been many web site usability studies. [2] Agarwal and Venkatesh (2002) and [33] Keevil (1998), for example, developed methods for measuring the usability of web sites. [2] Agarwal and Venkatesh (2002) converted the Microsoft Usability Guidelines into weighted scores, where the rating of each item is multiplied by its weight and a total score is obtained by adding all the individual weighted scores. [33] Keevil (1998) used a simplified checklist approach. The items in the checklist have a dichotomous scale ("Yes" or "No" response), and the total score is just the sum of the checked items. [18] Gehrke and Turban (1999) studied five usability factors for business web sites: download speed, content, ease of navigation, security, marketing aspects. [31] Katerattanakul and Siau (1999) studied and developed a method of measuring the information quality of web sites, which included intrinsic, contextual and representational factors and accessibility of information. [61] Teo et al. (2003) conducted an empirical study of the effects of interactivity on web users' attitudes. They found that an increased level of interactivity had positive effects on users' perceived satisfaction and perceptions of the effectiveness, efficiency and value of a web site, as well as their overall attitude towards a web site. [8] Chau et al. (2000) conducted an empirical study of the impact of information presentation modes (e.g. graphics or text) on online shopping. They proposed that information presentation modes should vary based on user familiarity with the shopping items; for example, images should be used for unfamiliar items and text for familiar ones. [28] Iwaarden et al. (2004) conducted a usability study of Northeastern University's and Erasmus University's web sites using [13] Cox and Dale's (2002) key quality factors in web site design. They found that the quality dimensions applicable in the service sector are also applicable to web sites.

Even though there have been so many web usability studies, [2] Agarwal and Venkatesh (2002) found that most lacked focus on specific users and industries, and were conducted using a small sample size. [2] Agarwal and Venkatesh (2002) conducted a large-scale study of 1,475 web evaluations across four industries (airlines, bookstores, auto manufacturers and car rentals) and two-user roles (customer and investor). It was the first study that attempted to find the industry-user (product-task) relationship with web usability. However, their study had the following limitations:

- The study was too general, as it did not focus on any specific customer group. The subjects were customers of shopping malls in general, which could be anybody, including students, working adults, retirees, etc. Usability studies should be targeted to specific users in order to derive an accurate result ([22] Helander, 2000).

- The subjects were not allowed to choose the industry or the web site they would like to evaluate - the web sites were assigned randomly to the evaluators. This posed a problem, because an evaluator might not have been a true customer of the web site.

- The number of web sites chosen (five to six per industry) was too small to represent each industry adequately.

- Too few industries were evaluated (only four); it is difficult to derive anything substantive about the usability preferences of the user (i.e. customer and investor) based solely on a small number of consumer industries.

- Even though content validity was tested using expert and user assessments, and construct validity was tested by comparing the calculated usability metric with a multi-item scale of overall usability, the percentage of variance (R² ) explained by the factors and the contribution of each factor to the overall usability were not derived. This was because single-item scales were employed, and these cannot be used for statistical tests such as factor analysis and multiple linear regression.

- It is uncertain whether the evaluators had the capacity to understand the usability factors, since the evaluators were chosen randomly from physical shopping malls. There was no filtering of the evaluators; therefore, some might not have had sufficient computer literacy and experience in using the internet to perform the evaluation.

- It is uncertain whether the evaluators understood the meaning of the complex usability factors in the study such as content, ease of use, promotions, made-for-the-medium and emotions. There was no verbal explanation or hands-on lab tutorial included in the evaluation procedure to ensure that evaluators grasped the meaning, answered the questionnaire and performed data entry of the answers.

- The evaluation procedure was cumbersome and took considerable time because too many steps were required. There were many items, and to derive an item score the evaluators had to perform three subsequent steps - assign a weight, provide a rating and calculate the item score by multiplying the two values.

The usability evaluations included in-depth informational content of the web sites such as relevance of the web site, depth and breadth of the content, timeliness of the information, element of challenge and plot. However, only five minutes was allowed for the evaluation procedure making adequate assessment of information content difficult.

Thus, the present research aims to overcome the limitations in [2] Agarwal and Venkatesh's (2002) study. The objectives in this analysis are as follows:

- To identify and compare the crucial web site usability factors of different web industry groups from the perspective of SIUs.

- To recommend specific web site usability guidelines for designing web sites that target SIUs.

SIUs were chosen as the sample for this study. A large number of student-related industries (36) were examined to derive a clear understanding of SIUs' usability preferences for the various industries. Based on the service-oriented nature of the web sites from the 36 industries, three distinctive groups of industries emerged. Web industries offering information and services were grouped as "personal services"; those mainly interested in online sales were grouped as "purchase services"; those offering educational content were grouped as "study-related". The 36 web industries and their groups can be found in the Appendix.

The evaluators were given the freedom to choose web sites among the industries, and each industry had more than 300 evaluated web sites. A five-point Likert scale was used; therefore, the percentage of variance and contribution of each factor to the overall usability can be calculated through multiple linear regression. The evaluators were specifically selected from university students who had three to five years' experience using the internet. They were given a hands-on laboratory tutorial and verbal explanation to help them understand the usability factors and to perform the evaluations. The evaluation procedure was simplified (with fewer steps) to allow quick and large sample-size evaluation to be conducted. Information content was excluded, as the evaluators would take considerable time to perform such evaluations. In most web sites, there are too many webpages and too much text to be read to provide a reasonable judgment of the information content; therefore, it is not practical for large sample-size research.

The preliminary results of this study were presented in [69] Yeow and Nathan (2003). This paper presents the final results with more detail on the literature, method, results, discussion and conclusions.

Literature review

Factors affecting overall web usability (OWU)

OWU measured the SIUs' overall perception of web usability, and it has become an important key to the success of electronic commerce (EC) web sites ([2] Agarwal and Venkatesh, 2002; [63] Tilson et al. , 1998; [34] Kirakowski et al. , 1998; [16] Ezquerra et al. , 1999). OWU plays a major role in the successful design of a web site ([2] Agarwal and Venkatesh, 2002; [63] Tilson et al. , 1998; [34] Kirakowski et al. , 1998; [16] Ezquerra et al. , 1999; [40] Liu and Arnett, 2000; [35] Kumar et al. , 2004). Jakob Nielsen, a usability guru, estimated that 90 per cent of all EC web sites were too difficult to use ([4] Barnum, 2002). When users find a web site hard to use, most feel uneasy and leave the site without accomplishing their objectives.

