MOBILE WEB USABILITY DEVELOPMENT AND MAINTENANCE Caroline Collier, Georgia College & State University, Campus Box 012 Milledgeville, GA 31061 (478) 445-5721 [email protected] Ajantha Dahanayake, Georgia College & State University Campus Box 012 Milledgeville, GA 31061 (478) 445-5721 [email protected]

ABSTRACT Web designers have to manage the design of new mobile Web sites or new versions of existing mobile Web sites in very short time periods. Due to these very short time periods, mobile Web usability is often not considered in the Web site development process. Even when the issue is considered, the usability of the mobile Web site is usually evaluated in the latter phases of the development process, when its implementation is almost complete. Therefore, correcting the detected usability errors implies a complete redesign of the application which cannot usually be afforded. This paper proposes a methodology for mobile Web usability development and maintenance. This methodology incorporates mobile Web usability throughout the mobile Web site lifecycle. Mobile Web usability appears then as a fundamental property of the mobile Web site, itself. INTRODUCTION In recent years, the Internet has become an essential means of communication. The World Wide Web is particularly used to carry out many different activities in areas such as business, leisure, learning, and so forth, causing a tremendous growth in the total number of Web sites in existence [11]. However, most of the currently existing Web sites are not usable to varying degrees by mobile device users. For this reason, mobile device users are unable to use a significant part of the information included in the Web [3]. The mobile device either cannot display the Web site at all, or cannot display the Web site in a size that fits the mobile device screen or can display the Web site, but it is difficult to interact with the Web site itself (such as difficult navigation through the Web pages of the Web site). The lifecycle of Web sites is currently very short [8]. In fact, Web designers have to manage the design of new Web sites or new versions of existing Web sites in very short time periods. This has a detrimental effect on the quality and usability of the final product. Although some research studies are being carried out on mobile Web services and mobile Web security, mobile Web usability is not being considered in the Web site development process. Moreover, when the issue is considered, the usability of a Web application is usually evaluated in the latter phases of the development process, when its implementation is almost complete. As a result, correcting the detected usability errors implies a complete redesign of the application which cannot usually be afforded. To avoid these situations, developers should consider mobile Web usability from the very beginning of the product development process.

However, a shortage of development methodologies, which incorporate mobile Web usability as a fundamental property of the product, exists. Such methodologies should be designed and implemented within organizations to increase the mobile Web usability awareness of Web developers. As a result, this should facilitate the development of usable mobile Web sites. These methodologies lead to the production of usable applications, but they would also lead to the development of higher quality products and facilitate their maintenance. In this paper, we propose a methodology for the mobile Web usability development and maintenance process. The establishment of this methodology in an organization will ensure that mobile Web usability is incorporated throughout the development of the mobile Web site. It will also increase awareness of the importance of mobile Web usability. MOBILE WEB USABILITY OVERVIEW Mobile Web Usability and Web Usability One could argue that mobile Web usability should follow the same methodology as Web usability. However, mobile Web usability diverges from Web usability, as both have different characteristics. Mobile Web users utilize mobile devices to access some Web sites whereas “typical” Web users utilize their desktop computers or laptops to surf the Web. This fundamental difference has several consequences including the fact that mobile device users face limited input and output capabilities [12]. Other consequences encompass the importance of context and goaloriented intentions of the mobile Web users. Most desktop computers are equipped with a large keyboard including 104 keys (PC English keyboard) and a mouse, whereas the input devices vary greatly from one mobile device to another one. A mobile phone uses a limited 12 button keypad and a PDA can include a touchscreen and a stylus which works in combination with the touchscreen but provides higher precision. Despite the creation of new and innovative input devices for mobile devices, it is generally easier to work with input devices designed for a desktop computer than it is with the ones for a mobile device. This has an impact on Web usability. For example, if a Web site requires many “clicks” to navigate through its Web pages, this Web site may not be user friendly or not “usable” for a mobile device user. Table 1 presents a list of input devices and their descriptions. Input devices Keypad Touchscreen Stylus Trackball Clickwheel

Description Classic 12 button mobile keypad Touchscreen that allows pointing to an area of the screen A stylus normally works in combination with a touchscreen. The stylus provides higher precision. A little sphere that acts like a trackball mouse. A wheel that is normally placed either below the screen or on the side of the device that lets the user quickly move up and down as well as clicks on link or items on the screen. Table 1: List of input devices used by mobile devices