There are many factors that might affect the overall usability of a web site. To ensure content validity, a literature review has been conducted of previous studies that identify these factors. The following sections discuss the seven factors found to affect overall web site usability, and these factors are used in the present research framework. These factors are called web interface usability factors (WIUFs), and they are equivalent to the factors in the Microsoft Usability Guidelines as shown in Table I [Figure omitted. See Article Image.].

The other factors in the MUG ([32] Keeker, 1997; [2] Agarwal and Venkatesh, 2002) were excluded, as the present study is limited to the information architecture of web sites. Information architecture is defined as the way in which a web site and its pages are organised, labelled and navigated to support browsing and searching throughout the web site ([55] Rosenfeld and Morville, 2002). It does not take into consideration the in-depth information content of a web site, such as the relevance of the web site, depth and breadth of content, timeliness of information, promotional aspects, elements of challenge and plot. As previously mentioned, evaluation of the information content of a web site requires considerable time and is not feasible for large-scale evaluation. Besides, it has already been established in research by [2] Agarwal and Venkatesh (2002) and [11] Cole et al. (2000) that the information content of a web site is very important for any industry and task; therefore, information content should be a prerequisite for any web site. Since this is already clearly established, testing the information content variables would be redundant.

[2] Agarwal and Venkatesh (2002) tested the content validity of the above seven WIUFs through expert and subject assessments. First, the WIUFs were validated because they are the actual design practice used by market leaders such as Microsoft. Second, the factors were validated and refined by four experts (in usability, information systems and statistics), two PhD students and 40 undergraduates ([2] Agarwal and Venkatesh, 2002). Last, the WIUFs were validated by 30 experienced web users ([2] Agarwal and Venkatesh, 2002). However, [2] Agarwal and Venkatesh's (2002) study did not include content validation through a literature review. The following sections present the literature review of the WIUFs.

Web site attractiveness: use of colour and font (UCF) and use of graphics and multimedia (UGM)

[34] Kirakowski et al. (1998) conducted a survey of the various factors influencing customers' decision making in using web sites to determine the significance of each factor. The factors were then listed in descending order of importance:

- attractiveness;

- ease in navigation;

- ease in finding items; and

- clearly labelled items.

The attractiveness of a web site was found to be the most important factor in influencing customer decision making. Attractiveness refers to the aesthetic appeal of a web site, with reference to the UCF and the UGM. [64] Tractinsky (1997) found that aesthetics correlated with perceived usability in an investigation of automated teller machines. Several other web usability studies have confirmed that attractiveness significantly affects the usability of web sites ([39] Lindgaard, 1999; [65] Tractinsky et al. , 2000; [5] Brady and Phillips, 2003).

Elements such as colour and graphics are believed to affect the perceived usability of a web site. [39] Lindgaard (1999) found that colour can be an important predictor of web site usability. Several authors have also linked the beauty of a web site to its usability. [58] Shenkman and Jonsson (2000) found that the best predictor of a typical user's overall judgment of a web site is its beauty.

Aesthetics have also been linked to users' emotions. Attractive appearance is believed to trigger an observer's emotions, and if a positive emotion is triggered, a positive outcome is believed to be the result ([26] Isen, 1993). [50] Norman (2002) found that users feel better when using a more attractive product; it is believed to give users a sense of perceived satisfaction with the product even though the product may not inherently possess qualities that satisfy users.

Clarity of goals in web site (CGW)

[66] Turban et al. (2006, p. 662) defined CGW as the extent to which the goals of a web site are understandable. The example given was the www.obo.co.nz web site, which clearly stated its goals (in the "About OBO" link) as community building, product sales, and research and development. It further supported its goals by providing online discussions, sponsored players and an image gallery for community development; an online purchasing mechanism for product sales; and online surveys for research and development ([46] New Zealand MED, 2000, p. 12).

Other goals of a web site may be saving costs, increasing processing efficiency, improving customer service, providing a convenient marketing channel, etc. ([66] Turban et al. , 2006, p. 667). Goals of a web site depend on the business models of the company. [66] Turban et al. (2006, p. 92) have presented several business-to-consumer business models, such as transaction brokers, information portals, community portals, etc. The business model of the web site should be made clear to facilitate the full use of the web site for both businesses and customers. [56] Sabberwal and Chan (2001) found that alignment between business and information system goals (when the web site goals match business goals) improves business performance. [54] Rivero (1999) highlights that clear web site goals result in customer needs being better met, customer time being saved and company resources being optimised. Even though the CGW is crucial, usability gurus [49] Nielsen and Tahir (2002) found that of the world's 50 most popular web sites, only 84 per cent (and not 100 per cent as expected) had an "About Us" link that explained the objectives of the web site.

Web site security: trustworthiness of web site (TOW)

[63] Tilson et al. (1998) found that various factors affected customers' online purchasing decisions. They are listed in descending order of importance:

- web security for credit card transactions;

- easy return or exchange methods;

- detailed descriptions of items;

- price of items;

- privacy of personal information; and

- pictures of merchandise.

It was found that web security is the most important factor. In another study, [6] Brynjolfsson and Smith (2000) found that TOW is probably the single most important factor in EC. [38] Lee and Turban (2001) developed a model to explain EC trust for consumer internet shopping. Their research proposed that EC trust is dependent on many factors, such as the competency and benevolence of the seller; the reliability, understandability and security of the web site; consumer protection and effective legal recourse when using the web site; and success stories or referrals from peers.

[66] Turban et al. (2006) proposed that an EC security mechanism such as secure order transactions using secure sockets layer would increase the TOW, particularly of web sites that require online monetary transactions. They also proposed that TOW is increased by other factors, such as the possession of a well-known brand name or organisation and success of the initial transaction with the company. [53] Pennington et al. (2003), however, proposed that TOW is decreased by users' uncertainty towards technology, lack of initial face-to-face interactions and lack of enthusiasm among trading parties.

Interactivity of web site (IOW)

IOW refers to whether a web site encourages interaction between the business and customers, such as learning about the business products and services (e.g. e-newsletters, product demonstrations and customer forums) and providing a channel for sharing and submitting feedback (through e-mail, feedback forms, online surveys and the virtual community) ([66] Turban et al. , 2006). IOW is related to the feedback and made-for-the-medium factors (such as community, personalisation and refinement) in the MUG ([2] Agarwal and Venkatesh, 2002). All of these factors provide a sense of control for users in their use of a web site.

The internet provides a platform where different media are present simultaneously, for example the presence of text, graphics, animation, audio and video together in a single web site. [15] Dix et al. (1993) believe that such multimedia becomes a powerful tool of communication only when the users are able to assume a sense of control in their interaction with the web site. In addition, [67] Watson et al. (2000) found that attractors - web site features that attract and interact with visitors, such as games, puzzles, prizes, contests, etc. - can increase IOW.

Ease of web navigation (EWN)

EWN refers to how quickly and easily a web site visitor locates required information. Web site navigation must be easy, as users want surfing around the site to be predictable, consistent and intuitive enough to obviate the need for much thinking ([10] Cheung and Lee, 2005; [27] Iwaarden et al. , 2003). [34] Kirakowski et al. (1998) found EWN to be the second most important determinant of customer decision making. [18] Gehrke and Turban (1999) surveyed 130 e-commerce customers and found that EWN was the third most important factor. Since users rarely land directly at what they are seeking in a web site, EWN has become a fundamental factor affecting user perceptions of web usability ([48] Nielsen, 2000).

EWN is dependent on the use of navigation aids such as the navigation bar, navigation column, site map, site search, directory and hyperlinks ([66] Turban et al. , 2006; [9] Chen et al. , 1997). To avoid confusion while surfing a web site, there should be consistency in the use of navigation aids, such as use of the same format and style in the navigation aid and in approximately the same location on every webpage ([66] Turban et al. , 2006). EWN can be improved by having a wide and shallow web site structure, because this reduces the number of clicks visitors must make to access the needed information ([66] Turban et al. , 2006).

Download speed of web site (DSOW)

The issue of download speed is vital, since surfers normally turn away when they must wait too long for a web site to appear or load. [22] Helander (2000) found that download speed is one of the most important factors in whether users revisit a web site.

Certain designs or tools used in web sites (such as graphics or images requiring high bandwidth) can delay the download speed. [63] Tilson et al. (1998) included DSOW as one of the important factors affecting the usability of EC web sites. [66] Turban et al. (2006) recommended that every page on a web site should have a download speed of less than 12 seconds, because visitors are likely to hit the "stop" or "back" button and go elsewhere if they have to wait longer. [49] Nielsen and Tahir (2002) evaluated the world's 50 most popular web sites and found that the average download time was 26 seconds, much longer than [66] Turban et al. 's (2006) recommendation. Only 28 per cent of the web sites had an acceptable downloading time of >10 seconds ([49] Nielsen and Tahir, 2002).

Research framework

Figure 1 [Figure omitted. See Article Image.] shows the research framework. One dependent variable and seven independent variables were identified and tested. The independent variables were EWN, CGW, TOW, IOW, DSOW, UGM and UCF. All seven variables are believed to contribute to the OWU (dependent variable). There is one moderating variable, the web site industries. The relationship between the independent variables and the dependent variable is tested for all 36 web industries.

Research methodology

Web site evaluation method

There are two popular methods of addressing web site usability issues - heuristic evaluation and laboratory testing ([29] Kantner and Rosenbaum, 1997). The heuristic evaluation method is based on evaluation by a pool of experts who inspect and use a part of a web site to identify usability problems which they believe would affect the end-users ([47] Nielsen, 1994). In contrast, the laboratory testing method selects samples from the user population of a web site and asks the selected users to use a part of the web site and report what they think works or does not work, or are appropriate or inappropriate.

The present study has not been conducted to solve usability problems of a single web site, but rather to derive patterns of SIU perceptions of web site usability in general through empirical findings. Thus, the laboratory testing method was adopted in this study to conduct empirical research into many web sites to determine how WIUFs affect the OWU for several different industries.

Material

Based on the review of previous literature and formulation of a research framework (as presented in the previous sections), an online questionnaire was developed. The online questionnaire had two sections labelled A and B.

Section A began with questions on the demographics of the evaluators, which included gender, race and age. This section also required evaluators to choose, categorise and evaluate any web site from 36 different industries (see the Appendix). These 36 industries were included in the study because of their relationship to the needs and interests of students. This ensured that the industries were relevant to the SIUs, who assumed the role of customers. The data obtained through this section were nominal and categorical in nature, and these data were used as a moderating variable.

Section B consisted of questions on web usability factors. In this section, evaluators were required to rate the web sites they had browsed for the seven WIUFs (EWN, CGW, TOW, IOW, DSOW, UGM and UCF). Finally, they were also required to rate the usability of the web sites as a whole under the OWU. A semantic differential scale ([59] Snider and Osgood, 1969) from 1 to 5 (poor to excellent) was used to rate the WIUFs and the OWU. The data obtained were used as independent and dependent variables.

Subjects

The SIUs who participated in this study are identified as web site evaluators. They were between 18 and 21 years of age and had three to five years' experience in using the internet.

The convenience sampling technique ([43] Malhotra, 1999) was used for the selection of web site evaluators, because employing a random sampling method to select a large number of web site evaluators from across Malaysia was not feasible. The research needed to gather a large number of subjects, provide sufficient briefing and hands-on tutorials, and conduct the web site evaluations in assigned computer laboratories. Thus, 400 SIUs were selected from a large university (Multimedia University, which has about 20,000 students) located in the central region in Malaysia.

Evaluation procedure

Prior to the evaluation web site evaluators were given an hour's lecture on the definitions and literature (as presented in the previous sections) of the WIUFs, OWU and the 36 student-related industries to ensure they had a clear understanding of the research. The evaluators were also given training in evaluating web sites using the questionnaire in an hour-long, hands-on tutorial in a computer laboratory.