The output capabilities of mobile devices are determined by their screen, which can range from small monochrome to VGA/SVGA size displays. The screen limits the amount of information for simultaneous display and therefore, the bandwidth for user interaction. Applications need to consider this limitation, for example, by distributing information across multiple pages or adapting the content of the application. Desktop computers and mobile devices differ in screen sizes. On one hand, while buying a desktop computer, a person looks for purchasing the largest screen his or her budget allows. On the other hand, the mobile device buyer is looking for portability and accepts a small screen size to ensure the device portability. Due to this diversity in screen size and resolution, a Web page can have different displays on a desktop computer and on a mobile device. The concept of context plays a major role in mobile Web usability. Desktop users surf the Web while seated in front of a desk. For example, they can be at home, at work, or even at a cyber café. The context can then provide diverse levels of noise and distraction. However, mobile users are more likely to use their mobile devices in “uncomfortable” and “noisy” places. This can include a bus while traveling to their workplace, at lunch in a restaurant, or even between two meetings on a business trip. The mobile devices are usually used in a context where concentration is not easy because of the level of noise, the presence of other people around the users, and the position of the user (one can be standing up in a bus while using the mobile device) [19]. Mobile users typically have different interests than users of desktop computers. Mobile users are likely to have more immediate and goal-directed intentions than desktop Web users. Their intentions are often to find out specific pieces of information that are relevant to their context. An example of such a goal-directed application might be the user requiring specific information about schedules for a journey he/she is currently undertaking. Equally, mobile users are typically less interested in lengthy documents or in browsing. The ergonomics of the device are frequently unsuitable for reading lengthy documents; users will often only access such information from mobile devices as a last resort because more convenient access is not available. The content of the mobile Web site needs to be adapted. This includes summarizing the text to reduce the text length, but keeping the main idea. It also involves shortening the words themselves. For instance, on a mobile Web site, the word “Entertainment” could be replaced by a shorter word such as “Fun” [12].

Mobile Web Initiatives As a first step towards mobile Web usability, the W3C has published the Mobile Web Best Practices 1.0 on November 2, 2006 [20]. A later version was released on July 29, 2008. The document was created by the Mobile Web Best Practices Working Group as part of the Mobile Web Initiative. The scope is based on the fact that the quality of the user's Web experience via a mobile device depends significantly on the usability of Web sites, the browsers, and the device itself. Although the document recognizes that browser usability and device usability are important (for reading, navigating, and interacting with content), it focuses primarily on Best Practices for improving site usability. It presents 60 recommendations for delivering Web content to mobile devices. The Mobile Web Best Practices Working Group briefly describes each recommendation and answer succinctly to how to do it and how to test it. In the “how to test” section, the Group differentiates human test from machine test. These recommendations are an important part of mobile Web usability. However, nowadays, Web engineers have to manage the creation of new Web sites or new versions of existing Web sites in very short time periods. Consequently, the conformance to these recommendations is often overlooked, which has a detrimental effect on the quality of the final product. MOBILE WEB IN REAL LIFE: TWO CASE STUDIES Two case studies are used to investigate mobile Web usability in practice. The underlying motives for this investigation are to derive starting points for the development and maintenance of mobile Web usability. The research method is exploratory in nature. We use interviews (by phone and emails), review Web sites, and search the Internet. Alpha University Alpha University is a state university located in the southern region of the United States. The competition between universities in this region is fierce. Each university wants to attract and retain the brightest students and faculty, as well as raise donations from private companies and alumni to support their expansion and improvements. Therefore, Web sites are powerful promotional tools. Alpha University understood the power of the Web and did a major redesign of its own Web site in 2006. The Alpha University Web site attempts to reach a diverse audience. It is seen as a promotional tool to attract new potential students (and their parents), as well as new faculty and staff members to fill open positions. It is also an informational tool to inform its current students, faculty, and staff. Finally, it is a communication tool to keep in touch with alumni and friends as the university needs to rely more and more on external donations. As a result, there are 17,000 Web pages (including Internet and Intranet sites) and almost 60,000 files in the Alpha University Web site. The Alpha University Webmaster, along with the team of the University Communication division, supervises the updates of the University Web sites. To make sure that the Web site stays updated, the Webmaster delegates the updates of several parts of the Web site to 158 content managers. A content manager is a person who works in a specific school or department and is aware of changes at the school or department level which impacts the content of the Web