Each evaluator was to select about 40 different web sites for the evaluation, which would yield approximately 16,000 (400 × 40) data points. They selected web sites by querying common search engines (such as Yahoo! and Google Search) using each of the 36 industries as a keyword, then randomly choosing one or two of the search results to be evaluated. They also examined the chosen web sites' content (e.g. the business function of the company) to determine whether it matched the industry. They were trained to perform this selection procedure during the hands-on training. After selecting a web site, each evaluator spent ten minutes surfing the web pages before evaluating it. The evaluators were given freedom to choose the order in which the webpages were retrieved from a web site. They would then open another browser window to access and answer the online questionnaire and submit it to the web server. On average an evaluation took about 20 minutes, including the initial ten minutes of surfing. There were two computer laboratories (with a total of about 80 computers) dedicated for this purpose. Both laboratories shared a 10 MBPS connection to the internet. The computers used Intel Pentium microprocessors and the Microsoft Windows operating system. Qualified personnel were stationed in the laboratories to provide clarification on the questionnaire to evaluators whenever needed.

The evaluators were required to evaluate at least one web site for each of the 36 industries. To prevent a web site from being evaluated too many times, the online questionnaire was programmed to allow the same web site to be evaluated a maximum of 16 times. In the event that all the web sites selected by different evaluators were evaluated the maximum number of times allowed, the final data points would still have at least 1,000 unique web sites evaluated (16,000/16).

The evaluators were given approximately eight weeks to perform the 40 evaluations. They were advised not to perform too many evaluations per day (not more than five) to avoid threats to validity such as maturation (e.g. fatigue) and repeated testing. They were given course credit upon completing and checking the validity of the evaluations (through data filtration - see the following section). The web site evaluations were given as an assignment for an undergraduate university subject, Electronic Commerce. The idea behind this assignment was to expose students to EC web sites and the various aspects of web site design. Therefore, the evaluations served two purposes - the students' education and the present research. In this way, both data collection costs and time were saved compared with collecting data in physical shopping malls, as in [2] Agarwal and Venkatesh (2002). Also, the method employed in this research ensured that the evaluators were given sufficient knowledge to perform their evaluations accurately and that a more conducive environment was provided, i.e. in computer laboratories rather than noisy shopping malls. In addition, this approach ensured that the subjects were the right focus group, i.e. SIUs who must act as customers instead of anyone who stepped into a shopping mall who might not have any knowledge of the internet. Finally, large numbers of evaluations (more than 300) were gathered for each of the 36 student-related industries, instead of being limited to four industries with five or six web site evaluations in each industry, as in [2] Agarwal and Venkatesh (2002).

Data analysis

Prior to data analysis, data filtration was performed by removing data points with inconsistent and incomplete answers. Two questions were repeated (but rephrased in a different way) in the questionnaire to detect inconsistencies in the answers. If different answers were given, the data point was removed. In addition, evaluators were required to key in the uniform resource locator (URL) and the industry category for each web site evaluated. Every web site evaluation was checked for consistency between URL and web industry by visiting the web site. For example, if an evaluator wrongly chose the "education" industry for a URL that supposedly came under the "automotive" industry (as confirmed through browsing the web site), that data point was removed during the filtration process. After the filtration process 12,310 data points (76.9 per cent of the total data points) were finalised for statistical analysis. Among the data points, there were 5,595 unique web sites. The data points were analysed using the Statistical Package for the Social Sciences (SPSS) software. Multiple linear regression (MLR) was used to test for significance and to rank the independent variables (the seven WIUFs) in relation to the dependent variable (OWU).

Results

Predictors of OWU and industry effect on web usability

Table II [Figure omitted. See Article Image.] shows that all seven WIUFs significantly affect the OWU. UCF has the most effect on OWU for all industry categories, while the other six WIUFs differ in importance across the industry categories.

The R² value of 0.68, 0.67 and 0.70 indicates that 68 per cent (for personal services), 67 per cent (for study-related) and 70 per cent (for purchase services) of the effect on OWU is explained by the seven WIUFs tested in this study. From the standardised coefficient of ß values in Table II [Figure omitted. See Article Image.], the seven WIUFs of every web industry group were ranked, and the results are presented in Table III [Figure omitted. See Article Image.].

Discussion

Comparison of WIUFs among the three web industry groups

Table III [Figure omitted. See Article Image.] shows that different web industry categories have different sets of crucial WIUFs. Several inferences can be made from the trends in the SIU perceptions of usability shown in Table III [Figure omitted. See Article Image.]:

- Regardless of industry category, UCF is still the most important factor. This finding confirms the appeal of web site aesthetics for SIUs.

- CGW is important for all industry categories as it is among the top three most important factors in all categories.

- TOW is more important for personal services than for study-related web sites. Personal services web sites generally provide SIUs with a wide range of information important in decision making, e.g. banking and finance web sites facilitate decision making about finances; dating and matchmaking web sites provide crucial information in choosing a partner. Hence, TOW is found to be very important in such personal services web sites.

- UGM is more important for personal services and purchase services web sites than for study-related web sites. UGM and UCF are both aesthetical variables; however, for study-related web sites SIUs have distinguished them very clearly by placing UCF as the highest and UGM as the second lowest in order of importance. Perhaps, graphic and multimedia elements reduce the download speed of study-related web sites, which would be undesirable when the students do their assignments.

- DSOW is the least important factor for both personal services and purchase services, but it is the third most important factor for SIUs in study-related web sites. Perhaps, due to the heavy contents of study-related web sites, students place a higher importance on the quick downloading of information when it comes to education and studies.

- Both, IOW and EWN are found to be rather low in importance for all categories of industry. Nevertheless, they do significantly affect SIU perceptions of the OWU.

Use of colour and font

From the findings in Table III [Figure omitted. See Article Image.], UCF emerges as the most important predictor of OWU for SIUs. The importance placed on UCF is consistent across all web industry categories. This finding indicates that SIUs are attracted to the aesthetics of web sites. When comparing SIUs with general internet users, similarities are found. For example, [34] Kirakowski et al. (1998) found that the attractiveness of a web site is the most important factor for general internet users. In addition, [2] Agarwal and Venkatesh (2002) found that UCF (classified under the MUG media use factor) is an important usability factor for general internet customers.

From a previous study ([45] Nathan and Yeow, 2003), it was found that SIUs prefer the use of special fonts, such as Comic Sans and other artistic fonts, over formal fonts such as Times New Roman and Arial. This is contrary to the practice adopted by the world's 50 most popular web sites (e.g. eBay and Microsoft), 96 per cent of which used formal fonts ([49] Nielsen and Tahir, 2002). As for colour contrast in web sites, it was found that SIUs prefer lighter coloured fonts on a darker background. Again, this differs from the practice of the world's most popular web sites, of which 84 per cent used a white background ([49] Nielsen and Tahir, 2002).

Clarity of goals in web site

CGW is an important WIUF for SIUs. For a web site to be useful for SIUs, it should have clear goals with regard to what it intends to offer (e.g. online purchasing, community development or after sales services). CGW ranked second for factors affecting the OWU in purchase services and study-related web sites and third for personal services (Table III [Figure omitted. See Article Image.]). These results strengthen [2] Agarwal and Venkatesh's (2002) and [49] Nielsen and Tahir's (2002) claims that CGW is important when web users take on the general customer role. The present results confirm that CGW is important for SIUs as customers and for most student-related industries.

Some purchase services industries, such as computer hardware and software, may have a greater need for CGW because SIUs need to identify what their web sites can offer before spending more time on them, perhaps leading to an online purchase. For example, if the goal of a computer hardware web site is to sell its line of computer hardware while an SIU is looking for computer games, he/she will probably look elsewhere. Therefore, these industries have to make their goals very clear to attract the right SIU customers. This finding also concurs with the guidelines presented by [17] Gehl (2005), who stressed the importance of highlighting the goals of the web site at the top of the site to present them clearly to customers.

Use of graphics and multimedia

UGM, which is another measure of aesthetics in web sites, has varying importance for the three industry categories (Table III [Figure omitted. See Article Image.]). It is the third most important factor for purchase services, fourth for personal services and sixth for study-related industries. There seems to be an obvious connection between industries in the purchase services category and the UGM. These industries generally require much UGM, including animation, video clips, background music, etc. to make the web sites come alive and be more attractive to visitors. Pictures and graphic images of products are good drivers of online purchasing, because people want to see what they are buying - SIUs are no exception ([52] Pearsall, 2007; [62] Thomason, 2007). In addition, [24] Hoffman and Novak (1996) and [1] Agarwal and Karahanna (2000) suggested that multimedia capabilities, richness and interactivity in web sites have the potential to engage users in ways unlike other media. They totally engross users, making them forget the time and have fun shopping, which is important for the success of purchase services industries.

Despite the fact that many personal services industries, such as financial information, employment, health/medicine, management consultancies, business-to-business and stockbrokers, are generally seen as more formal in nature, SIUs do place importance on aesthetics (UCF and UGM) for the web sites of these industries (Table III [Figure omitted. See Article Image.]). Most firms in these industries offer almost similar products or services; for example, web sites under "financial information", such as Bursa Malaysia's www.bursamalaysia.com, Wall Street Online's www.wall-street-online.com and Bloomberg's www.bloomberg.com all offer financial information services. Since it has been found that SIUs see UCF and UGM as important WIUFs for personal services industries, a firm in any of these industries can make itself more appealing to SIUs by designing its web site using attractive colours, fonts, graphics and multimedia to differentiate it from its competitors.

Trustworthiness of web site

TOW is important for SIUs, as it is ranked the second most important factor affecting the OWU for personal services industries (Table III [Figure omitted. See Article Image.]). Most web sites in this category are sought by SIUs for information which can lead to important decision making. For example, online real estate agencies would require much credibility, because real estate (be it for rent, purchase or sale) usually involves significant financial commitment. Therefore, SIUs would be very cautious in choosing web sites with high trustworthiness (even though most of the monetary transactions occur offline). This finding is contrary to [2] Agarwal and Venkatesh's (2002) study, which found that TOW (known as "character strength") is important for general internet customers for industries with high outlay and non-repeat purchases. In the present study, personal services industries are information brokers, and they mostly do not involve any online purchase; nevertheless, SIUs place high importance on TOW. This is a unique characteristic of SIUs.

TOW is ranked fourth for purchase services industries. Most web sites offering purchase services generally require credit card information for their online transactions, thus the need for TOW in these industries is obvious. For example, travel companies offering travel packages, such as www.asiatravel.com (Asia Travel) and www.malaysiaairlines.com (Malaysia Airlines), require payments to be made online.

TOW is found to be relatively less important for study-related web sites. This is a very interesting discovery which is highly SIU-specific. Most study-related web sites such as those for books, child education, general education, research and online libraries are sought by SIUs for academic research and information gathering to meet their educational needs. These web sites do not aid SIUs in making critical monetary or life decisions or online purchases. Hence, SIUs do not place great importance on TOW in study-related web sites.

Interactivity of web site

IOW is found to be more important for study-related web sites (ranking fourth) than for personal services (fifth) and purchase services (sixth). [2] Agarwal and Venkatesh (2002) found that industries with transaction capabilities require more IOW than industries which use web sites as an internet presence. The present research contradicts this. SIUs seem to perceive IOW as more important for internet presence (study-related web sites) than for web sites offering transactions (purchase services). Perhaps, this can be attributed to the greater involvement of SIUs in study-related web sites than in purchase services web sites. IOW tools aid in getting SIUs involved in the web site, particularly when they are doing assignments and research. For example, the use of real-time chat, online forums, shout-boxes and blogging involves SIUs and enables them to identify with the web sites. These capabilities in study-related web sites help SIUs in their information searching and gathering processes. This contrasts with purchase services web sites, where information gathering may not be as extensive as in study-related web sites; for example SIUs only need product information (for their purchasing decisions) and not extensive literature reviews (for their assignment or research).