site. Content Managers update the HTML documents they are responsible for and post the modified Web pages to a development server. They then email the University Webmaster who checks the updated files on a development site for proper HTML contents, respect of the University style guidelines, and proper content. Once the University Webmaster approves the updated files, he/she transfers them from the development site to the public site. While the interest in Web accessibility has been a concern for Alpha University since 2005, the interest to make their Web site usable on mobile devices is very recent and is still at a very early stage. One staff member (M.M.) working with the Webmaster attended an Apple conference in November 2006. Part of the conference was dedicated to the mobile Web and how to create a Web site that is accessible and usable on mobile devices. M.M. gained awareness of the importance of mobile Web and how it is different from the original World Wide Web. As a consequence, he started to share his new knowledge with the Webmaster and built a prototype of the Alpha University employee directory as a separate mobile Web site. The Web site was developed using an application based on AJAX open source code. M.M. transformed the static application into a dynamic application in order to use the existing database of employee’s contact information which is also used for the Web version of the employee directory. For testing and evaluation, the following was performed: • M.M. tested the Web site using his own iPhone and his co-worker’s HP IPAQ. • M.M. sent a message to the Alpha University iDreamer list serv. The iDreamer list serv groups Alpha University faculty and staff that have an interest in iPod and other mobile devices. M.M. did receive feedback from three different iDreamers. They were positive and appreciated the mobile Web site. However, the testing has been limited to these three positive comments. M.M. recognizes the feedback from only three different iDreamers is very restricted and that then more testing is needed. Moreover, the informal feedback did not mention crucial information such as the types of mobile device and browser used. He identifies the following challenges in the quest for mobile Web usability: • Lack of awareness: Staff needs to be aware of the mobile Web and its possibilities. M.M. is so far the only Alpha University staff member who attended a mobile Web conference. Awareness is still an issue for the mobile Web. • Lack of training: Staff needs to be trained to create usable mobile Web application. As the University Communication staff masters the traditional Web technology and concepts, this expertise is limited to this area and doesn’t include the mobile Web area. • Lack of methodologies: A precise methodology needs to be established to develop usable mobile Web applications and maintain them. Mobile Web usability evaluation should provide accurate results so a methodology needs to establish how and when the evaluation needs to take place. • Lack of time: The lifecycle of a Web site should be short and ensuring the mobile Web usability should not be a lengthy process. The staff is busy with the original Web site. C University C University (CU) is a private Christian University located in the United States. CU students may choose from 61 baccalaureate majors that include more than 100 areas of study, 26 master's

degree programs, and one doctoral program. CU accounts 4,800 students including 630 graduate students from across the U.S. and 60 nations. As it states in its vision, CU will build distinctive and innovative programs. One of these innovative programs is the CU Connected program. Thanks to this program, CU has provided an iPhone or iPod Touch to all incoming freshmen starting Fall semester 2008. This is the beginning of a large project. The mobile devices will ultimately be used to provide: • Interactive campus map ( 3-D map that will progress as you walk) • University Calendar • Course Registration • Food order (including payment online) • Social networking such as Facebook • Weather • Voice reporter • Podcasts from professors • Entertainment Web sites (i.e. buy a concert ticket online) • Class conversations (Hybrid classes) • Class Syllabus • Possibility to answer online surveys posted by professors CU has adopted the idea proposed by the 2008 Horizon Report which states: "As new devices… are released that make content almost as easy to access and view on a mobile as on a computer, the demand for mobile content will continue to grow. This is more than merely an expectation to provide content: this is an opportunity for higher education to reach its constituents wherever they may be." Almost all of these project goals require making mobile Web applications usable. The ideas to feed this project initiate from a group of eight teams: •

•

• • • •

•

The Administrative & Infrastructure Team: It examines ways university administrators can leverage converged devices for standard administrative tasks and revenue development; The Application & Programming Team: It develops front-end and middleware applications to help users take full advantage of the new capabilities offered by converged devices; The Digital Media Interaction Team: It explores ways that campus media can use new strategies to inform the university population about news and events; The Living & Learning Team: It considers the social applications of converged devices for residence life and student affairs; The Pedagogical Innovations Team: It develops new teaching and active-learning strategies based on ubiquitous deployment of converged devices; The Podcasting Group: It is a separate university initiative, which is in charge of coordinating its efforts with the other mLearning teams to think about mobile devices as a platform for educational media; The Social Interactions Team: It seeks ways to use converged devices to strengthen student and faculty community;

•

•

The Student Research Team: It gathers requests for features and applications from students and also relays student responses to applications being developed by the other teams; The Study Coordination & Invention Team: It offers synthesis and coherence among the teams and providing broad theoretical direction.