Ease of web navigation

EWN is ranked in the bottom three for all three industry categories, with relatively more importance in purchase services, followed by personal services and finally study-related web sites. This finding agrees with [2] Agarwal and Venkatesh (2002), who found EWN (under the ease of use category) to be moderately important across all industries. This also agrees with [34] Kirakowski et al. (1998) and [18] Gehrke and Turban (1999), who found EWN to be important for any web site, but not the most crucial factor. EWN enables users to surf web sites effectively. Web sites offering purchase services "seal the deal" when an online transaction is made. Usually, users surfing these web sites have something specific in mind that they are looking to purchase. Thus, clear navigational tools to help users locate sought information are especially important in industries where making sales is the central purpose.

Download speed of web site

DSOW is the least important factor for both personal services and purchase services but the third most important factor for study-related web sites. This is contrary to many studies, such as [63] Tilson et al. (1998), [49] Nielsen and Tahir (2002), [2] Agarwal and Venkatesh (2002) and [18] Gehrke and Turban (1999), which found DSOW to be an important or the most important factor. However, these studies were conducted a few years ago, when computers, servers and networks were slower and had low bandwidth. This contrasts with the present situation, in which internet connections (including the ones used in this research) are facilitated by fast network connections, such as broadband, leased line connection, Integrated Service Digital Network, etc. and these speed up the downloading of web sites.

The finding is also interesting in that SIUs, who downplay the importance of DSOW for personal services and purchase services industries, place greater importance on DSOW when it comes to study-related web sites. SIUs use study-related web sites for educational purposes, i.e. to conduct research and gather information. Study-related web sites such as Online Library and Research offer large files of information, which often means that downloading could take a long time. Hence, SIUs require faster access to this information.

Web site usability guidelines

Based on the research findings and the discussion above, the usability guidelines for designing SIU-centric web sites in the three industry categories are presented in Table IV [Figure omitted. See Article Image.].

Conclusion and recommendations

This study has established a research framework to measure the usability of student-related web sites from three web industry groups. The results indicate that different industry groups have different crucial WIUFs. Possible explanations for the WIUF rankings in the various industry groups have been provided. The results have also been compared with those of previous studies, including those by [2] Agarwal and Venkatesh (2002), [1] Agarwal and Karahanna (2000), [34] Kirakowski et al. (1998), [63] Tilson et al. (1998), [49] Nielsen and Tahir (2002) and [18] Gehrke and Turban (1999).

The research discovers that SIU preferences for web usability are similar to those of general internet users (researched in the previous studies) in terms of the importance of UCF, UGM, CGW, IOW, TOW and EWN. However, SIUs have a greater need for TOW than general internet users (see the discussion on TOW). In addition, SIU preferences for these usability factors differ significantly, depending on the industry group concerned.

The research also finds that DSOW is the least important WIUF (contrary to the findings of the previous studies), probably because of higher-speed internet connections nowadays. However, this is subject to the industry group - SIUs place greater importance on DSOW when it comes to study-related web sites. This finding reveals a peculiar characteristic of SIUs, who attribute greater importance to UCF and DSOW in study-related web sites, but very low importance to UGM.

The present research provides an easier method of measuring web usability through using the seven WIUFs. These factors are supported by the literature of previous studies and practitioners (since they are a part of the MUG). It also offers a method to determine which WIUFs are crucial for a specific industry group with consideration of the users. It simplifies [2] Agarwal and Venkatesh's (2002) method, thus allowing more data points to be collected.

Companies can identify the usability needs of their customers by replicating the study and using their customers as subjects to evaluate their industry's web sites. As for those involved in student-related industries, it is recommended that Table IV [Figure omitted. See Article Image.] be used as a guideline for designing more usable web sites. To increase the usability of a particular web site, designers first need to determine the industry group to which a particular web site belongs and then focus on strengthening the WIUFs that are associated with it. Furthermore, web site designers can use the seven WIUFs as a yardstick to compare their web site usability with those of their competitors. In making this comparison, designers should consider the different emphases for different web industry groups (as shown in Table IV [Figure omitted. See Article Image.]) and assign higher weightings for the higher ranking WIUFs.

Future studies can be conducted by replicating the present study to discover the perceptions of usability of various web site market segments (e.g. working adults, retirees, etc.), genders, races and countries.

The limitation of the study is that the evaluators were all Malaysians; thus generalising the findings to all SIUs may require a confirmatory study with SIUs from other parts of the world. Even though some WIUFs (e.g. in-depth information content factors) were excluded in this study (see the literature review), the results show that the majority of WIUFs that affect the usability of web sites have been addressed (high-R² values of 0.67 to 0.70 in Table II [Figure omitted. See Article Image.]).

3-6 January

24-26 September

5 June

22-27 March

10-13 August

References

1. Agarwal, R. and Karahanna, E. (2000), "Time flies when you're having fun: cognitive absorption and beliefs about information technology usage", MIS Quarterly, Vol. 24, pp. 665-94.

2. Agarwal, R. and Venkatesh, V. (2002), "Assessing a firm's web presence: a heuristic evaluation procedure for the measurement of usability", Information Systems Research, Vol. 13, pp. 168-86.

3. Allen, M. (2002), "A case study of the usability testing of the University of South Florida's virtual library interface design", Online Information Review, Vol. 26 No. 1, pp. 40-53.

4. Barnum, C. (2002), Usability Testing and Research, Longman, New York, NY.

5. Brady, L. and Phillips, C. (2003), "Aesthetics and usability: a look at color and balance", Usability News, Vol. 5.1, available at: http://psychology.wichita.edu/surl/usabilitynews/51/aesthetics.htm.

6. Brynjolfsson, E. and Smith, M. (2000), "Frictionless commerce? A comparison of internet conventional retailers", Management Science, Vol. 46, pp. 563-85.

7. Calero, C., Ruiz, J. and Piattini, M. (2005), "Classifying web metrics using the web quality model", Online Information Review, Vol. 29 No. 3, pp. 227-48.

8. Chau, P.Y.K., Au, G. and Tam, K.Y. (2000), "Impact of information presentation modes on online shopping: an empirical evaluation of broadband interactive shopping service", Journal of Organizational Computing and Electronic Commerce, Vol. 10, pp. 1-22.

9. Chen, B., Wang, H., Proctor, R.W. and Salvendy, G. (1997), "A Human-centered approach for designed world-wide-web browsers", Behaviour Research Methods, Instruments & Computers, Vol. 29, pp. 172-9.

10. Cheung, C.M.K. and Lee, M.K.O. (2005), "The asymmetric impact of web site attribute performance on user satisfaction: an empirical study", Proceedings of the 38th Annual Hawaii International Conference on System Sciences (HICSS-38), Big Island, HI, (CD Rom) .

11. Cole, M., O'Keefe, R.M. and Siala, H. (2000), "From the user interface to the consumer interface", Information Systems Frontier, Vol. 1, pp. 349-61.

12. Copas, G.M. (2003), "Can internet shoppers be described by personality traits?", Usability News, Vol. 5.1, available at: http://psychology.wichita.edu/surl/usabilitynews/51/personality.htm.

13. Cox, J. and Dale, B.G. (2002), "Key quality factors in website design and use: an examination", International Journal of Quality & Reliability Management, Vol. 19, pp. 862-88.

14. Chowdhury, S., Landoni, M. and Gibb, F. (2006), "Usability and impact of digital libraries: a review", Online Information Review, Vol. 30 No. 6, pp. 656-80.

15. Dix, A., Finlay, J., Abowd, G. and Beale, R. (1993), Human Computer Interaction, Prentice-Hall, Hemel Hempstead.

16. Ezquerra, N., Braal, L.G., Cooke, C. and Krawczynska, E. (1999), "Interactive, knowledge-guided visualization of 3-D medical imagery", Future Generation Computer Systems, Vol. 15 No. 1, pp. 59-73.

17. Gehl, D. (2005), "12 ways to increase online sales", Entrepreneur.com, available at: www.entrepreneur.com/ebusiness/ebusinesscolumnist/article79002.html.

18. Gehrke, D. and Turban, E. (1999), "Determinants of successful website design: relative importance and recommendations for effectiveness", Proceedings of the 32nd Hawaii International Conference of Information Systems, Maui, HI, available at: http://ieeexplore.ieee.org/iel5/6293/16785/00772943.pdf.

19. Ginige, A. and Murugesan, S. (2001), "The essence of web engineering - managing the diversity and complexity of web application development", IEEE Multimedia, Vol. 8, pp. 22-5.

20. Greenspan, R. (2003), "College students surf, spend", ClickZ Network, available at: www.clickz.com/stats/big_picture/demographics/article.php/1583871.

21. Greenspan, R. (2004), "Online Asians lead in income, e-commerce", ClickZ Network, available at: www.clickz.com/stats/big_picture/demographics/article.php/3326041.

22. Helander, M.G. (2000), "Theories and models of electronic commerce", Proceedings of the IEA 2000/HFES 2000 Congress, Vol. 2, pp. 770-3.

23. Hix, D. and Hartson, H.R. (1993), Developing User Interfaces. Ensuring Usability through Product and Process, Wiley, New York, NY.

24. Hoffman, D.L. and Novak, T.P. (1996), "Marketing in hypermedia computer-mediated environments: conceptual foundations", Journal of Marketing, Vol. 60, pp. 50-68.

25. Hoffman, D.L., Novak, T.P. and Venkateshi, A. (2004), "Has the internet become indispensable?", Communication of the ACM, Vol. 47 No. 7, pp. 37-42.

26. Isen, A.M. (1993), "Positive affect and decision-making", in Lewis, M. and Haviland, J.M. (Eds), Handbook of Emotions, Guilford, NY, pp. 261-77.

27. Iwaarden, J., Wiele, T., Ball, L. and Millen, R. (2003), "Applying SERVQUAL to websites: an exploratory study", The International Journal of Quality & Reliability Management, Vol. 20, pp. 919-35.

28. Iwaarden, J., Wiele, T., Ball, L. and Millen, R. (2004), "Perceptions about the quality of websites: a survey amongst students at Northeastern University and Erasmus University", Information & Management, Vol. 41, pp. 947-59.

29. Kantner, L. and Rosenbaum, S. (1997), "Usability studies of WWW sites: heuristic evaluation vs. laboratory testing", Proceedings of the 15th Annual International Conference on Computer Documentation, Salt Lake City, UT, pp. 153-60.

30. Karat, J. (1997), "Evolving the scope of user-centered design", Communications of the ACM, Vol. 40, pp. 33-8.

31. Katerattanakul, P. and Siau, K. (1999), "Measuring information quality of websites: development of an instrument", in De, P. and DeGross, J. (Eds), Proceedings of 20th International Conference of Information Systems, Charlotte, NC, pp. 279-85.

32. Keeker, K. (1997), "Improving web-site usability and appeal: guidelines compiled by MSN usability research", available at: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnsiteplan/html/improvingsiteusa.asp.

33. Keevil, B. (1998), "Measuring the usability index of your website", Proceedings of the Sixteenth Annual Conference ACM International Conference of Computer Documentation, Quebec, Canada, , pp. 271-7.

34. Kirakowski, J., Claridge, N. and Whitehand, R. (1998), "Human centered measures of success in web site design", Proceedings of the 4th Conference on Human Factors & the Web, Basking Ridge, NJ, .

35. Kumar, R.L., Smith, M.A. and Bannerjee, S. (2004), "User interface features influencing overall ease of use and personalization", Information & Management, Vol. 41 No. 3, pp. 289-302.

36. Lebhar-Friedman, Inc. (2000), "Internet commerce study reveals demographics, shopping trends", DSN Retailing Today, Vol. 39 No. 14, p. 17.

37. Lecerof, A. and Paterno, F. (1998), "Automatic support for usability evaluation", IEEE Transactions on Software Engineering, Vol. 24, pp. 863-87.

38. Lee, M.K.O. and Turban, E. (2001), "Trust in B2C electronic commerce: a proposed research model and its application", International Journal of Electronic Commerce, Vol. 6, pp. 75-91.

39. Lindgaard, G. (1999), "Does emotional appeal determine perceived usability of websites", in Straker, L. and Pollock, C. (Eds), Proceedings of CybErg: The Second International Cyberspace Conference on Ergonomics, The International Ergonomics Association Press, Curtin University of Technology, Perth, Western Australia, pp. 202-11.

40. Liu, C. and Arnett, K.P. (2000), "Exploring the factors associated with website success in the context of electronic commerce", Information & Management, Vol. 38, pp. 23-33.