As the teams have specific assignments, none of the team is officially in charge of the mobile Web usability. One of the Web programmer leaders (J.L.) has been instrumental in the development and implementation of this project. He is part of the Application & Programming Team. This team gets the requirements from the other teams and is in charge of their ultimate development and maintenance. The mobile Web applications are tested by the experts of the team first. They are then demonstrated to the team which provided the requirements for the application and from which the idea was initiated. Comments and feedback are reported and potential modifications are made. If modifications occur, a new version of the application is developed and each version is presented to the initial team until acceptance. Once the application is approved, the application is tested on a group of CU students. Comments and feedback are recorded and potential modifications are made here again before releasing the application to the entire university population. J.L. considers that the process works well but also mentions the process faces some challenges: • Lengthy and chaotic process: The development process can be lengthy and chaotic. J.L. agrees that a formal methodology will be helpful to organize the development process itself. • Lack of reporting standards: J.L. recognizes that there are some reporting issues since comments and feedback are open and sometimes unclear. A more detailed and standardized reporting method is needed. These reporting standards should be incorporated in the methodology. Conclusion In both case studies, the lack of methodology for the development and maintenance of mobile Web usability constitutes a challenge. Both case studies would benefit from a methodology which could incorporate mobile Web usability throughout the lifecycle of the Web site and streamline the development and maintenance processes.

ENGINEERING MOBILE WEB USABILITY Software Engineering Methodologies Developing Web applications is a complex process which requires adherence to a particular framework or methodology in order to produce good quality products. These methodologies define concrete steps of the development process in order to reduce its complexity. Software engineering methodologies describe the phases required for software application development, as well as the way in which these phases are integrated into the process model [13]. The most commonly applied software engineering models are the following: • Waterfall model • Prototyping model

•

Spiral model

The waterfall model [1] is a sequential methodology which establishes a fixed order for the development process. This specific methodology divides the software methodology into six activities: 1. system engineering 2. analysis 3. design 4. coding 5. testing 6. maintenance The prototyping model [16] is based on producing prototypes of the software application in a short period of time. The prototypes are then tested and the results are used to produce subsequent enhanced prototypes until all the requirements are fulfilled. The spiral model [16] integrates features from the waterfall and the prototyping models and adds an element of risk analysis in the development process. The process model is configured as a spiral where each iteration consists of four major activities: 1. planning 2. risk analysis 3. engineering 4. customer evaluation In all these different models, three main phases are always included: 1. definition 2. development 3. maintenance The definition phase consists of requirements definition and application analysis whereas the development phase covers the design and implementation of the application [16]. Together, these phases constitute the lifecycle of the application. Selecting one or other methodologies depends on the size and complexity of the software application to be developed, as well as on time constraints. However, software engineering methods cannot be directly applied to Web application development due to the special features of hypermedia. In order to overcome this situation, some new methods have been proposed. These include hypertext design model (HDM) [2] and relationship management methodologies (RMM) [5]. In addition, different methodologies (frequently sharing some features such as some lifecycle stages) have been produced: sequential, iterative, prototype-based, and so forth. Incorporating Mobile Web Usability into the Lifecycle We suggest adopting the following lifecycle phases for a Web application: requirements, analysis, design, implementation, and maintenance. This section describes each lifecycle phase and explains the decisions to be made within each lifecycle phase.

A preliminary evaluation of the company must be performed in the product requirements phase. This preliminary evaluation will determine whether the company is able to proceed to the design of usable mobile Web applications. It aims to detect any limitations regarding the company’s staff, technical, and economical resources in the mobile Web usability area. As a result, the company’s ability to deal with upcoming problems is measured and the most deficient areas are identified. One of the main obstacles in this stage is the lack of trained staff. Solving this problem may require additional financial resources, and the company would have to weigh the benefits of a usable mobile Web site with the financial burden associated with it. In addition, the available of the trained staff to conduct manual evaluations need to be taken into account. Besides, the company needs to recruit mobile users to perform mobile Web usability evaluation. Thus, it is desirable to include groups of users with different mobile devices, various browsers, and diverse levels of familiarity with the Web site. It is also important to make the users interact with their mobile devices in different contexts. Adopting the Mobile Best Practices 1.0 helps identifying usability obstacles in the analysis phase. These guidelines are aimed at avoiding unnecessary design barriers and promoting the application of user characteristic specifications and their operational environment in a positive way. The results obtained in the analysis phase are formalized in the design phase. Applying techniques related to the previously adopted Mobile Best Practices 1.0 is advisable at this stage, as it gives some orientation to build the mobile Web site and avoids pitfalls that will be source of a non-usable Web site. These techniques are intended to be independent from the development technology used. The next activity to be addressed is the implementation phase. It is essential to remember the universal design principles in order to make appropriate decisions. When producing content in this phase such as images, tables, and forms, it is necessary to provide equivalent content if appropriate. This requirement leads to the fulfillment of some of the most relevant requirements in the Mobile Best Practices. For instance, in the case of forms, entering data and text is a very time consuming and error-prone task for mobile device users. Everything possible should be done to minimize the amount of clicks required by users. This may imply having to do extra effort on the application back-end to collect user-preferences and use the information to provide reasonable defaults and pre-filled forms [12]. The idea is to produce equivalent content that provides the user with the same functionality as the original. The evaluation phase is one of the most relevant phases in the usable mobile Web application lifecycle. Results of the design and implementation phases are checked to evaluate the fulfillment of the specifications. In order to reach a reliable conclusion, manual evaluation by experts, mobile device and browser emulators, and Mobile Web Best Practices checkers must all be applied. After removing all usability barriers detected by automatic evaluation tools and by experts, evaluations with users will help to detect the remaining user-specific obstacles. All these evaluation methods are complementary and necessary. However, evaluations with mobile device users should not be carried out as frequently as automatic ones. In this stage, it is useful to document process and results in a report in order to avoid repeating the same mistakes. A detailed report aimed at improving the development process. Producing good quality products should also be an objective of the development process. Maintenance is a critical phase in any Web site lifecycle. As the Web is essentially dynamic, its contents change frequently. From the mobile Web usability point of view, the maintenance phase is understood as the phase where usability is monitored in order to measure its evolution.

When the content of a mobile Web application is updated, it is difficult to say whether its usability level has increased or decreased since up to now only qualitative metrics have been used. It is advisable to use quantitative metrics, in order to measure more accurately the usability level and its evolution through time. Whatever the evolution is (positive or negative), the decisions made and the factors involved must be reported and reviewed. An application which reports these facts and evaluates mobile Web usability quantitatively with a predetermined frequency is a powerful and essential tool within the context of this lifecycle. The main objective of taking into account mobile Web usability issues during the whole lifecycle is to improve mobile Web usability so a broader range of end users can be reached.

PROCESS MODEL FOR USABLE MOBILE WEB APPLICATION DEVELOPMENT AND MAINTENANCE Mobile Web applications tend to be implemented in short time frames and due to this feature a specific development methodology is necessary. This methodology should clearly define the decisions to be made in each phase. The lifecycle is frequently defined as a group of phases that do not have to follow a determined sequence. However, a process model defines the sequence followed by the phases of the lifecycle [17], and it is important to select an appropriate one for the development of usable mobile Web applications. As it is the case for Web accessibility [7], iterative methodologies fit better when developing usable mobile Web applications. The iterative process model, in contrast to the classical waterfall model, enables the development of first prototypes in the earlier phases of the process. This feature facilitates mobile Web usability during the whole development process. Consequently, usability errors are easier to find and repair. As a result, errors are not passed to subsequent phases and similar errors are avoided in the rest of the process. It is beneficial to use a platform for the reporting of errors detected. In this way, keeping track of errors improves the development process as it avoids the same errors being made again. In order to determine the appropriate process model it is necessary to take into account that a company may face two possible scenarios: (1) development of a new Web application or (2) mobile Web usability improvement of an existing application. An initial evaluation has to be performed in order to improve the mobile Web usability of an existing Web application. In this way, usability problems will be detected in order to analyze and correct them and avoid passing them on to subsequent phases. Solutions to the errors detected will be implemented in an iterative way, allowing the correction of errors and the “cleansing” of the application. Both scenarios (creation of a new application or improvement of an existing application) need to predict and plan mobile Web usability evaluations during the development process. Once the application has been implemented, an evaluation is performed. According to the results, if the objective proposed in the specifications is not met, a re-analysis is carried out. The application will then be redesigned. Following this methodology implies the improvement of the prototype with each iteration. When the required usability level has been fulfilled, the development phase will finish and the maintenance phase will start. In the maintenance phase, periodical usability evaluations have to be made in order to know whether the updates made to the application have a detrimental effect on the required usability level. If these updates have decreased the overall application usability level, the evaluation report must be analyzed and the detected errors fixed by designing and implementing new solutions.

Requirements Definition Manual Expert Evaluation New Web Application ?

No

Evaluation

Yes

Automatic Checkers

User Evaluation

Analysis

Design

Implementation

Manual Expert Evaluation No

Specifications Fulfilled?

Evaluation

Yes Maintenance

Figure 2: Mobile Web Usability - Process Model

Automatic Checkers

User Evaluation

MOBILE WEB USABILITY EVALUATION Mobile Web usability evaluation is an essential component of the usable mobile Web development process. Mobile Web usability evaluations should be performed frequently throughout the development process as described in the previous section. Different methods can be used to carry out these usability evaluations, such as manual evaluation by experts, automatic validation of the mobile best practices, and mobile usability testing with users. A valid mobile Web usability evaluation methodology should combine all these methods. However, the most widespread practice in the Web developer community is to perform a manual evaluation by the Web developers themselves. The developers may use a few mobile devices (frequently their own mobile devices) and check on a few browsers (the browsers they usually used on their own mobile devices). At Alpha University, two mobile devices and two different browsers were used to perform the evaluation of the employee directory mobile Web site. The evaluation appears to be incomplete and subjective as it was performed by the Web application developer himself. This incomplete evaluation resulted in products with poor usability levels. Manual Evaluation by Experts As with Web usability (by this term, we mean the usability of Web sites created in the attempt of being seen through desktop computers), mobile Web usability evaluation requires performing inspection methods. In this way, heuristic evaluation can be extremely useful as experts are able to evaluate mobile Web applications according to sets of mobile best practices [4]. Some of these practices can not be automatically tested as they require human judgment. For instance, Mobile Web Best Practice # 7: BALANCE – “Take into account the trade-off between having too many links on a page and asking the user to follow too many links to reach what they are looking for” [20]. The idea behind this practice is to design the mobile Web application so that frequently accessed information is easily reached with a minimum number of page retrievals. Navigation to less frequently accessed information may take more retrieval as a result. A guideline is that users become frustrated if it takes more than four retrievals to reach their objective. Whether this can be achieved depends on the nature of the site and, in particular, how items in menus are grouped together to provide understandable themes. This practice requires human judgment to decide what a good balance is for a particular Web site. Another convenient technique is performing a walk-through in order to detect any usability barrier which obstructs the completion of specific tasks [9]. In this way, experts are able to determine the main executable tasks in a Web site and browse all the particular solution paths by using different types of mobile devices such as iPhone, Nokia, etc. It is possible to use the mobile devices themselves but also, experts could rely on device and browser emulators. As this list is not complete, it constitutes a start and addresses the most popular mobile devices emulators and their browsers simulators.

Simulators/Emulators Windows Mobile Emulator

URL http://www.microsoft.com/downloads/details.aspx?Family Id=C62D54A5-183A-4A1E-A7E2CC500ED1F19A&displaylang=en iPhoney: iPhone Emulator http://www.marketcircle.com/iphoney/ Opera Mini Emulator http://www.operamini.com/ Blackberry Emulator http://na.blackberry.com/eng/developers/downloads/simul ators.jsp ACCESS NetFront http://www.accesscompany.com/products/netfrontsdk/index.html OpenWave Simulators http://developer.openwave.com/dvl/tools_and_sdk/phone_ simulator/ Nokia Simulators http://www.forum.nokia.com/info/sw.nokia.com/id/db2c6 9a2-4066-46ff-81c4caac8872a7c5/NMB40_install.zip.html Table 2: Mobile Devices and Browsers Simulators/Emulators Automatic Mobile Web Best Practices Validation The advantages of this method are several in terms of cost effectiveness since free automatic mobile Web Best Practices evaluation tools, which obtain the evaluation results in a short period of time, are used. Currently, there are a few mobile Web checkers with diverse characteristics. Checker URL Characteristics MobileOK Basic http://validator.w3. MobileOK Basic Checker tests the Checker org/mobile/ compliance of a Web site to the Mobile Web Best Practices published by the W3C Mobile Web Best Practices Working Group. The current checker is at its beta stage and is asking for comments from the public. There are 2 levels of conformance, basic and pro. TAW mobileOK Basic http://validadores.ta TAW OK Basic is a tool for analyzing wdis.net/mobileok/ mobile web best practices. Specifically en/ it checks mobileOK Basic conformance as described in W3C mobileOK Basic Tests 1.0 based on W3C Mobile Web Best Practices 1.0. Ready.mobi http://ready.mobi/ This testing tool evaluates mobilereadiness using industry best practices & standards. The free report provides both a score (from 1 to 5) to determine how well your site performs on a mobile device. Table 3: Mobile Best Practices Checkers

The checkers execute their validation online on a Web server. However, the lack of methods for validating these tools creates a number of problems and a number of experts criticize the Mobile Best Practices 1.0 and the associated checkers. These experts define the practices as “compromises” [6] [12]. The main objective of these mobile Web checkers is to verify the content of a Web page or Web site according to a set of guidelines and to return a report, detailing all errors discovered. They offer guidance on error correction by providing the Mobile Best Practices associated with the detected errors. However, the guidance is limited to the display of the guidelines as shown in the W3C’s Mobile Best Practices document and sometimes is illustrated by an example on how to correct the error. The “repair” service doesn’t provide a custom solution for repairing the specific error of the tested Web site. Mobile Web Usability Testing with Users An accurate mobile Web usability evaluation methodology requires testing the Web application with different groups of users. This method will detect real usability barriers for the end users. This process is even more significant for achieving the overall goal of mobile Web usability when expert evaluation has been performed by people involved in the development of the Web application, as they will be accustomed to the interface features. A typical usability test is usually carried out in controlled environments such as testing laboratories where experts can observe and collect data from users. The thinking-aloud technique, consisting of users continuously vocalizing their thoughts, feelings, and opinions while interacting with the site, is very useful since it allows the detection of barriers found by users in real time. However, we argue that in the case of mobile Web usability testing, a laboratory test is not appropriate. Mobile Web applications are used in very different environments For instance, the user may be travelling. In fact, the user accesses the application whenever and wherever required to perform a certain task. As a result, a mobile Web application is in prone of continuous context changes. The context influences the application, which needs to adapt to those context changes [18]. The term context should be understood as extending beyond simple geographical location to include the physical environment, users or participants, the activities in which they are involved, and the interaction between the three [15] [19]. Figure 3 provides a graphical representation of the context model and show how the environment, the participants and the activities interact with each other as well as with themselves respectively. In addition, Table 1 offers a better understanding of these three categories by presenting their characteristics.

Figure 3: Graphical Representation of Context Model (adapted from [19]) Category Environment Participants Activities Interactions

Representative Characteristics Location, Orientation (of objects) Physical properties, Brightness and noise levels Availability, quality (of devices and communications) Location, Orientation Personal properties (e.g., age, gender, education, preferences) Mental state, Physical health, Expectations Tasks and goals (of participants) Events in the environment (e.g., weather) Co-location, Group dynamics, Social situations Participant/environment relationships (e.g., worker/workplace) Season, time-of-day, day-of-the-week In mobile and ubiquitous computing, the notion of context is often equated

Table 4: Representative Characteristics for the Context Model [19] Context in mobile Web usability evaluation is a primordial concept in mobile Web usability as it influences the user’s attention level. The mobile Web application may not be the primary focus of the user’s activities as the user may be trying to juggle interaction with a mobile device along with other elements in the environment (e.g., riding a bicycle with friends on a busy street while receiving directions from a navigation system). The amount of attention that a user can give to a mobile application will vary over time, and a user’s priorities can also change unpredictably. Moreover, various tasks can be set up in order to encourage users to browse the system, and it may be useful to collect data from this interaction so that usability parameters, such as effectiveness in completing the tasks, can be calculated. If the effectiveness value in performing

a specific task is low, its solution paths should be analyzed in order to detect any existing usability barrier. Enquiry methods such as questionnaires and interviews are broadly used in usability testing and can also be applied to mobile Web usability testing. The questions within these questionnaires should be designed in such a way that users’ answers help evaluators to determine the most significant mobile Web usability barriers in the system. The results of the complete mobile Web usability evaluation must be summarized and documented in a report. This report should contain all the detected errors, clearly identifying the method used for the validation as well as the list of the evaluated Web pages. If an area of the Web site was not part of the evaluation, it needs to be mentioned in the report. The more detailed the evaluation report is, the easier the detection and corrections of future mobile Web usability will be. A standard organizational template for developing this report can be useful and facilitate the production of accurate results for the evaluation process. The organization is faced with two options. The first option consists on the creation of a customized organizational template. The organization then decides to create its own template. The second option consists on the adoption of an existing template such as the Common Industry Format (CIF) for Usability Test Reports [10]. CONCLUSION Nowadays, as the number of mobile device users increases, Web engineers cannot ignore the importance of mobile Web usability. Our mobile Web usability development and maintenance methodology proposes a process model for usable mobile application development and maintenance. The mobile Web usability is incorporated throughout the mobile Web site lifecycle. It becomes a fundamental property of the mobile Web site itself. The methodology also describes the components of a complete mobile Web usability evaluation: manual expert evaluation, automatic checker evaluation, and user evaluation. At a larger scale, our research attempts to educate and build outreach materials to help Web engineers achieve a better understanding of the complimentary aspects of an effective mobile Web usability solution. However, this is only one step and other steps are needed to lead a more widespread development and maintenance of usable mobile Web sites.

REFERENCES [1] Boehm, B.; Port, D.; and Al-Said, M. (2000). Avoiding the Software Model Clash Spiderweb. Volume 33, Issue 11 (November 2000). Pp. 120-122. [2] Garzotto, T.; Paolini, P.; and Schwbe, D. (1993). HDM-a model-based approach to hypertext application design. ACM Transactions on Information Systems. Volume 11, Issue 1. Pp. 1-26. [3] Harper, S. (2008). Mobile web: reinventing the wheel? ACM SIGACCESS Accessibility and Computing. Issue 90. Pp. 16-18. [4] Hvannberg, E.; Law, E.; and Larusdottir, M. (2007). Heuristic evaluation: Comparing ways of finding and reporting usability problems. Interacting with Computers. Volume 9, Issue 2. Pp. 225-240. [5] Isakowitz, T.; Storh, E., and Balasubramanian, P. (1995). RMM: A methodology for structured hypermedia design. Communications of the ACM. Volume 38, Issue 8. Pp. 34-44. [6] Moll, C. (2007). Mobile Web Design. Cameron Moll. First Edition. [7] Nielsen, J. (2001). The usability lifecycle. Retrieved on April 12, 2008, from http://www128.ibm.com/developerworks/library/it-nielsen3/ [8] Nielsen, J. and Loranger, H. (2006). Prioritizing Web Usability. New Riders Press, Berkeley CA. [9] Nielsen, J. and Mack, R. (1994). Usability inspection methods. New York: John Wiley & Sons. [10] NIST Industry USability Reporting (2001). Common Industry Format for Usability Test Reports. Retrieved on May 12, 2008, from http://www.idemployee.id.tue.nl/g.w.m.rauterberg/lecturenotes/Common-Industry-Format.pdf [11] Olssen, C. (2000). To measure or not to measure: Why web usability is different from traditional usability. Proceeding of WebNet2000. [12] Passani, L. (2008). Global Authoring Practices for the Mobile Web. Retrieved in April 5, 2008 from http://www.passani.it/gap/. [13] Pressman, R. S. (1992). Software engineering: A practitioner’s approach. New York: McGraw-Hill. [14] Rubin, J. (1994). Handbook of usability testing: How to plan, design and conduct effective tests. New York: John Wiley & Sons.

[15] Ryan, C. and Gonsalves, A. (2005). The Effect of Context and Application Type on Mobile Usability: An Empirical Study. The 28th Australasian Computer Science Conference. Volume 38. [16] Sommerville, I. (2004). Software Engineering. Boston : Pearson/Addison-Wesley, 2004. [17] Scacchi, W. (2001). Process Models in Software Engineering. Final version appeared in Marciniak, J. (Ed.) Encyclopedia of Software Engineering 2nd Edition. John Wiley and Sons, Inc. New York. [18] Spriestersbach, A. and Springer, T. (2004). Quality Attributes in mobile Web Application Development. Lecture Notes in Computer Science (LNCS). Volume 3009. Pp. 120-130. [19] Tarasewich, P. (2003). Mobile commerce opportunities and challenges: Designing mobile commerce applications. Communications of the ACM. Volume 46, Issue 12. Pp. 57-60. [20] W3C (2008). Mobile Web Best Practices 1.0. Retrieved in August12, 2008, from http://www.w3.org/TR/mobile-bp/