41. Lohse, G.L. and Spiller, P. (1999), "Internet retail store design: how the user interface influences traffic and sales", Journal of Computer-Mediated Communication, Vol. 5 No. 2, available at: www.blackwell-synergy.com/doi/full/10.1111/j.1083-6101.1999.tb00339.x.

42. Looney, M. and Lyman, P. (2000), "Portals in higher education", EDUCAUSE Review, Vol. 35 No. 4, pp. 28-36.

43. Malhotra, N.K. (1999), Marketing Research: An Applied Orientation, 3rd ed., Prentice-Hall, Upper Saddle River, NJ, pp. 334-5.

44. Moraga, A., Calero, C. and Mario Piattini, M. (2006), "Comparing different quality models for portals", Online Information Review, Vol. 30 No. 5, pp. 555-68.

45. Nathan, R.J. and Yeow, P.H.P. (2003), "A usability study of interactivity and design of web sites as perceived by student internet users", in Khalid, H.M., Lim, T.Y. and Lee, N.K. (Eds), Proceedings of 7th South East Asian Ergonomics Society and 4th Malaysian Ergonomics Conference (SEAMEC), Hilton, Kuching, Sarawak, 19-22 May, pp. 296-301.

46. New Zealand MED (2000), E-commerce: A Guide for New Zealand Business, New Zealand Ministry of Economic Development (MED), Wellington.

47. Nielsen, J. (1994), "Heuristic evaluation", in Nielsen, J. and Mack, R.L. (Eds), Usability Inspection Methods, Wiley, New York, NY.

48. Nielsen, J. (2000), Designing Web Usability: The Practice of Simplicity, New Riders Publishing, Indianapolis, IN.

49. Nielsen, J. and Tahir, M. (2002), Homepage Usability: 50 Web Sites Deconstructed, New Riders Publishing, Indianapolis, IN.

50. Norman, D.A. (2002), "Emotion and design: attractive things work better", Interaction Magazine, Vol. 9 No. 4, pp. 36-42.

51. Pastore, M. (2000), "US college students use net for shopping", ClickZ Network, available at: www.clickz.com/stats/big_picture/demographics/article.php/432631.

52. Pearsall, C. (2007), "eBay marketing secret 54 - photos are key!", available at: www.ebay-marketing-secrets.com/.

53. Pennington, R., Wilcox, H.D. and Grover, V. (2003), "The role of system trust in business-to-consumer transactions", Journal of Management Information Systems, Vol. 20, pp. 197-226.

54. Rivero, M. (1999), "Web objectives save time, money", New Hampshire Business Review, Vol. 21 No. 27, p. 6.

55. Rosenfeld, L. and Morville, P. (2002), Information Architecture for the World Wide Web, 2nd ed., O'Reilly, Cambridge, MA.

56. Sabberwal, R. and Chan, Y.E. (2001), "Alignment between business and IS strategies: a study of prospectors, analyzers, and defenders", Information Systems Research, Vol. 12, pp. 11-33.

57. Saliba, C. (2000), "Kids, young adults now e-commerce force", E-Commerce Times, available at: www.ecommercetimes.com/story/4257.html.

58. Shenkman, B.O. and Jonsson, F. (2000), "Aesthetics and preferences of web pages", Behaviour and Information Technology, Vol. 19 No. 5, pp. 367-77.

59. Snider, J.G. and Osgood, C.E. (1969), Semantic Differential Technique: A Source Book, Aldine Publishing Company, Chicago, IL.

60. Stark, J. and Meier, R. (2001), "A longitudinal study of usage and satisfaction levels of internet shopping by college students", Journal of Computer Information Systems, Vol. 41 No. 4, pp. 65-8.

61. Teo, H.H., Oh, L.B., Lui, C. and Wei, K.K. (2003), "An empirical study of the effects of interactivity on web user attitude", International Journal of Human-Computer Studies, Vol. 58, pp. 281-305.

62. Thomason, L. (2007), Seven Ways to Increase Your Online Sales, NetMechanic, Inc., Plano, TX, available at: www.gotwww.com/tips/increase_on_line_sales.php.

63. Tilson, R., Dong, J., Martin, S. and Kieche, E. (1998), "Factors and principles affecting the usability of four e-commerce sites", Proceedings of Human Factors and the Web, Basking Ridge, NJ, 5 June, available at: http://research.microsoft.com/users/marycz/hfweb98/tilson/index.htm.

64. Tractinsky, N. (1997), "Aesthetics and apparent usability: empirically assessing cultural and methodological issues", Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, Atlanta, GA, , pp. 115-22.

65. Tractinsky, N., Katz, A.S. and Ikar, D. (2000), "What is beautiful is usable", Interacting with Computers, Vol. 13 No. 2, pp. 127-45.

66. Turban, E., King, D., Viehland, D. and Lee, J. (2006), Electronic Commerce: A Managerial Perspective, Pearson Education, Upper Saddle River, NJ.

67. Watson, R.T., Berthon, P., Pitt, L.F. and Zinkhan, G.M. (2000), Electronic Commerce: The Strategic Perspective, Dryden Press, Fort Worth, TX.

68. Xie, H.I. and Cool, C. (2000), "Ease of use versus user control: an evaluation of web and non-web interfaces of online databases", Online Information Review, Vol. 24 No. 2, pp. 102-15.

69. Yeow, P.H.P. and Nathan, R.J. (2003), "An empirical study of factors affecting the usability of web sites from the perspective of student internet users", Proceedings of the 35th Annual Conference of the Nordic Ergonomics Society (NES 2003), Reykjavik, Iceland, , pp. 339-42.

Appendix

Appendix. A total of 36 student-related web industries and their corresponding groups

Table AI [Figure omitted. See Article Image.]

  

Corresponding author

Paul H.P. Yeow can be contacted at: [email protected]

  

Illustration

Figure 1: Research framework

  

Table I: Description of WIUFs and equivalent factors in the (MUG)

  

Table II: Result of MLR test between the seven WIUFs and OWU for student-related industries

  

Table III: Importance of the seven WIUFs affecting the OWU for student-related industries

  

Table IV: Web site usability guidelines for student-related industries

  

Table AI: