This document was uploaded by user and they confirmed that they have the permission to share
it. If you are author or own the copyright of this book, please report to us by using this DMCA
report form. Report DMCA
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 2 of 67
Abstract: This research is aimed to generate different usability issues and guidelines for WAP Applications. It starts with problem findings in WAP applications from different sources and then we analysed all those problems and generated usability guidelines and issues, concentrating on user interface issues. To evaluate the research we have performed an experiment with some participants and also evaluated the applicability for these guidelines.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 3 of 67
Table of contents Author: ..............................................................................................1 Dhaval Vyas, (MSc Computer Science)............................................1 Table of Figures:.................................................................................7 Figure-1...............................................................................................8 Mobile devices.....................................................................................8 Figure-2...............................................................................................8 Structure of the research...................................................................8 Figure-3...............................................................................................8 Model for efficient WAP-browsing...................................................8 Figure-4...............................................................................................8 Flexibility in Navigation.....................................................................8 Figure-5...............................................................................................8 An example proxy mechanism..........................................................8 Figure-7...............................................................................................8 An example of wizard........................................................................8 Figure-8...............................................................................................8 An Example of Links..........................................................................8 Figure-9...............................................................................................8 An Example of Discoverability..........................................................8 Figure-10.............................................................................................8
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 4 of 67
An example of Vertical scrolling.......................................................8 Figure-11..............................................................................................8 An example of Accredited Information...........................................8 Figure-12.............................................................................................8 An example of Design of Evaluation.................................................8 Figure-13.............................................................................................8 An Example for visibility...................................................................8 Chapter 1: Introduction.....................................................................8 1.1 OVERVIEW OF USABILITY:....................................................................................................................................9 1.2 OVERVIEW OF MOBILE DEVICES AND THEIR USABILITY:...........................................................................................10 1.3 APPROACH USED IN THE RESEARCH......................................................................................................................11
Chapter 2: - Problem Gathering in WAP.......................................12 2.1 INTRODUCTION:................................................................................................................................................13 2.2 EXECUTIVE SUMMERY:......................................................................................................................................13 2.3 PROBLEMS IN WAP:.........................................................................................................................................13 2.3.1 User Personal Problem........................................................................................................................14 2.3.2 General Usability Problems.................................................................................................................14 2.3.3 Interface Problems................................................................................................................................15 2.3.4 Application Specific Problems..............................................................................................................17
Chapter 3: Usability Guidelines & Issues......................................17 SECTION 1: GENERAL USABILITY.............................................................................................................................18 3.1.1 Introduction:.........................................................................................................................................18 3.1.2 User-oriented issues:............................................................................................................................18 3.1.2.1 Know the user concerns: .......................................................................................................................... .....18
Speech Recognition.......................................................................................................................................20 Scrolling.........................................................................................................................................................20 Entering Numbers..........................................................................................................................................20 Entering Text..................................................................................................................................................20 3.1.2.2 User Tasks: .................................................................................................................................. .................20 3.1.2.2.1 Identify user activities:............................................................................................................. ..............21 3.1.2.2.2Classify activities:.................................................................................................................... ...............21 3.1.2.3 Design interface according to user activities:.............................................................................................. ...22
3.1.2.3 Other usability support:....................................................................................................................23 3.1.2.3.1 Minimise the user's memory load............................................................................................. ..................23 3.1.2.3.2 Consistency........................................................................................................................... .....................23 3.1.2.3.3 Feedback.................................................................................................................................................... .24 3.1.2.3.4 Internationalisation:........................................................................................................................ ............24
3.1.3 System-oriented issues:........................................................................................................................25 3.1.3.1 Iterative design and usability testing:......................................................................................................... ....25 3.1.3.2 Compatibility Issue:.................................................................................................................................... ...25 3.1.3.3 System Architecture Issue:............................................................................................................................. 26
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Chapter 4: Evaluation......................................................................52 4.1 INTRODUCTION:................................................................................................................................................52 4.2 METHODOLOGY:...............................................................................................................................................53 4.3 DESIGN OF EVALUATION:....................................................................................................................................53 Initial Interview................................................................................................................................................ .........53 User Tasks.................................................................................................................................................... .............53 Final Interview.......................................................................................................................................... ................53
4.3.1 Initial Interview:...................................................................................................................................54 4.3.2 User Tasks:...........................................................................................................................................54
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 6 of 67
4.3.3 Final Interview:....................................................................................................................................54 4.4 ANALYSIS:.......................................................................................................................................................54 4.4.1 Visibility of system status......................................................................................................................54 ................................................................................................................................................................ ..........55 Figure: 13 Example for visibility....................................................................................................... ...................55
4.4.2 Match between system and the real world............................................................................................55 4.4.3 User Control and freedom....................................................................................................................55 4.4.4 Consistency and standards...................................................................................................................56 4.4.5 Error prevention...................................................................................................................................56 4.4.6 Recognition rather then recall..............................................................................................................56 4.4.7 Flexibility and efficiency of use............................................................................................................57 4.4.8 Aesthetic minimalist design..................................................................................................................57 4.4.9 Help users recognise, diagnose and recover from errors.....................................................................58 4.4.10 Help and documentation.....................................................................................................................58
Chapter 5: Final Discussion & Conclusion:...................................58 Bibliography:....................................................................................60 Appendix A: Personal Information................................................65 Appendix B: User Tasks & Results.................................................65 Occupation or Work ........................................................................66 LESS THEN A YEAR
Which of the following area services you use more?.....................66 How long have you been using the Mobile phone?.......................66 Less then 1 2-3 years more then 4 years.....................................66 Have you used any WAP phones before?.......................................66 YES NO...............................................................................................................................................................66
Do you have any knowledge about HCI or Usability studies?.....66 Not heard Just heard
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 7 of 67
Table of Figures:
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Figure-1
Mobile devices
Figure-2
Structure of the research
Figure-3
Model for efficient WAP-browsing
Figure-4
Flexibility in Navigation
Figure-5
An example proxy mechanism
Figure-7
An example of wizard
Figure-8
An Example of Links
Figure-9
An Example of Discoverability
Figure-10
An example of Vertical scrolling
Figure-11
An example of Accredited Information
Figure-12
An example of Design of Evaluation
Figure-13
An Example for visibility
Page 8 of 67
Chapter 1: Introduction
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 9 of 67
This dissertation is an Integrative Research on investigating different usability issues in WAP headsets. Many people have witnessed the failure of the initial version of WAP, in 2000. This dissertation is aimed to investigate different human computer interaction (HCI) issues in WAP to make it usable. Our research will generate set of usability guidelines and some important issues about HCI. Throughout this research we have used lots of different usability principles to generate guidelines. 1.1 Overview of usability: Most researchers have found out lots of usability problems in WAP phones from their experiments. Here, we would like to introduce usability principles. Many usability experts have described usability in different way but mostly all those usability principles are very similar to each other. Jakob Nielsen describes, in [Nielsen 93], usability as combination of Learnability, Errors, Satisfaction and Memorability. Learnability measures novices' ability to reach reasonable level of performance. Errors refer to error-rate, error recovery and number of catastrophic errors in a system. Satisfaction refers to the pleasant usage of the system. Memorability refers to the ability of casual user to remember how to use the system properly. In [Dix 98], the usability is described as a combination of Learnability, Flexibility and Robustness. Learnability is further divided into Predictability, Synthesizability, Familiarity, Generalizability and Consistency. Here, predictability is the ability of the user to determine the future effect, synthesizability is the ability to assess the past activities in current state, familiarity refers to how familiar the system is for the users having their real-world experience, generalizability refers to the generic view of the system that can be used by user from all other systems, consistency refers to the similarity in input-output behaviour in similar situations. Flexibility is further divided into Dialog-initiative, Multi-threading, Task migratibility, Subtitutivity, Customisability. Here, dialog-initiative refers to the preemptiveness of system constraints, multi-threading means allow the user to accomplish symultaneous tasks, task migratibility refers to the ability to pass on a task between user and system, subtitutivity refers to allowing equivalent values of input-output to be arbitrarily substituted for each other and customizability refers to modifiability of the interface by the user. Robustness is further divided into Observability, Recoverability, Responciveness, Task conformance. Here, observability refers to user's ability to evaluate the states of the system from its presentation, recoverability refers to the ability to correct one's action after recognising the error, responciveness refers to how the user perceives the rate of communication from system and task conformance is the degree to which the system provides all services that user expects.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 10 of 67
Depending on system's requirement and behaviour, these usability principles should be applied. 1.2 Overview of mobile devices and their usability: There are lots of different types of mobile devices currently available in market. Some of them are shown in figure 1. But, in our research we have mostly talked about cellular phones.
Figure 1: Mobile Devices Courtesy of [RedWhale]
Mobile devices are basically used for personalised and on the move access to different services and data. But there are lots of challenges in mobile computing compared to desktop systems, as mentioned in [Satyanarayanan 95]. Briefly, mobile systems are resource poor, inherently hazardous, having less energy source and having very slow and unreliable communication. Coming to mobile phones, poor resources mean- a very small display (i.e. text of 5-10 lines with less resolution, depending on phones), limited input capacity (e.g. push-button), slow processor, low memory, small battery and dynamic contexts (e.g. passing through tunnel). Considering the given limitation, there are lots of other HCI challenges to overcome as described in [Kuutti 2002]. First, “develop better ways to provide good user interface”. A ‘painful’ interface will not make users to visit that site again. Present the information in a way that user can perform tasks well and they also feel good about the interface. Second, “develop better ways of interaction”. Having to change the focus to a new interactive system, from static systems, can be very difficult for users. Interface must require ‘minimal attention’ from the user, considering different contexts [Jesper 2002]. And third, “develop important and interesting services that can be compared to desktop systems”. To consider the usability of a WAP headset, one has to consider three conceptual interfaces namely- Service Interface, User Interface and Device Interface [Pekka 2001]. Here, external/device interface is the interface that is provided by WAP phones and its accessories, user interface is the interface that is provided by WAP browser (e.g. Nokia browser, Openwave browser) and service interface is the interface that is provided by the WAP sites.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 11 of 67
1.3 Approach used in the research To limit the scope of the research, we have only discussed usability issues about User Interface and Service Interface. We have discussed Device Usability issues very briefly at the end of the report since it is as important as the other two. There has been lots of work done to improve WAP usability. We have used different WAP usability reports and research papers to generate guidelines. Lots of work has been done in past and still going on, to make the Web more usable. It would not be possible to apply all of the Web-Usability guidelines to WAP, but we believe that the people that use Web sites and WAP sites are Humans, and they all have got relatively same perceptions and similar expectations, as a whole, from a service. So, in this report at some times we have tried to apply Web site's usability guidelines to WAP applications and also generated WAP specific guidelines. Figure 2 shows the design of this research, which is divided, in three parts.
Problem Domain
Guidelines & Issues for WAPUsability
Evaluation
Figure 2: Design of the research In Problem Domain, we have investigated different usability problems in WAP applications. For this we have used different sources of information currently available. In Guideline & Issues for WAP-Usability, we have characterised different WAP problems, found in Problem Domain, into different categories namely- General Usability, Design Considerations, Navigation, Support Services and Performance & Reliability, as described in [Becker 2001]. We have also analysed each problem and try to found out different issues regarding those problems and generated some usability guidelines. We have also found some research-papers on a specific issue, which we have used, for our analysis. In Evaluation, we have used heuristic evaluation method with four participants using different test-WAP-sites, to apply and verify our guidelines. We have also found some new issues came across in our evaluation. Rest of the report is explained in following section.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 12 of 67
Chapter 2 discusses some of the problems found form different usability experiments of WAP applications. Chapter 3 discusses different WAP-usability issues characterised in General Usability, Design Considerations, Navigation, Performance & Reliability and Support Services. Here, we have also described briefly issues of Device Usability. Chapter 4 discusses the design methodology of evaluation carried out for the research. And also some analysis of the results. Chapter 5 discusses the overall discussion and conclusion of these guidelines and their applicability on those test-WAP-sites.
Chapter 2: - Problem Gathering in WAP
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 13 of 67
2.1 Introduction: In this section we will talk about the problems that users had faced in the initial version of WAP. By this we will create a platform that will be helpful for us to gather usability and design issues and guidelines for WAP applications. Here, we have tried to gather information about WAP from currently available sources and some test-reports, rather then doing our own experiment for finding out problems. These reports have described feedback from the users who used the WAP mobile phones for certain amount of time. They have also addressed the usability and design problems in WAP applications at that time. This section mainly uses the material based on a report called “WAP Usability” [Nielsen 2000], produced by NielsenNorman Group, but it also uses some other papers about WAP, which have been referred separately. So, this chapter only talks about the problems that where found in these experiments. 2.2 Executive Summery: Apart from the problems caused by slow data transfer-rate and smaller size, problems found in initial version of WAP can be described as following. WAP usability failed miserably. To accomplish some simple tasks, the users were taking much longer then expected, at the time of experiments. This was not only because failure in the connection set-up but also WAP sites were designed pretty badly that users had to spend lots of time while selection and navigation. WAP phone providers had used built in portals to satisfy their monopoly. But these portals were not designed well and there was also lack of usability resources available. Spending money for good services will be the only concern of any user. But in WAP, users were charged for the services which were not good and sometimes when users were not even getting the proper services. 2.3 Problems in WAP: The slow connection and higher download-time were very frustrating aspect of WAP. Many times connections were failing before user can download data. While requesting any service, most of the time, users were getting errors like e.g. "Internal Error", "Server Error", "No access", "Connection Time-out", etc. Users couldn't understand what to do in those situations. One of the reasons for this unreliable connection was the slow GSM connection. We have divided the problems in following criteria: 1. User Personal Problem 2. General Usability Problems 3. Interface Problems 4. Application Specific Problems
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 14 of 67
2.3.1 User Personal Problem Users were also annoyed because of the payment scheme used by WAP services. Users were charged on the calling time-rate and as a result they had to pay, even if they didn’t connect and had not been able to access any data. Being charged for something that doesn't work seemed unfair and unacceptable to the users. In some WAP sites, outdated content and incomplete services were major problems as companies launched sites without the necessary commitment to maintenance or professional editorial support. As a result of that users lost their confidence in those sites. Also, there were no indications of trustworthiness of the WAP services. Users cannot simply rely on information provided in the service. In WAP, there was a lack of clear differentiation between services. Some services were having similar kind of interfaces; e.g. newspaper sites were having similar kind of headings and options and many of them didn’t use logos so it was very difficult to differentiate them. 2.3.2 General Usability Problems After the period of experiment the impression of users was largely negative towards WAP. Although there were some good WAP sites which provided comparatively better service to the users but they were much less and those sites were not easy to find. As we have mentioned above that connection time was very slow in almost all the applications and the data quality was too variable. But still some users enjoyed using Internet on mobile phones and other advantages having mobility. In the usability experiment of WAP, described in [Nielsen 2000], the improvement in the users' performance was very low compared to their initial knowledge. At the initial stage, it may be acceptable if users were poor in performing their tasks as they were facing new interface, and they had to learn it first. But in case of WAP, the users didn't perform well at the end of the experiment even after having good use of the WAP phones. This clearly shows that WAP provided very poor learnability. Some WAP services were hard to use because of mismatch between their information architecture and users' tasks. Desktop computers, which have more resources and better interaction capabilities, still face the problem of poor “task analysis”; so a better task analysis becomes inevitable in mobile systems, which is having fewer resources and smaller screen then desktop computers. WAP users, having small screen, don’t like to read all the text by scrolling down. And even, the screen has poor typography and it becomes very difficult to scroll through pages of small number of lines of information and try to piece them together in mind. They just want a very precise content. As we have mentioned earlier that all the WAP phones were having a built-in portal and in which the developers had also provided on-line help. But a very few number of users could
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 15 of 67
find out that which simply indicates how bad these portals were. Even in the phone menuals they have not described about the help. While accessing WWW, users get the indication about whether the application is crashed, or how much progress is made, or what is the status of the service (i.e. System Response); but in WAP sites, users were not informed any of these things while they were waiting for service to be downloaded. So, sometimes when users couldn’t connect, they ended up with just a waste of time and money. Even-though the screen size was a big limitation, for some users it was not the screen size, which irritated them, but the contents that the service providers didn't re-purpose to work on that screen size. E.g. in a newspaper site, if a story is Device Usability: For users the experience with the Phone itself was OK. Here, one of the problem faced by users was that in Ericsson Phone, the scrolling button was on top left edge of the phone which was slightly awkward as users had to change the way that they hold the phones. 2.3.3 Interface Problems WAP designers used non-standard and invented language for labels and menu choices, which user were not been able to recognise easily and as a result of that it caused delays in users' navigation. The need for simple language is even stronger in WAP design, because there is no room to explain non-standard terminology with effects or icons. Initially, users were not aware of scrolling the pages, as there were no indications of page layout. And even scrolling is not a good option in small devices for navigation. Sometimes the phone is waiting for the user to input more information, but doesn't make this clear to the user. Often, the reason is, the phone is waiting for the user to enter information into text boxes but it can be only found by scrolling further down the page. And sometimes, there were no indication that there is any more text to scroll down to. Error Messages: The presentation of error-messages was itself a big issue. Users were very annoyed with the quality of error-messages been shown on the screen. There were no indications about these error messages or the meaning of these messages, in the handbook of their phones. And as mentioned above there were no explanations given on the screen for the errors. And as a result users could not understand what the problem was and what to do next. E.g. after entering any site name, if user gets the error like, "Problems connecting the Server", the uses had no idea on what to do. Should they try again or go for the alternative? And in addition to it there were no indications about where the failure was, at the portal or by the information provider. So, even the users wanted to complaint, they had no idea of where should they go. There was lots of instability in services, on some days users were able to access some sites but not on other.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 16 of 67
Some sites provided the reasons for the errors that were coming on the screen, like "Upgraded and will be available soon." or "site temporarily unavailable". By this users could even understand that what's wrong with it and what to do next. But this doesn't solve the problem as users were charged for this. One of the reasons for these problems was the incompatibility between the phone microbrowsers, different networks, and portals. This was a very big usability problem and has to be solved by all parties including site designers and engineers trying to agree to and then adhere to tried and tested design principles. Navigation Issues: Although there are lots of differences in designing mobile systems and desktop systems, the basic usability principals are same to some extent. Navigation in WAP mobile phones should follow the same principle as those are in the Internet. In WAP, user navigation was restricted through only 3 provided mechanisms: Menu options, Search engine, and Bookmark. This is to some extent acceptable as WAP was into its initial stages and some constraints regarding their usage. But some of them are not working properly. In the navigation point of view, NTT DoCoMo's I-mode provides users the flexibility to customise their own menu so that their favourite sites or interests come up automatically. While in WAP, they had planned this facility but it was not working at the time of the experiment. Labelling: Labelling was not proper and unclear in some portals, and as a result users didn't understand the meanings from a given link or given menu. By this, sometimes user couldn't find out the required services easily. WAP applications should follow the rules of the labelling of the Internet. There should be some simple instructions like "click here" or "try this", from which users can get even some ideas about what will happen after they click it. Which was not there in the WAP sites. Telling people what will happen before they click through to a new screen is a basic, long established principle of Internet design. Back Button: At any point user may feel of going on wrong track and he has to come his previous place and it is totally unacceptable if user has to connect to the server again and download the page which he has just seen. In WAP, if a user wanted to go back then he had to re-download the whole page again, which also mean that user is paying for the same thing again. This will be a very big problem if you have to go long way back to the initial page. Searching: Sometimes, it was difficult to find the desired site, as users were getting lots of results and links with no level of relevance, as they had not designed the search-engine properly. Bookmarks: As we saw that searching for a desired page was not easy and labelling in the menu option was also inefficient, this was the only reliable mechanism for navigation in the sense that all user must have visited and used these book-marked pages. But, the phones have provided less number of bookmarking capacities e.g. 15 to 30, which was not enough for all users.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 17 of 67
2.3.4 Application Specific Problems WAP phones provided very less learnability as users had very less improvement in their working speed even after one week of experience. Some sites were badly designed that more screens were used for comparatively less number of informations. For a small-screen design, it is more important to conserve screen space and present as much new information as possible. Given the cumbersome and slow process for moving between WAP screens, it is even more important for WAP design than for other small-screen designs to optimise the communicative value of each screen. These were the common problem in all different type of WAP sites. One big problem that was common in most of the applications, supporting real-time information e.g. sports, news, travel, whether, etc, was lack of regular maintenance or updating. Users were getting out-dated information, so they found it very unreliable. Other problem was the readability of contents provided by those sites. Users didn’t like to read bulk of pages of information. In the newspaper sites, they didn’t provide information with different level of details. The users had to scroll through the bulk of pages if the story is long. And for each new page the user had to wait until download completes. This was not only waste of time but was also hard to continue on some interesting stories. In order to make the headlines precise, some of the newspaper sites used same criteria and as a results newspaper sites were quite similar in look. Users were not been able to differentiate between them. This way, some newspapers lost their identity. As there was limitation on the screen of phones the users were not been able to view pictures, and images. Some users preferred watching news on TV or reading it from the newspaper for getting more interactions and more detail of the story. [Mobile 2000] Some sites providing financial services like banking, stock exchange information, and other related services, were not developed at the full-strength because these applications are very crucial and require more secure and high-speed connection, which was not there in WAP. At that time these sites were providing very simple services. Some entertainment applications, such as movie listing, TV listings, music sites, had the same problem of having out-dated information. The labelling was also not good and users failed to understand its meanings, e.g. in one of the TV listing sites there was a label recorded by a users called “Unmissable TV”, which doesn’t actually mean what was there in it. So, in this way users had to waste some time. And in this type of applications e.g. movies, music, TV listing, where long list of information could be there, there was no “home” link so the users had to navigate all the way back to go to the first page by clicking on the back button.
Chapter 3: Usability Guidelines & Issues
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 18 of 67
This chapter will talk about different usability assessment issues, which are very useful to consider the overall application usability of Wireless Internet [Becker 2001]. After analysing these usability assessment issues, we have generated specific usability guidelines. This chapter in divided in 6 sections. Section 1: General Usability Section 2: Design Consideration Section 3: Navigation Issue Section 4: Performance & Reliability Section 5: Support Services Section 6: Device Usability Section 1: General Usability 3.1.1 Introduction: To be successful for any technology or product, better user orientation in co-operation with a good system design is necessary. This section is divided into two parts- “User-Oriented Issues” and “System-Oriented Issues”. 3.1.2 User-oriented issues: This is one of the most important aspects of usability that most developers should keep in mind. "There is a widening customer experience gap online. Companies who bridge this gap will win" [Hurst-Gellady 2000]. From a research it was found that approximately 70% of retailers lack a clearly articulated Ecommerce strategies in their online business applications [Spiegel 2000]. Of course, there are lots distinctions between the WWW and WAP, but we believe that the basic usability will be similar to some extent. In this section we will basically talk about user-orientation and other related usability issues. 3.1.2.1 Know the user concerns: Users are pretty much concerned about their money spent on the quality of service they get. All they want is getting efficient, reliable and faster service. There is no way WAP developers can satisfy all the requirements from different user group. At the same time, most common requirements can be satisfied very easily. A first time user should get a clear idea of the application’s usefulness in the first screen of the application. Otherwise the chance is that he/she just might leave. This issue is more important in WAP as there are some limitations: -
At present, using WAP is very expensive. Some users cannot let themselves to get thorough experience of it and then get used to it.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 19 of 67
-
In Europe WAP is used more as problem solving service rather then a fun service. People will not use it for surfing or to entertain themselves. Users will want to use it to accomplish a task [Bergman 2000].
-
The user might be distracted due to the fact that he/she could be in a demanding or stressful environment.
-
Users will definitely not like much browsing i.e. inputting and scrolling to get information.
Analysis: As we have maintained above that many users would use WAP phones as a problem-solving unit, and they will definitely not like to do much browsing and/or inputting and scrolling. Considering different WAP services, we can categorise these applications as “regular”, “urgent” and “critical” depending on its usage. Mostly, user will use mobile phones for getting quick information. E.g. consider a user with three different situations. He might be travelling in his car and first, wants to check his mails (as a regular activity); second, he wants to pay his bills (as a critical activity); and third, he wants to check the traffic news on his way to the destination (as an urgent activity). Also in the above example, while travelling in a car user is more concerned to looking at the road and driving the car. So, for using the phone the user concentration will be divided into other activities as well. In order to fulfil all this requirements, developers have to create contents depending on these application as well as its situations. They should also provide effective interaction styles, considering the criticality, urgency and frequency of usage of the application. In fact, in order to classify all different activities the developers have to do detailed “Task Analysis”. It helps the designers to find out the critical tasks for the usage helps them to consider these activities at the early stages and provide more support and efficiency to those tasks. It also finds out frequency of use for tasks. Designers can assign priority to different tasks by their level of criticality, frequency, user attention and time spent. Users are not experts of the technology. Even some of the users don't like the complexity of any system. It is not a good idea to put complex or technological information to the users. The communication with users should focus primarily on the functionality and the benefits of the systems instead of technology standards [Mobile 2000]. One other attribute of mobile user is they are always on the move and their concentration is divided into few activities. In this situation their readability becomes very poor. So, the information should verbally be very easy to read and contents should be very optimised in a way that suits the users.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 20 of 67
As described in the user concerns that users do not like to input and scroll in order to get simple information. Scrolling in WAP phones can be only applied when all the menu options are well-structured and most important options comes first in the list. A more detailed analysis on scrolling and input issues is given in the section called Design Consideration. In all of these situations information that is displayed on the screen should be essential and intuitive. KISS- "Keep it simple and stupid" [Mobile 2000], is a well-known slogan to support the easiness of the interactive applications. Here, we have chosen a model, as shown in figure-3, which are different browsing styles and can help develop efficient browsing in WAP phones. It shows different mechanisms, in ascending order, to provide more efficiency and easiness to user. This can vary as new technologies and its new mechanisms are coming to the market.
Speech Recognition
Scrolling Clicking (Selection) Entering Numbers is less efficient Entering Text
Figure 3 Model for efficient WAP-browsing These are different WAP-browsing styles that developers should apply in descending order, in order to provide efficient browsing. However, this model also depends on different mobile devices and application needs. 3.1.2.2 User Tasks:
Depending on user group and application domain, users perform specific tasks. These tasks are also performed at different frequencies. Developers should try to identify the main tasks that the majority of user will perform. In order to generate design interface for different application, the developers have to do a detailed Task Analysis. Task analysis is the study of the way people performs tasks with existing systems [Dix 98]. There are many different ways of doing Task Analysis, but we will follow the one that is described in [Openwave 99].
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 21 of 67
1. Identify user activities 2. Classify activities 3. Design interface according to activities 3.1.2.2.1 Identify user activities: This is the first step to determine what tasks the user will perform with the application and in what order. Here, developers should not create confusion between the user actives with the application features. One of the best ways to find what task user performs for a specific application is to conduct set of Inquiries. There are different methods for doing inquiries; Contextual Inquiries [Beyer-Hugh 1997], Ethnographic Studies, Interviews, Cognitive Walkthroughs etc. By this developers can make out set of scenarios. Based on these scenarios, developers can conduct detailed task analysis by dividing the identified activities into sub-activities and sub-goals. [Bergman 2000]. 3.1.2.2.2Classify activities: After the first step the developers can get the complete set of activities and tasks for an application. Now all these activities will have different level of importance to the users as well as the application itself. To classify them all the developers have to consider following groups [Openwave 99]. 3.1.2.2.2.1 Required Activities: These are the compulsory activities for all users that all of the application users will have to perform. These activities always compromise the user-friendliness of the systems and as a result it degrades the overall usability. Some activities like logging-in or set-up activities come in to this category. Developers should try and reduce these activities as fewer as possible. 3.1.2.2.2.2 Main-path Activities: These are the activities that are most frequently used. These activities should be as simple and easier as possible, so the users don’t need to learn or spend much time over it. Even the interaction and presentation styles should also be very easy. 3.1.2.2.2.3 Semi-main-path Activities: These are the activities that have used by many users but not the majority of users. So, these activities should be simple and intuitive, but users have to make sure that these activities should not become an impediment for users who do not perform them. 3.1.2.2.2.4 Side-path Activities:
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 22 of 67
These are the activities that are used occasionally by most users when they use the application e.g. 80% of users will use it 20% of time. Replying to e-mail is one such activity. It makes sense to implement access to side-path activities in a menu that is not immediately accessible. 3.1.2.2.2.5 Rare-path Activities: These are the activities that are never used by majority of users. So, developers should try to remove these types of activities. It would be better to have a simple system than one with a lot of obscure options. 3.1.2.3 Design interface according to user activities: Once all the activities are prioritised, they should be presented on the user screen in a way that best suits the user needs and situation. E.g. Starting with the most important activities, the user interface should be made assure that the higher priority activities are always available with optimal user interaction and fewest user key-stokes. And also these activities should be given better ‘presentation methods’. A more detailed analysis about how to present information is described in Design Consideration section. One other very important issue is to structure all these activities/tasks [Openwave 99]. E.g. consider wants to complete a task and in order to do that, he has to go through 2 sub tasks to complete his goal. In this case structuring of tasks becomes very crucial. To explain these in more detail consider following figure 4.
Sub-Task A
Sub-Task 1
Critical Line
Sub-Task 2
GOAL
Sub-Task B
Figure 4: Flexibility in Navigation To complete a goal user has to go through sub-task 1 and 2. Developers can provide flexibility here by adding some subtask as A and B to provide more options to user [SAP
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 23 of 67
2002]. E.g. A user want to buy something from any online shop from his mobile. Now, there are many different ways that user will buy that product. He may want to straightaway buy that product if he has already got all information about that product or he may want to buy it after getting some information about that product. The user should be given both the options at the same time. Another issue here is about the Critical Line. In Project Management studies there comes an issue about ‘Milestones’ and ‘Deliverables’, in which the Critical Path to achieve that Milestone should be as short as possible [Project Management]. That concept can also be taken here to structure the tasks here. As we can see from the figure that goal should be achievable through minimum amount of dependencies to provide more efficient navigation, of course depending application type. Here, developers have to "Balance" between flexibility and efficiency. 3.1.2.3 Other usability support: 3.1.2.3.1 Minimise the user's memory load While using desktop computers, users have more time and fewer things to remember. Whereas in handheld computers, users have small screen and they also have to remember more information as the task that he will be performing is of very short-term. And also they might be one the move for many times, so for hand-held users it would be difficult to face lots of information considering the limited memory of humans [Telenor 2000]. In the conventional desktop systems, there is well known rule that “20% of the functionality is used 80% of the time” [Bergman 2000]. By simplifying this, developer should cut out this 80% in order to reduce complexity and increase usability. So, the user interface should be as precise and as simple as possible. In order to navigate, user should always give some sort of selection criteria rather then input field. Recognition is better then Recall [Nielsen 10]. In situations of choice the users should be asked to make simple decisions rather than complex ones. Divide the problem if possible. It would be helpful for the user if the WAP browser can remember some of the previous data entered by user. This can be done by better cookie management [Telenor 2000]. 3.1.2.3.2 Consistency It is very important to provide an interface that uses same template or principles throughout the application. It is also recommended to provided interface that is consistent with real world.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 24 of 67
Consistency can be provided in many other ways like labelling and positioning [Telenor 2000]. In naming, the function should be named from its functionality and the same functions should have the same name. In positioning, same functions should be on the same position throughout the site. Having consistence positioning can provide good recognition to users and as a result they can work faster. 3.1.2.3.3 Feedback Feedback is a confirmation of any user or system activity. For frequent and minor actions, the feedback should be modest; but for some major and less frequent actions the feedback should be substantial [Shneinderman 98]. E.g. as a low level feedback, when the user goes to a new page, on that new page user should be informed about that new page by a header-name and also, if possible, should also indicate the information about the previously visited page. This will make user feel confident and comfortable with the new page. And this is a well-known guideline for designing traditional web sites; to repeat the user's choice on the destination page for confirmation of their task achievement [Nielsen 2000]. Feedback should both give an indication of whether the input had a valid format as well as and the result of any operation from user. System should also give feedback about its progress and status. Some icons indicating appropriate visual representation can be useful to provide system progress. System can also provide user’s position in the site navigation hierarchy. This is described in Design Considerations in more details. 3.1.2.3.4 Internationalisation: Internationalisation refers to the process of designing a core application considering the needs of foreign users [Brew 2002]. Designing an international user interface may or may not include language conversion but should include special needs of other countries and cultures [Nielsen 94]. This section will provide some important consideration for using internationalisation as described in [Brew 2002, Nielsen 94]. Many countries have character sets that extend beyond A-Z of English and ASCII character set. Some Asian languages have characters that extend 8-bit ASCII code. Some applications have to use more then one language, which makes the representation more complicated. Some languages use special characters ‘å’, ‘ø’, ‘ü’, etc. which have to be sorted by simple characters, e.g. ‘ø’ with ‘o’. The problem can also arise when presenting different currency-standards and time measurement standards. In different currency standards the main problem is decimal point. E.g. In US: In India:
$ 10,000.00 10,00,000 Rs.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 25 of 67
Developers also have to review the application if it has to be made for international market and also have to find out how different users will interact with it. After this developers can change or accommodate some assumptions so that user can adapt to a variety of countries and languages by minimal breaking of the application. 3.1.3 System-oriented issues: 3.1.3.1 Iterative design and usability testing: Iterative design and usability testing are the issues that are not only applicable to the development of WAP sites but all types of computer-based systems. These are very important processes in systems development life cycle and also need some management. For this research we cannot go into details of these issues, we can only give a very brief introduction of their importance. The developers cannot predict all the requirements of the system at the beginning of the development, moreover in the development process these requirements tend to grow or change. So, the better way to develop a product is to make a potential prototype of the system and apply them to some users and get feedback on the prototype to make an improved design. And by this evaluation, the final product can be improved. The definition of the this iterative design is, as described in [Dix 98]: “Iterative Design is a purposeful design process which tries to overcome the inherent problems of incomplete requirement specification by cycling through several designs, incrementally improving upon the final product with each pass.”
Iterative Design in conjunction with Usability Testing can improve the system productivity. Usability Testing is very important as it provides direct information about how the people use computers and what their exact problems are with concrete interfaces being tested. [Nielsen 94]. 3.1.3.2 Compatibility Issue: Different mobile phones supported different browsers, e.g. Nokia and Ericsson used their own browser; Motorola, Seimens support Phone.com browser (now Openwave). Now, all the browsers support different WML structure, so it can be possible that an application running well on a browser may not run so well on other. For example, Nokia assumes a "title" tag for a WML 1.x card should be displayed at the top of the screen (but only one line); Openwave assumes there will be content on the card describing the card and thus relegates the title for use as the default bookmark title. [Little 2002]. All these mobile devices run on different operating systems like, Palm OS, Windows CE, Symbian, etc This issue has affected the mobile usability to the worst. We have already described different problems caused by compatibility. Here we will describe some approaches to solve those problems, as shown in.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 26 of 67
Develop applications targeting a single device. Developers can readily write an application that delivers an outstanding user experience for any small number of devices. It has, however, drawback as it may only work for specific device group. Develop application that can be used by all devices. By assessing all devices' implementation of the platform, and selecting the subset that works effectively the same all devices, developers can write only one application. Of course, this may be impossible for some applications. Use a “run anywhere” server. By this, server can translate the original application to the format supported by the device. This can be very efficient approach and also is costly. 3.1.3.3 System Architecture Issue: As we have mentioned in our introduction that usability in mobile device depends on three interfaces namely: user interface, service interface and external interface [Pekka 2001]. Particularly in service interface, the usability can be enhanced by better network and communication mechanisms. In most simple scenario, developers manually build WAP sites by using WML. But there are lots of other mechanisms currently available that can translate the HTML contents into WML or other specific contents, at some third place. These mechanisms are called proxy. One of the proxy mechanisms is shown here in figure 5.
Figure 5- an example proxy mechanism Courtesy of [Arizona 2001]
There are lots of reasons that a proxy mechanism can improve the usability of mobile devices. First, as we have mentioned in introduction about the problems with slow connection in mobile technologies, having a proxy decreases the delays caused by wireless communications. Second, it acts as Point of Presence (PoP), which causes the faster response
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 27 of 67
time (in terms of feedback and feedthrough), if it is near to the clients [Dix 99]. For more information please see the section Performance and Reliability. Third, it can support heterogeneous devices and can provide better quality of service by employing different transcoding mechanisms. Transcoding mechanism is a very big issue within itself. Here, we can only provide brief information about different transcoding mechanisms. WebExpress is an example of this kind of proxy systems, which can employ different mechanisms supporting compression, image conversions, caching, differencing, header reduction, protocol reduction, etc. The efficient this mechanisms better usability in service can be provided [WebExpress].
Section 2: Design Consideration: In previous section, we have introduced user’s concerns. This section will describe usability issues related to design layouts, information presentation, and interaction mechanisms. 3.2.1 Input Layout Here, one important thing should be noted that in order to provide overall usable inputtingprocess primary issue is to provide efficient input mechanisms (e.g. graffiti, T9, etc.), and the input lay-out of the application becomes secondary. In this section we will describe different input layout issues and guidelines for WAP enabled phones. As to follow simplicity in the application developers should provide fewer fields for input so that it can save the time of the users. "In mobile phone interface design, every click loses half of your users". -Nick Knowles (Chief Technical Officer, KiZoom)
A research described in [SAP 2001], proved that error-rate in textual input is twice as higher then in numerical input and the speed in textual input is also twice as slower. So, the developers should prefer numerical input over textual. And if it becomes inevitable to use textual input fields keep the input mechanism in mind by utilising the first letter of buttons. And if possible developers should use selection of links more then inputting something, in case of navigation. In applications where more information from the users are needed, try to use fields that can fit in only one page, as scrolling is supported by all browser. Also try to use input field like check box, radio-buttons, so that users don't need to write much [SAP 2001]. It is also important for the developers to label those input fields in a way that prompts the users to recognise whether it as a textual input field or numerical input field [Telenor 2001]. It is also important to have some default value in any field [SAP 2001].
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 28 of 67
In applications where more information from the users are needed, try to use field that can fit in only one page, as scrolling is supported by all browser. Also try to use input field like check box, radio-buttons, so that users don't need to write much [SAP 2001]. Developers can use wizards or forms for entering information. In [Openwave 10], they have suggested the developers to use wizards instead of forms by saying that wizards give better user experience especially for novice users. And up to some extent it is a good idea to use wizards as there is less input needed on a page. But for expert users or returning users it is not a good idea as they feel it as tedious. Other problem is that the back functionality doesn't work well in wizards, as users have to navigate through all the wizard-pages that he/she has visited to go back to the original place. If developers are using wizards, they should use the page information on the screen as shown in figure 6. And avoid having a separate "submit" card. The last OK of the wizard should in fact be a submit option [Openwave 10]. XYZ Bank
XYZ Bank
Payment Step 1 of 2
Payment Step 2 of 2
Credit card number: [ ]
Expiration date: (mm/yy): [ ]
Menu
Menu
Next
Pay
Figure 6: An example wizard To use any mobile phones, users have to enter pin first. This way the device itself provides security by identifying them. Developers should only ask for username and password when application requires more security e.g. applications like online shopping or online banking. For other applications cookies can work fine to identify users. 3.2.2. Output Layout: At first instance it can be said that having small user screen, the WAP usability is heavily affected by amount of information that can be presented on the screen compared to a Web page. However, it is very important to know how users uses and perceives information from both web and WAP pages. While using web pages users don't read each and every word but they just scan it [Neilson 99]. This is because they get lots of information at the same time and some of them they don't want. However, in WAP applications users are on the move very often, so the way they use and read the information is very different. Researches showed that, in small display device (i.e. mobile phones), user's ability to read and understand information is not adversely affected [Jones 99]. It is also said that user is more active when he uses the small screen device then any desktop computer. So from all this, it is reasonable to say that limited screen is not a disastrous for the usability and the
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 29 of 67
contents presentation is more important issue. Taking this concept in mind we will generate some guidelines and issues to improve the output (or presentation) of WAP devices. This issue is further divided into “Getting relevant information” and “Presenting information”. 3.2.2.1 Getting relevant information: Building a WAP site often means transformation from its Web site. For this the developers have to reduce its contents and have put only the relevant and most important contents in the first place. Finding the most important information requires a firm grasp of the subject area as well as an understanding of the users' needs and tasks - the development team has to involve users throughout the design process [SAP-MiniApp]. There are some generic rules described by Dr. Nielsen [Nielsen 99] to get the relevant information. Aggregation: Showing single information that is a collection of other smaller ones. Summarisation: Presenting a large amount of information in a smaller form without compromising the quality. Filtering: Eliminating some irrelevant information that users don’t care. Elision: Showing everything with examples to provide practical understanding and to reduce the words. 3.2.2.2 Information Presentation: From researches it has proved that only 20% of display is used for 80% of time.[Bergman 2000]. And as we know in mobile device the user screens are very small, especially in mobile phones, it becomes necessary to use the user display optimally. We have divided this issue in different presentation and interaction issues. 3.2.2.2.1 Links: Different browsers use different ways of presentation. E.g. Nokia browser uses only underlined text to present links, as “Link”. In Phone.com browsers (now Openwave), they present links as Link, [Link] and . By presenting the links as underlined text gives the users the Familiarity of web sites enables users to discover the links very fast. Developers should not use underlined text for any other purpose and do not put text within <> or []. This might be mistaken as for a link. The developers have to take into account the different WAP browsers supporting different length of text in a line and wrapping as well. E.g. some phone.com browsers does not support more than 16 characters, and Nokia browser does not support wrapping and allows only 18 characters in a line [Telenor 2000]. Another issue is about the words or the sentence contained by those links. Telling people what will happen before they click through to a new screen is a basic, long-established principle of Internet design, which can also work well for mobile applications [Nielsen 2000].
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 30 of 67
Use the smallest meaning words or abbreviations for links, because for some times if there a longer word or phrase is used as a link then users may misunderstand that assuming it as two different words. And even while selecting those links in some browser only one part of the link will be selected, which may mislead the user. Clicking a link can lead to some information contents or can lead to some more sub-links. It will be a good idea to let users know what they can expect by clicking those links. There are some important issues here that designers should consider [Telenor 2000]. Labelling - Link that leads to lists of topics should be distinguished from links that lead directly to services or content. Naming them properly can do this. E.g. "NEWS" will lead to different category of news like, sports, technology, latest news, etc.; and the link shown as some breaking news can lead to contain straight away. Information at different level - Link that lead to lists of topics and links that lead directly to services or content can be shown at different level of link-structure in the portal. By separating links - A page containing links to both lists of topics and links directly to content or services can be shown separately by putting them in different groups see figure-7. XYZ News - America asks Iraq - Tony Blair supports - Sports - Local News Menu
Back
Figure: 7 Example of Links 3.2.2.2.2 Page Layout: It is very annoying for the users to read any document having lack of continuation. The designers should use the whole card for information presentation, otherwise users have to wait for new card to be loaded. If the information is longer e.g. large articles or email, there should be a link "MORE" shown at the very last of page. And this "MORE" link should be easily discoverable and recognisable. For this designers have to show it with different size of font with different type of color. 3.2.2.2.3 Writing & Naming: In Naming and Writing, consistency and simplicity are the two most important characteristics for any well designed system.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 31 of 67
Use consistent grammars in menu and link items that are used in WWW e.g. don’t use Back instead of Previous and Home instead of Main. Also within a site, use the same words throughout, for the same functions [Shneiderman 98]. User uppercase and lowercase letters appropriately. As we know Upper-case words are slower to read so they should be used for only important information, which needs more user concentration [SAP 2002]. E.g. for a Flight-No: user Uppercase SQ203A and for general text use lower-case words.
Space restrictions may necessitate the use of abbreviations in many situations: in dialog boxes, for field names in the applications, for lines in the menu bar or pull down menus, and for function key names. But do not use abbreviations if sufficient space is available [SAP 2002]. Use abbreviations only if you are sure that the user can easily understand them. Refer to standard abbreviations such as listed in dictionaries. Designers can also use new abbreviations, if there is no accepted abbreviation available and more information is necessary. There are some rules defined for this in SAP design guide [SAP 2002], which we have shown here, briefly. - Right truncate words: Create abbreviation by omitting the end of words and use period. E.g. "abbreviation" becomes "abb." - Omit vowels: Vowels that do not influence the pronunciation of the word very much can be left out. E.g. "group" becomes "grp". - Create acronym: created new terms consisting of the first capitalised letters of the word fragments or individual words. E.g. for "Human Resources" use "HR". 3.2.2.2.4 Graphics: Normally, the most limiting factor in WAP phones is small screen. So, developers should use graphics very carefully. They should only be used for Company-specific logos [SAP 2002] (to provide the users the familiarity and recognition about that site), Navigational status of the phone and as a part of content only if that is inevitable. As a consequence of the limited screen, a good strategy to use graphics that are stretched horizontally. 3.2.2.2.5 Fonts: Designer would want to provide as much information as possible, efficiently. And it would also be logical to provide text with smaller fonts on small screen. Researches have shown that the font size and its type have effects on the overall usability. [Legge 85] found that reading speed improves with the increasing text size, but upto some critical point. [Tullis 95] found that as the font becomes smaller the error rate increases. Still there is not much work done on fonts of the small screen. Here, there are some guidelines most of which basically derived from the desktop PCs but can also be applied to small screens.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 32 of 67
Do not use capital letters for no reasons. It reduces the readability speed for the users. Only use them where they are originally used or when it is very important to draw the user's attention. Bold fonts should be only used for important information. Use big fonts where there is need to show the status of phone and activity the phone is doing. That is because it is very important to provide the feedback for the user's activity. Some status messages like "Dialling", "Call Divert" are very important to be shown to users and as we mentioned above that bigger fonts improves readability, it becomes essential to show those messages in bigger fonts. Even some error messages should also be shown in bigger fonts. Some researches like [Tullis 95] has shown that even some font types are also very important for usability. They have experimented with Arial, MS Serif and MS San Serif; and proved that as the font size increases the reading time and the error rate decreases in Arial. Some times fonts become very crucial when novice users are using the device. We have already devoted a section to provide usability to novice users in Customer Support part. Fonts can provide efficient discoverability. Discoverability means easiness to find things or options from the screen [Openwave 99]. It would be easier for the novice users, if text materials are presented with different font size and font style, and now a day considering the available color screens different font colors will also be very helpful to the novice users. E.g. in the following figure it is not only easy to discover different links but also it is good for the user to recognise where they are in the navigation. XYZ -NEWS -Sports -------Cricket - Tendulkar on 100 - Dravid is the …. MORE Menu
Back
Figure: 8 Example of Discoverability 3.2.2.2.6 Main Card By main card we mean the first page of any WAP portal or WAP site. In WWW the first page is called the home page, and it the starting point to find any service or application within the site or portal. Now, it is very important for the developers in WWW to make the Home Page as usable as possible so that any kind of user (specially novice users) get full and efficient advantage of the services provided by the whole site. This theory can be applied to WAP sites as well. Ideally, the main card should always have an option that matches as close as possible to any user’s initial thought about what is the task and what is the best way to start with [Movie 2001]. For this a proper Task Analysis is needed.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 33 of 67
To improve the usability of the main card, all other issues of menus, labelling, list, fonts, etc, all that we have described in this report are important, which is needless to mention here again. Increasing the length of main card incurs relatively small loss in usability while at the same time can provide for a significant gain through the variety of options offered [Movie 2001]. A greater number of options on cards indirectly increase the usability of application either by allowing shorter time required to reach the target card or by increased level of satisfaction for users who want to find out more details along their search path. Long card also mean more scrolling, so the interface should be well structured. For this, options or links that are presented there, should be most important and frequently used by most users. Even in terms of marketing, the first card is very important e.g. in any commercial WAP site, the first card is very important to attract user by providing efficient usability. 3.2.2.2.7 Menu Design: WAP phones and other hand-held devices, usually, utilise hierarchical menu structure to interact with users. Menus were originally designed to exploit the fact that humans are better at recognising commands from list rather then recalling them from memory [Marsden 2001]. But the menus cannot be straight away translated from desktop PCs to hand-held devices, as there are lots of limitations in handheld devices. And also because users don't have access to all the available choices simultaneously on one screens, in the menu. To find some improvements considering the limitation described above, we have described some guidelines and issues here. Developers should try to build a menu system with a uniform structure and make similar parts of the application look and feel like consistency [Dix 98, Shneiderman 98]. To prevent applications to behave differently in similar situations, the developers have to use welldefined hierarchical structure [Ericsson 2001]. There are two approaches defined in [Ericsson 2001]. First structure, as shown in figure (a), uses strict relation within menu, so that uses can easily keep track of the location in the menu. But this approach some times becomes tedious, as users have to navigate through the whole sequence back to go to the different main menu options. Second structure, shown in figure (b), hyperlinks the menus together so that they can be accessible in different ways. This approach gives shorter interaction time while moving between different menus. But same time could be wasted, as users can't keep track of the navigation, especially novice user. The developers have to use one of the approaches depending on their application.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Figure (a)
Page 34 of 67
Figure (b)
Menus are part of navigation. The efficient the menus are presented the efficient and faster the navigation can be. One of the best ways to improve the usability of menus is to have better categorisations in the menu classification [Marsden 2001]. Considering the small screen display the menu categorisation is also limited to few lines. Research [Swierenga 90] says that handsets having 3-4 lines of options can perform similar to desktop computers. There is also an important issue about the tree structure in menus. [Miller 81] conducted that wide and shallow trees are more desirable then narrow and deep. Recent research [Norman 88] found out that "Concave menus are more efficient"- i.e. developers should give more options (i.e. wide structure) at the root and leaves and less options at the intermediate links. For more information about the tree structure see the Design Architecture section of chapter Navigation. In menu design if the developers can provide some extra information along with the options then that would be very helpful for the users. E.g. In movie listing applications, the release date along with the movie name or even the star-cast along the movie name can be very useful. There is also an issue about designing menus with small number of options. Of course this contradict with the long main card that we described in previous section. But in short, it totally depends on the developers to choose any of these options depending on the application that they are developing. More justification about "number of menu options" is described in following section of scrolling. Here, there are some more guidelines. If you want to use more menu options in a card, keep the frequently used options appear first. Much more importantly, let the users customise their own menu so that their favourite sites or interests come up automatically [DoCoMo's I-mode]. Another method is to presents the list with numbers that act as accelerator shortcut keys, allowing users to select an option by pressing the corresponding number on the keypad rather than scrolling to it.
3.2.2.2.8 Scrolling: To start with simple, scrolling is needed where there is more information to be presented beyond the screen limitation. Generally, scrolling creates some problems as users sometimes just scan the text and may forget to look further down. Scrolling also affects overall readability of the screen, as user has to divert his concentration towards the scrolling. Specially, where there is a huge limitation of screen size, the scrolling becomes very crucial.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 35 of 67
There is a slight difficulty to provide the straightforward guidelines for scrolling in smaller devices like mobile phones. The usability guru Nielsen, in his review of Web Usability since 1994, no longer sees scrolling as a "usability disaster". But he also suggests developers should not to use scrolling in small screen devices like WebTv [WebTv 97]. In contrast, some researchers from Middlesex University has provided different mechanisms to improve scrolling in Mobile phone applications, acknowledging that scrolling in small displays can be improved. Considering Nielsen's arguments about scrolling, it would definitely be a good idea to have only screen-sized pages (cards). By using small sized pages/cards not only scrolling can be reduced, the error that are being made from using scrolling can also be reduced. E.g. sometimes error can be made if a user may not scroll down to choose a particular option, as he just scans the page. Also if users are not prompted to scroll they can miss some information down-words by not scrolling [Nielsen 2000]. Also scrolling can reduce the readability and diver the user's concentration. So, all this issues argue to support Nielsen. Researchers from Middlesex University have done some experiments on different way to employ the scrolling and enhance the overall usability [Buchanan 2001]. Among them the best was Vertical scroll method, which displayed each menu option on a series of lines on the display with the user scrolling down to see the full list of option as shown in figure 9.
Figure: 9 Scrolling The main advantages of this method were it was easier to read from the screen with no wrapping, the average time complete a task was very low and also the average number of error were also low. The drawback of this method, considering Nielsen's comments [WebTv 97], was that user has to do some scrolling. But by providing increased length of pages can gain significant advantage of variety of options offered on the screen [Movie 2001]. A greater number of options on a page indirectly increase the usability of application either by allowing shorter time required to reach the target service or page; or by increased level of satisfaction for users who want to find out more details along their search path.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 36 of 67
After considering both arguments about scrolling, if a developer can provide a mechanism in which most important and primary options of links should be presented in the front and other less frequently used options/links should be presented last on the page [Jones 99]. By this way most of the users can choose their links with minimal scrolling from the screen. For this developers have to do Task Analysis and have to rate the activities. Other issue for scrolling is about the ergonomics of the scroll-button. In some mobile phones, e.g. Nokia, newer types of scroll-buttons like Navi-roller and others are used, which are actually very effective to use compared to other mobile phones like Ericsson, in which user have to use the push-buttons. In navi-roller users just have to scroll the roller on the screen, where in scroll-buttons user has to push up- and down buttons to scroll up and down, which is sometimes tedious.
Section 3: Navigation Navigation is the navigational schema in terms of breadth and depth of search path and traversal mechanism [Becker 2001]. An efficient navigation depends on architectural issues or the tree structure and effective presentation on the user screen. In this section we will talk about both of the issues. 3.3.1 Architectural Issues: In WAP application all pages are called Cards, which can be of any size depending on application. When a request for a WAP site is made these cards are sent in a bunch, which is called Deck. Now while users navigate through the site, the Deck should have sufficient Cards for the use. If the requested page in not in the Deck a new request has to be send to the server, which will take some time. So, it is recommended that a maximum navigation space should be provided within one Deck to avoid reloading during the navigation [SAP 2001]. In [Openwave 10] it is recommended to use decks of 500 bytes or less. They also say that having larger deck space can increase download latency. They also suggest setting the deck's expiration value high if the content is static. The user can navigate through pages by using menu options, links from the menu or may be from the back button. Now, all the cards/pages in a WAP application can be represented as a tree structure. Issues about different tree structures are presented in Menu Design section. Precisely, wide and shallow trees are more efficient then narrow and deep trees [Miller, Lee]. One of the reasons for this is wide and shallow tree mean more options or links are presented on each of the screen. And as more options are presented on the screen, user can complete the tasks faster, with less interaction time. And providing more options user has less memory load, so this serve as recognition then recalling. The navigation can also be utilised for efficiency and flexibility. From the figure-4 described in General Usability section, flexibility can be provided by allowing users to navigate through different paths, at the same time efficiency can be provided by providing user fast-
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 37 of 67
path [SAP 2002]. Developers should also use focused navigation by e.g. presenting list of goals that they can achieve from the site straight away [Jones 99]. 3.3.2 Presentation Issues: Menus are the representation of the "card architecture" in all WAP applications, as whenever the user selects any option from menu he will be shown a new card on the screen depending on the selection. Now, user always wants confirmation about their task completion [Nielsen 2000]. Whenever they navigate by clicking a link on a page next page should describe some information about the previously clicked link, means user should be given confirmation that they have gone to the right page that they wanted to go. One of the solutions, as described in [Marsden 2001, is to provide “visualisation” by giving the visual feedback on the screen. Actually, visualisation is one of the wildly used mechanisms for PDAs. In [Marsden 2001] they have used Iconic presentation and Contextual presentation. In iconic presentation, they have suggested to use animated icons to represent actions or verb as shown in figure. Here in addition to word-Call Divert, animation can add little understanding about the role of that menu link. Some context icons can be used to highlight a particular choice from the set of alternatives. It is also better to display full set of context icons on the screen then single menu per screen. In contextual representation, it will be better to give feedback to users about their position in the navigation, especially for novice users. E.g. as shown in the figure 9, show the menu number and its section number in the top-right corner. 1-7-2
Shows the navigation hierarchy
Memory Status
Menu
Back
Figure: 10 Contextual Information
3.3.3 Back Button: It may happen at any time that user may find himself on wrong path and may want to go back. So, developers should always provide back functionality. Different phones browsers use different provision for back buttons. Some phones have separate back button and some have back function in menu.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 38 of 67
One of the problem in current back function in mobile phones is sometimes users cannot go back in a site where timer is used. This type of pages often disables back function. In this case the phone browser should ignore the timer function and allow the user to go back. Some browser (e.g. Nokia browser), uses the same button for back function and clear function [Nokia]. In this case when the users are dealing with wizards and they want to go back then they have to clear all the fields first and navigate back to the page they want. One of the best ways to provide intuitive back function is to set a "Back" link on the WAP site. By having a back link, user will not be confused about the back buttons, as different phones uses different ways for back function. Other important thing is to have a back functionality, which takes the user back to first page or "Home". This can be done either setting a Home function in menu or by putting a link on the site itself.
Section 4: Performance & Reliability Up to now we have addressed different issues about presentation of layouts, different interaction layouts, different user services, etc. However, the overall usability has lots of influence over performance and reliability provided by the system. In general, there are no specific guidelines to provide better system performance or reliability because different applications have different levels of performance requirement. This section will apply the general issues to improve the performance of WAP applications. In web, performance can be measured in terms of consumer wait time and system response time [Backer 2001] and reliability can be defined in terms of site crashes, downtime, error messages, and consistent response time. We can apply the same concept for WAP. In this section we have tried to discus them in following manner. 1. System response 2. Error prevention & recovery 3. Domain-Specific Application 3.4.1 System Response: One of the most important issues about System Response is that it affects human behaviour and productivity and as a result it affects the overall usability. E.g. It is obvious that a system, which takes long time to respond, will annoy its users. A system with higher response time will definitely please the users but it should not be much faster that its users don't even understand what have happened. There have some research been made for a proper response time while interacting with WWW, but none with WAP or other mobile applications. In any computer system, response time is the number of seconds the system takes to perform and satisfy that activity after the user initiates it. But in reality, to calculate the system response, issues like -interaction complexity, technical feasibility, costs, task complexity, user expectation, error rates, error handling procedure, have to be considered. The overall
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 39 of 67
productivity not only depends on speed of the response but also on rate of human errors and how the system helps to recover them [Shneiderman 98]. So for our research here, we will try to find out what factors and issues to consider getting optimal System Response from WAP applications. Looking at the previous usability studies of WAP [Nielson 2000], system response was the most frustrating bit for the users. One of the reasons for this was user's expectation about the WAP phone's response from their previous experiences. Generally, user uses a mobile phone to make and receive calls, so while using WAP enabled phones they can wait up to the same response time that they can wait while making call from a simple mobile phone. Second thing, probably most of the WAP phone users expect its response time similar to WWW as WWW is the main and widely used source of Internet applications. So, a proper response time that user can tolerate from any WAP service can be found from the two aspects that we have just mentioned. Of course, we cannot provide any concert experimental evidence here but this can be very useful for future research about WAP usability. Other issue to improve the system response is about Point of Presence (PoP) as described in [Dix 99]. WAP applications can have single-user interaction or multi-user interaction depending on the applications. And as we know the slowness of mobile communication causes delays in feedback (response from user's own action) and feedthrough (response from other user's action), it becomes really important to consider the system architecture. Architectural design has important effects on usability and notably the temporal properties. In both cases, the PoP is the replication of server that has almost all the capability of server for a particular application. PoP effectively form a third place where computation and data may reside. The crucial point about PoP is that it can be very near to user (or client) and as it can improve the speed of the response. E.g. a PoP can be a local cell's base station for mobile phones or a it may be a desktop computer for PDA like devices. Other issue to improve the system response is through better caching and cookie management. Cache management is important for quick access for previously used cards and to control the display of time-sensitive data and cookies are used to store data so that user can be flexible to enter less input [GSM 2001]. Here are some common guidelines for both of them. - Do not store any time-sensitive data in cache. - Set proper expiry value for the deck, so that old decks can't be used after certain time. Response time is also affected the underlying protocols and technologies used. In Europe, all mobile phones uses GSM networks that are relatively slow then GPRS. DoCoMo's I-mode technology (very similar to WAP) uses GPRS networks and has already gained lots of advantages against WAP. This report will not look into these technologies but we can definitely generate an issue that to improve system response better protocols and technology are needed aseptically for WAP.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 40 of 67
3.4.2 Error Prevention & Recovery: Error prevention and its recovery are part of system response but as their importance in usability we have separated them from system response. Even the experienced users are some time makes error in operation computers [Card 1980, Brown-Gloud 1987]. One of the better ways to reduce errors is to provide improved error messages. Improved error messages can lead to lower error-rate, increased subjective satisfaction and repairing the errors [Shneiderman 1982]. In particularly using WAP phones, when the user time is more critical and users have less attention towards the system (compared to desktop systems), the errormessages becomes more important. Mobile devices use menus, options, lists, etc for the interaction and as a results the interface becomes very complicated. So, developers should not use "modes" in interaction, as it becomes impossible for users to reverse their actions [Norman 1983]. It is better to avoid the errors before they could occur. First we will see the error prevention. One of the most frequent errors come from mismatch between the "user input" and the "system input structure" [Shneiderman 98]. One of the ways to deal with this problem is to use proper instructions on the screen for the users. Sometimes it would be helpful to use selection as input. By this way, system can not only prevent the error but it can also make faster progression, as there will be less input needed for the system. Sometimes user tries to enter command that is not supported by application [Shneiderman 98]. In this case developers should provide some default inputs [Dix 98]. Other way of preventing errors is to foresight the applications and investigates if there are any resources available at the site. It is far better to say, "service is unavailable" then allow the user to sign-in and then show them the error [Nielsen 2000]. System can also guide the users through prompting them to navigate on specific set of path to achieve their goal. Some help can be provided by the different input mechanisms used by the device. E.g. some phones use T9 facility for input, which remembers the previous user commands and provides list of pre-entered commands that may be used again by the user. But this comes in device usability, which we are not going to look in this report. Users can sometimes make mistakes and may want to recover from that mistake. Recoverability is the term used in [Dix 98], which means to clear or overcome the error of interaction that has been recognised. The recoverability can be achieved by moving forward or moving backward depending on application types. Error messages are also one of the ways to recover from the errorsome interaction. Error message works for different causes. It can serve as a feedback, to prompt the user about the mistake and can also help recognise the mistake. Considering the complex interfaces of mobile phones, it is really difficult to differentiate which fuctionalities are device specific and which are application specific fuctionalities [Bergman 2000]. So, if an error occurs the error message should explain if that is application specific or device specific. This helps the users if they want to complaint about that error [Nielsen 2000]. There are some guidelines about how to use error-messages [SAP 2001].
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 41 of 67
-
Error message should contain the actual problem of the error and the reason for it, in a language that a common user can understand.
-
And if possible also provide some hint for solving that problem.
Here, language becomes very critical usability, as we know that user screen is very small and the error message should also be informative and easily understandable. And any technical information and cryptic errors should be avoided. Like error messages, the success messages are also important. If the system has performed a task successfully then the user should receive the confirmation for that [SAP 2001]. Especially in WAP phones, where there is limited screen space and most of the time users are on the move and as a result they cannot provide 100% attention to the system, it becomes necessary to provide confirmation for most of their tasks. By providing the success messages systems can provide more satisfaction. 3.4.3 Domain-Specific Application: Throughout in this report, we have presented guidelines and issues that concern most of the application domains. In this section we will talk about some of the specific requirements for domain-specific WAP applications. Domain-specific applications like News, Sports, Shopping, Entertainment, Travel Information, Financial Information etc have their own specific requirements like real-time support, accredited information, security, etc. This section will provide some issues and guidelines for building domain-specific WAP sites. 3.4.3.1 Payment Issue: As with all things involving money; security and reliability are critical. There are some WAP sites that offers on-line shopping or other services that requires users to pay. In this section we will provide good user-interface issues for payment service to get better usability from [Allnet 1]. User should always be given information about the exact price (i.e. including VAT) of the product on the same first page when he/she selects it, with proper currency and decimal point. To provide secure navigation, the browser should disable the back function so that user can’t go back to the original page and he cannot place another transaction from that page. Now, the confirmation of the transaction is also very important. But it should also depend on the level of the price of product. If a product is worth e.g. £1, then it is not necessary to provide confirmation of that transaction. And as the amount increases the level of security and feedback should be more effective. Developers should only ask for password or pin, if the transaction is very important e.g. transaction of a big amount or if the service requires more confidentiality. One of the reasons for this is because mobile phone, actually, asks for initial pin-number when the user starts using it, so it would not be worth taking users time to ask password if the transaction is not big. But for all the transactions, it should just give the
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 42 of 67
confirmation e.g. “Transaction Successful”. This gives the user good control and a feeling of safety, which is an important competitive quality for an e-commerce WAP service. If it becomes necessary to ask password from users, the developers should follow these tasks. As we know that inputting from mobile headset is a tedious task. The browser should allow numeric passwords, so it can be entered easily. Another issue about password is when user types a password most browsers mask it and user cannot see if he/she has typed the write word. This situation becomes worst when the user has to enter some letter, in which he/she has to be very careful. E.g. to type ‘v’, user has to press 8 three times. So, there is a big probability of getting password error, which can make user highly frustrated. Another reason, while entering the password in “*****” format, there is a chance that user might forget the typed letters and have re-type all over again. So, developers should not use masking mechanism for password. One advantage of not masking the password is, mobile phones are very personalised and whatever the user is doing cannot be easily seen by others as most of the users use it at chest height or stomach height, by keeping the head down. So, mostly the privacy of password cannot be breached. 3.4.3.2 Timeliness Issue: Timeliness is necessary for almost all applications, mainly supporting News, Sports, Whether, Stocks, etc. applications. And some times it is also important for Entertainment applications e.g. in Movie listing, TV listing, etc. Users always want to see the newest information and hates out-dated information. One of the simplest but the most important and inevitable thing, to apply timeliness, would be to maintain or update sites more regularly. Maintaining information along with efficient presentation can increase usability performance. This section will talk about some guidelines to present real-time information. For some application specific WAP site, it would be better that the first page should be about the latest information about that site. E.g. for News site, the latest and most important news should be on the main card and other news can be selected from different categories. For a whether site the latest whether information for any specific place should be on the first page rather then the whole menu. Allow users to bookmark some categories so that they can straightaway go there rather then from the first page. On the main page “alerts” should be given, when new information arrives. These alerts should be easily discoverable with e.g. “New” logo or blinking texts. 3.4.3.3 Accredited Information There is a big problem of accredited information in the Internet. However, it has been reduced up to some extent by digital certificates and signatures, but mostly in some finance and stock related applications. There has not been much development done in these applications of WAP or other mobile applications. But, even in case of any other application domains like e.g. News, Sports, etc, people need some kind of confidence in the given information. But the application developers can put reference form where the information has
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 43 of 67
come. This can at least give some confidence to users. E.g. in the following figure 11 of Yahoo-News WAP page, all the news links mention about its source “Reuters”. Yahoo! News - Terror alert on 'Alligator Alley' (Reuters) - Powell to push Iraq - 14 killed when van goes off Maine bridge (Reuters) MORE
Menu
Back
Figure: 11 Accredited Information Section 5: Support Services By support services, we mean those services, which provides additional support to user-tasks. The application-providers may provide these services in addition to and support their core service. This section will gather guidelines for support services like "Online Help Service", "Bookmark", and “Searching". It will also provide guidelines for specific user group and issues of flexibility and reliability. 3.5.1 Online Help Service: There may be a situation while using the WAP service, where user will need help from the system. Help can be the application help or the device help. Here, we will only talk about the application level help. Online help provides lots of advantages to developers as well as the users. In general there can be several advantages like- there will be no need for any manuals, the information can be easily updated, and the information can be easily available [Shneiderman 98]. At application level, it can provide some facilities like- explanation of errors, explanation of displayed questions, description of specific terms (or abbreviations), different guidelines, examples etc.[Shneiderman 98]. Here, we have provided different guidelines to create online help for WAP applications [Duffy, Keassley]. - Developers should anticipate the problems that the users may encounter. This can be done by doing some experiments and reviews and then try to collect as many problems as possible and prepare accurate solution for them each. - Provide the help that can be easily accessible. Nielsen found in his WAP usability experiment that out of 20, only 1 person was able to find the help service from his mobile [Nielsen 2000]. It would not be sensible to provide help
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 44 of 67
service which users could not find. So place the help form the first page of the application. It would also not be sensible to provide help for each and every page. Provide help for those pages where users get more problems as explained in previous point. Also provide all help pages/cards at one level, so that all pages can be available faster and no need to get connected next time. Provide help in a way so that user can proceed in their task. Also give stepwise guidance to accomplish a particular goal. And considering the worst case, provide some alternative ways to help users, if they cannot use online help. E.g. provide phone number or e-mail address of that service - Provide help with efficient presentation: The user should get their help from a single page. They should not have to read many pages to get answer of their help. So, each section should be self-contained. In the help contents developers should provide examples, so that users can understand it easily. "A good example can replace a thousand words". Also organise the topic hierarchy according to the structure of the site. In topics organise contents that the user can easily understand. The language and contents becomes very critical for some user groups like novices. Use the language that requires less verbal skills and should also be easy to understand. Avoid any abbreviations or any technical words. The text should be in small readable fonts and contents should be short but consistent considering the screen limitations. 3.5.2 Bookmark: Bookmark (or favourites) has always been a very useful way to browse the Internet. And considering the difficulties in mobile phones to enter the whole URL, the bookmark facility becomes very critical part of usability. Bookmark is a device-specific facility and users can bookmark only specific numbers of WAP sites. But the number of bookmarks available in Nokia 7110e was 10 and in Ericsson R320s it was 25, which are very small [Nielsen 2000]. The only way to overcome this is to increase the bookmarking capacity in mobile phones. 3.5.3 Searching: Usability studies showed that more then half of all users were search dominant [Nielsen 9707b]. While using WWW, they just go to search button and go and find their specific information. But in WAP, the issue is not the same. Because of the screen limitations the information is formatted differently in WAP sites so the searching service becomes slightly less important in WAP sites that provide specific information. Some big WAP portals like Yahoo!, Excite, etc., which provide lots of different services and have lots of pages of
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 45 of 67
information, should provide Searching service in their applications. Here, we have described some issues and guidelines to employ searching in WAP sites. The searching should very intuitive [Serco 2000]. And it is obvious that the users who will use the search mechanism may not be very familiar with the system. As a starting point, use scoped criteria so that user can select their own choice [Nielsen 9707b]. However, avoid using long navigation to find particular information. Provide flexibility of search. If possible allow user to type in search field for specific information without being taken to another screen [Serco 2000]. Also use some default criteria for users. It is recommended that users should be allowed to fill as fewer fields as possible, but sometimes it becomes necessary to ask more information from user to provide more sophisticated search result. In that case provide mandatory as well as the optional fields. The search result should also be very intuitive. The resulting list should be very short [Serco 2000], so that not much scrolling needed. If a search gets more result then the user should be asked for further refinement by adding extra criteria. The most probable result should appear on the top. Also provide some abstract descriptions about that result and also provide the level of matching of that search result e.g. give the relevance in percentage. 3.5.4 Customer Groups: As the price of mobile phones is getting lesser, more and more people are using them. The developers of WAP applications should consider different limitations and requirements, as there will be lots of different user groups, which may use the WAP phones. The customer groups can be divided into children, elderly people, novice users, less educated users, expert users, disabled people etc. And even in the disabilities, developers have to consider different disabilities like, color blind, deaf, mentally disabled, mobility disabled, etc. It is necessary to consider all of the user groups and their requirements and limitations in the development of WAP applications; however, to generate guidelines for all these user groups will be out side the scope of this research. Here we will provide guidelines for novice users. We believe that novice users are very critical for any kind of applications. One of the reasons for this is that they have actually very less or nil experience of the application and if they find the application very difficult then they will definitely not come again to use the same one. And if the application they found very easy to use then they will use it often and this will be profitable for application developers and/or service providers. It is said that WAP application should be that easier for new users that they should be able to use it without reading the manuals [Bergman 2000]. Novice users:
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 46 of 67
In order to develop WAP (or infect any) applications, the developers have to know their users very well. Developers have to know their needs, liking and limitations [Galitz 93]. In general we will call them user concern. The most common user concern we have described in section General Usability. To generate guidelines for novice users we will use the same approach. As from the name, novice users are the one who are the first time users. Developers have to design the application that can be easier to figure out and should be interpreted very easily by novice users. If the application is poorly designed, novices may find themselves completely stumped, unable to succeed, frustrated at wasting precious time without achieving a goal. On the other hand, if experts find the interface redundant and the information repetitive, they may quickly lose interest and move on to alternatives. So, balance between them is necessary here. Guidelines: All the issues and guidelines that we have described throughout the report should be applied to most user-groups. Here we have only gathered some typical and important guidelines that relate to particularly novice users. Novices require more familiarity: It is not necessary that a novice WAP user must have used WWW before, but if the WAP developers can provide similar sort of familiarity as in same WWW application then it would be helpful to novice users. E.g. in Yahoo WAP site the developers can provide the same logo as in its web site. This kind of familiarities can be provided considering other factors as well. [Dix 98]. Consistency: Consistency in interface design allows the user to build a mental model of the system, and as we know this is badly needed for novice users. Developers of WAP site interfaces can be consistence with Sequence of Tasks, Results of actions, Data across views, Functionality of tasks, and navigational patterns. E.g. the link "home" should always lead to start or home page. Simple: Even for any application and for any user group simplicity is the best policy. "KISS" (Keep It Simple and Stupid!) is well known slogan for any interactive application [Mobile 2001]. • Developers should use simple verbal language for menu options or links. • Because of the limited screen space, sometimes it becomes necessary to use abbreviations. Abbreviations should be either most simple ones or should be defined in the manuals. See Design Layout section for more information. • Do not use highly technological information. Predictable:
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 47 of 67
Predictability is closely coupled with Consistency, as consistency allows users to build the mental model of the system, which lead to make the application more predictable. As we have mentioned earlier that- telling people what will happen before they click through to a new screen is a basic usability principle [Nielsen 2000]. This is can be done by labelling the links with simple and most common meanings. Even while performing some task, the screen should provide the information about the page hierarchy and also number of pages remaining for a particular task. E.g. in wizards information like "2 of 7 page" on the top of the page. Feedback: Getting feedback from the action is also a very important and widely known principal for usability [Dix 98]. Developers can use Iconic feedback or Contextual feedback as described in Navigation section. Developers can also provide confirmation as a feedback. E.g. In some online shopping WAP site, after the user places the order a page stating, "Your order has been successfully placed" should be provided. * The error messages are also very important issue for novice users. The error message should be very intuitive and easily understandable for novice users. Recognition: As novice users are not familiar with the system, recognition becomes more important then recall. This guideline in general applies to all kind of applications all user groups as well. Recognition can also be achieved by some sort of synthesizability [Dix 98] e.g. some voice effect on each button press can be helpful. User's attention is also achieved by different presentation styles, e.g. different font size, different color on the screen, blinking texts or by some special characters like '*'. Less Input: The number of input should be kept less to perform a task. First-time users should be able to carryout simple task easily and successfully. The developers should follow the guideline of input mechanisms described in Design Consideration section. In any WAP sites the frequently used services or information should be the most easily accessible and should not be deep in navigation hierarchy.
Use Wizards: As we have earlier mentioned that wizards are more efficient then forms, especially in case of novice user. Wizards reduce the memory load from the users as they have less information to deal with and they also provide good contextual feedback to the user e.g. on the first line "page 2 of 4". Of course wizards are relatively slow compared to forms but for novice users they are more appropriate.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 48 of 67
Discoverability: The term discoverability has already been used in some [Openwave 99]. We came across this term in our evaluation. We asked some users, who haven't used WAP before, to use the WAP emulators. And we found that they were not been able to discover (or find) some links which were presented on the screen, at first instance. E.g. the link "More", when there is more links to be seen on next page. The main reason for this, we believed, was bad presentation of the screen. These links were mixed up with other links and as a result they were not easily discoverable. As a solution we suggest that for some information, which are different from common text used on the screen, developers should use different presentation (e.g. different fonts, color or size) and they should be separated from the common text. E.g. the scrolling indicator icon on the phone should also be well designed so that user can understand it. Error messages: Novice users are highly likely to make errors while using the system, so the interface design should be made to prevent the error then to recover it after it has occurred. More detailed information about error-message is described in Performance & Reliability section. Here, for novice users the error-messages become more critical. The language should be as simple and as understandable as possible. The message should not be general; users need more specific reason for errors [Shneiderman 98]. Developers can use more information on the page that can be scrolled down, as more simple information will prompt novice user to solve the problem. Section 6: Device Usability As we have mentioned earlier that all mobile head-set have three interfaces [Pekka 2001] and due to mobile handset interaction and user interface design conventions it is sometimes very difficult for the user to distinguish which part of the service is phone functionality and which belongs to the service [Bergman 2000]. It would be out side the scope of our research, to produce detailed guidelines for the mobile device. But one has to consider both the device usability and UI usability while designing the mobile system. This section will describe some issues and guidelines to improve device usability. Considering all mobile devices like mobile phones, PDAs, communicators, etc. we think mobile phones will have better future then the others. The following are the reasons for this: Mobile phones are used since late 80s in Europe and now it's used in the day-to-day life of most people in world. This shows that most of the people are more familiar with its use and functionality. So, any added functionality to the mobile phones will be easily accepted by the user groups then in any other mobile device [Bergman 2000 page 170-178]. Though having comparatively small screen interface then the other mobile devices, mobile phones do have better ergonomics then others and phones can be operated by one hand only, which is good considering some critical situations.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 49 of 67
Mobile phones are much cheaper then the others. 3.6.1 Challenges to Device Usability: There are some Basic Requirements to fulfil, which are inevitable while designing any usable system. At the user's part, the interaction is mainly carried out through sequences of button presses; and the phone may give feedback through tactile feedback, abstract sounds, or textual/animated effect on the user screen [Bergman 2000]. The mapping of user's key presses to the device's action is not always straightforward and simple. We cannot match each button to any single function. One of the ways to structure this phone functionality is a hierarchy of different factions. Users move through this hierarchy by "up", "down", "back" etc buttons. But still user has to make a mental model for this and he has to remember it. 3.6.2 Contextual Design and Task Analysis: 3.6.2.1 Contextual Design: Context of use together with specified target user group-should always be starting point for design process [Bergman 2000]. As from the name Mobile, it means variability, variability of all means. This variability, we can describe as different context of use. The analysis of the context can find out the focus on how and when the device should behave differently and offer different interaction/presentation options depending on the context of use [Rodden 98]. Contextualdesign is a huge research area, so it would be out-side the scope of our research to go in much detail.
3.6.2.2 Task Analysis: One would have noticed form this report, that WAP phone’s usability is somewhat dependent on WAP browser and WAP services (or sites). Actually, users perform their tasks on device (i.e. WAP phone) according to the presentation that they get from the UI of service and browser (i.e. WAP site). So, to consider Device Usability, one has to take into account different user-tasks that users perform on the WAP site and browser, at the early stages of design. These tasks can be ad hoc, time-critical, event driven, highly personal, short-term use, less attention [Berman 2000].
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 50 of 67
Task analysis is a process of finding out the user behaviour towards the system. It is a Black Art, and is often not given proper interest. Even the current Internet is having some problems for not employing task analysis. Task Analysis helps making the complete design process. It helps designer to find out the sequence of usage and scenarios of user's work format. From these scenarios more detailed Task Analysis can be made [Berman 2000]. The user goals and tasks can be divided into sub-goals and sub-tasks that all aims to a specific goal. As we have mentioned in previous section, different device functionality and services are mapped into the hierarchy and user has to go through this hierarchy by soft-keys and other buttons. By using the task analysis, after finding out different priority-level of user-tasks, the designer can set up easy button-navigation for most critical tasks and more traversal navigation to less critical tasks. Even some dedicated buttons can be set-up on the phone for a very important or extremely critical task. E.g. for switching off the mobile phones a separate button on the top right on the phone can be set-up. At the same time designer has to provide Familiar symbols and easily Learnable and understandable icons for different tasks/services button on the phones. 3.6.3 Input-output issues: 3.6.3.1 Output issues: Both the Mobile phone and PDA are having different layout and size. The user screen is comparatively bigger in PDA then in phone, which means PDA can be optimised for more information presentation by space utilisation. The user-screen is fixed in size and can only be optimised at the application-level by improving the various interaction mechanisms. These mechanisms are described in Design Consideration section. As there are limitations on the size of the phone, the designers should try to provide less dedicated buttons on the phone. Soft-key Issues: Almost all mobile phones have 2 soft-keys. These soft-keys work as function-keys of the calculators. They provide extra functions to use on the screen. There are some issues to consider in these Soft-keys. Number of soft-keys affects the usability. If there are less then 2 soft-keys on the device, then there should be other alternative keys (dedicated keys) or some mechanisms at the screen interface. But both of this will divert the user's focus, and will affect the speed of navigation [Bergman 2000]. Even having more soft-keys will make confusion while remembering them. It is reasonable to have at least 2 soft-keys. As soft-key works as functional-key the overall performance of system can be optimised by “number of soft-key pressed for a task”. Soft-keys also provide good recognition of functions as they always appear on the bottom phone screen, which help the novice users. 3.6.3.2 Input issues:
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 51 of 67
World is becoming fast-paced and people having no time to press a key three times to enter a single character. On the other hand, users also have ‘too little’ or ‘no’ time to train themselves. Even inputting data using regular keypads gets confusing when users try to enter special characters. As the number of cellular phone users are growing, low performance times and high error rates will lead to frustration, unnecessary stress and will discourage novice and elderly from taking advantage of the newer technologies. Users must therefore, be able to enter text efficiently. Compared to other mobile devices, input mechanisms in Mobile phones are slower then others because button interface is much more difficult while navigating links. For effective mobile phone usage there should be some alternative mechanisms for inputs. To provide high usability, some of the mechanisms can be described as following. Researches have proved that inputting numerical letters are much easier and faster then the textual letters. Some of the experiments have also shown that textual input is approximately twice as slow as numerical input and that the error rate is about twice as high [SAP 2001], but this has to be dealt at application-level. In the traditional Mobile phone 12-key pad is used for input. But it has some problems e.g. to input a single letter user has to press 3 or 4 buttons at the extreme cases. There is one other input method called T-9 [Nokia 2002], which uses a built-in dictionary and adds knowledge to the system itself. It only requires one keystroke. For example, to enter "this", the user enters 8-4-4-7. T9 then uses the combination of letters and compares it to the word possibilities in its dictionary to then "guess" the intended word. This creates problems as many multiple words may have the same key sequence. T9 then guesses the most common word. Users can then press "*" to view the next possible word. The performances of the novice and expert users are very different. The novice using T9 for the first time will have to go through a training and practices session whereas expert users will already be adapt. Expert users will already be familiar with the way T9 works. Voice Interface: Recent advances in wireless communication and speech recognition have made it possible to access the web from any place, at any time, by using only a phone. Some possible applications are browsing the web, getting stock quotes, verifying flight schedules, getting maps and directions, checking email, etc. Companies like IBM, Motorola, Lucent and AT&T founded in 1999 the VoiceXML Forum to deal with speech technologies that make Internet accessible [Voice 2002]. Speech technology contributes to universal design by creating access mechanisms that appeal to both people with or without disabilities. The other two major input entry methods are handwriting using Graffiti alphabet, which is a hand stroke based handwriting recognition system; and tapping on a soft keyboard. These two mechanisms are not used in Mobile phones for now but they are very effective in other mobile devices like PDA. In the previous research of early 90s, the screen keyboard was said to be faster then Graffiti; but as there are more improvements in handwriting recognition software, now Graffiti has likely the same speed as the screen keyboard. However, Graffiti is having some other advantages as well.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 52 of 67
3.6.4 Flexibility & Customisability: One of the best ways to provide the Usability in a system is to provide better flexibility and customisability in the system. Users should be allowed to customise the interaction in their own way, but up to some extent. User should not be forced to become designers of their system. Functions that allow, but not force, users to change some aspects of their device can significantly increase the subjective satisfaction with the device. E.g. let the users allow placing their favourite application on a button, which is ergonomically good.
Chapter 4: Evaluation 4.1 Introduction: In the case of any interactive system, the main goals of an evaluation will be to assess the usability and functionality of the product and find out problems, if there are any. For our research, the evaluation will be a critique of two different WAP sites currently available. Our main aim will be to find out how usable the current WAP sites are. For our evaluation we asked few people, each at a time, to perform set of tasks on Nokia Mobile Internet Toolkit v3.1 -emulator and asked them some questions about usability of
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 53 of 67
those WAP sites and their suggestions about the sites. Obviously, evaluating from any WAP emulator will not provide the same results, in terms of usability, as the original WAP enabled phones. In the evaluation of WAP emulators, some issues like ergonomics, interaction styles, input mechanisms, etc. will not be applicable as they are totally different in WAP enabled phones. But here, our main concentration will be on the presentation and design layout of these WAP sites. The main reason for this evaluation is to apply the guidelines that we have generated in our research and find out how the current WAP sites and portals apply those guidelines. We also want to know what factors and issues would be important for users and where they might have problems while using it. So, from the nature of the evaluation this would not be any statistical experiment but would be a simple critique. 4.2 Methodology: From the nature of our research, we cannot apply any field or laboratory studies to evaluate these sites. For our evaluation a method, developed by Jacob Nielsen and Rolf Molich, called heuristic evaluation will be the most appropriate. Of course, they had originally developed it for desktop user interfaces but we can use it for handheld device interfaces as it uses the most common principles. Heuristic evaluation luckily fits very well to our evaluation as it help to find out usability problems in user interface design [Nielsen 10]. It would also be advantageous to have some observers (or participants) for assessing the same sites as they all could find out different usability problems [Nielsen 10]. In addition to the heuristic evaluation, it would be helpful to use usability principles described in [Dix 98]. 4.3 Design of evaluation: We asked four people, 3 male and 1 female, in person, to use Nokia Mobile Internet Toolkit v3.1 - emulator and perform three tasks on two different WAP sites each. We have deliberately selected them as they all were familiar with mobile phones and none of them has ever used WAP before. The test design is shown in the figure-12.
Initial Interview
User Tasks
Final Interview
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 54 of 67
Figure 12- Design of Evaluation 4.3.1 Initial Interview: First, we asked our participants some general questions about their experiences in the Internet usage, mobile phone usage, their personal interests, etc. The full list of questionnaire can be found in Appendix-A. The main reason for these questions was to know their individual experience and interests. We also showed the demonstration of that Nokia emulator and taught them how to use it. 4.3.2 User Tasks: Here, we asked our participants to use two WAP sites, first wap.yahoo.com and then wap.hollywood.com and asked them to perform 3 tasks. The importance of this test was not just the completion of the given tasks but how they completed them and what did they like about the sites and what they didn't. Here, we have to interpret user's action in order to infer how these actions are related to the usability issues in designing the interfaces. Initially, we have also informed the users to speak, without hesitating, through out the test about the problems they may face while using the site. That is because they might forget some issues after the completion of the test. The more they say about the site the more it would help us to criticise it. We also generated the result of the test, which is shown in Appendix B. 4.3.3 Final Interview: After completion of those tasks we asked our participants to criticise and comment on the issues about the design layout and presentation of those two sites. 4.4 Analysis: It would be sensible to write a common critique from all four participants rather then their individual ones because there were lots of common views we found in this evaluation. We have divided our critique in 10 different parameters as described in [Nielsen 10] by Jakob Nielsen and allocated our critique into it. We have also applied usability principles described in [Dix 98] to make our critique more concrete. As we have mentioned earlier that in some cases there is a very blurred distinction between the user interface (provided by the WAP browser) and service interface (provided by the WAP application). Hence, in our critique we shall mention both them in conjunction as they complement each other in the overall usability. Here, we will use just Yahoo to address Yahoo-WAP site and Just Hollywood for Hollywood-WAP site. The following are the main points of our critique. 4.4.1 Visibility of system status
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 55 of 67
“The system should always keep users informed about what is going on, through appropriate feedback within reasonable time.” Whenever the system is processing some task, the browser provided by the Nokia emulator shows a rotating icon every time. This is also important because it is shown on the same topright corner of the Internet-browsers such as the Internet Explorer. In both Yahoo and Hollywood, on the top of each card there is one header showing the service-name or menuname. In Yahoo especially, the whole sequence of menu and sub-menu is also shown on the top of card when the user is in the sub-menu. E.g.
Y! Sports - Cricket Live News Menu
Back
Figure: 13 Example for visibility This can give good indication of the navigational hierarchy as well as current card status. The Hollywood shows only the current menu on the first line. In Yahoo, the visibility is given a good importance as they also have shown a service with an indication as “X” (cross mark), which was not available at that time. 4.4.2 Match between system and the real world “The system should speak the users' language, with words, phrases and concepts familiar to the user, rather than system-oriented terms. Follow real-world conventions, making information appear in a natural and logical order.” Both the sites have used very pleasing and unambiguous language in the links and in other text, so that almost all type of users can understand it. The logo of the site can provide more confidence to the user about their correctness and it can also provide better familiarity. Yahoo provides the logo but not Hollywood. The order and summarisation of the links should also be efficient, it should follow the rule of “most important links should come first”. Yahoo doesn’t follow this rule well; one of the reasons is that it is a big portal and it would be difficult to show all the important links. Hollywood is a very specific domain and it has done good link usage. 4.4.3 User Control and freedom
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 56 of 67
“Users often choose system functions by mistake and will need a clearly marked "emergency exit" to leave the unwanted state without having to go through an extended dialogue. Support undo and redo.” The browser supports back function by employing back soft-key on the screen, which works well. Hollywood’s site has facility of back, by providing a “back” link, in some pages where users might though to go back e.g. some leaf cards, where user has no further navigation. But this is not there in higher level pages. It also had “more” link at the same time is the users want to read more about that. But Yahoo didn’t have this facility. Hollywood site can be called to be of a specific domain, and Yahoo is a portal. In big portals there should be a search engine or help mechanism that can help user if they want to find some specific information, but these are not there in Yahoo’s site. 4.4.4 Consistency and standards “Users should not have to wonder whether different words, situations, or actions mean the same thing. Follow platform conventions. “ The consistency can be applied in positioning the widgets (or similar as in Mobile phones) and naming of links and options. In yahoo we observed that on some pages they have used ‘Back’ and in others ‘Prev’; this had actually caused problem for a participant working with us. He though it might be used for ‘Preview’ instead of Previous. Yahoo has used the same naming as it is in Yahoo web-site. There was also lack of consistency in Hollywood’s movie ‘Symptoms’, as some movies had long information and some had nothing written about them. 4.4.5 Error prevention “Even better than good error messages is a careful design which prevents a problem from occurring in the first place.” One of the main sources of errors, especially in mobile phones, is the input process. And here we have used an emulator and we are only going to evaluate the user interface, so this issue can not be applied here. In emulators most of the navigation is link-based so there is less chance for making errors. But we think, in this browser they haven’t kept any default selections, so user might just press a button but nothing works for him. 4.4.6 Recognition rather then recall “Make objects, actions, and options visible. The user should not have to remember information from one part of the dialogue to another. Instructions for use of the system should be visible or easily retrievable whenever appropriate.“ One of the most interesting issues that we found in this evaluation is Discoverability. We found that one user couldn’t find a link called “More”. This is used when there is more information available on next page. And reason that he said was he couldn’t easily
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 57 of 67
distinguish that link from the other links. Here comes the concept of discoverability, we have defined it as “how easily a user can distinguish (or discover) a link or other widgets from the interface”. If that more links had a different color or different font size/style (i.e. a mechanism that easily distinguishes a thing), the user would have easily found that link. Scrolling is always an important issue for mobile devices. Here this emulator showed good indications about the scrolling by adding an arrow at the bottom. Both of the sites have not used any good alert mechanism (e.g. symbolic “New”), which can be used in newly arrived news or new movie. In yahoo, as we have mentioned above, they have used how navigation hierarchy in the header, which gives good recognition of user’s task path. None of the sites have used any voice-synthesises, which can be applied here. 4.4.7 Flexibility and efficiency of use “Accelerators -- unseen by the novice user -- may often speed up the interaction for the expert user such that the system can cater to both inexperienced and experienced users. Allow users to tailor frequent actions.” Hollywood gives two way to book a ticket, one can book the tickets from the first page and for other users they can go and see the information about the movie and after they can book the tickets. In Yahoo, all long text is wrapped to next line and as a result it becomes difficult for user. And even user selects a link, it becomes very confusing to them. In Yahoo-News, one first screen a link, which just informs about that news in 1 or 2 lines and after clicking it user gets the whole news. They should also provide a summery of news because that news includes lots of scrolling to the user. But one good thing about it is, in each of the news they had provided their references, mostly from Reuters. This gives good confident in user as they have some evidence for those information. In Hollywood, while browsing different area code users can’t go back straight to first page. They have to re-navigate through the whole sequence again. In Yahoo TV-guide, in channel 4 schedule user saw the schedule always only up to 5pm, so there was less efficient updating. 4.4.8 Aesthetic minimalist design “Dialogues should not contain information which is irrelevant or rarely needed. Every extra unit of information in a dialogue competes with the relevant units of information and diminishes their relative visibility.” Both sites have followed this principle very well. There was not any decorative or fancy thing on the screen. User can do their important task without scrolling.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 58 of 67
4.4.9 Help users recognise, diagnose and recover from errors “Error messages should be expressed in plain language (no codes), precisely indicate the problem, and constructively suggest a solution.” This browser uses “back” functionality automatically when the user goes to some unavailable card or types a site that doesn’t exist. So, this way the browser prevents the confusion and also helps the user by taking them to the previous page. This functionality, however, is really good if it can be employed in the real WAP phones. The only error-message that all user came across was accessing the Yahoo-Mail. The users were getting error, but it was not well structured and didn’t prompt the users that well. The only understandable thing about that error was that, it just said, “XYZ has been moved to ABC. Go to new location”. For a normal user this can’t serve as a good reason or better way of prompting him. 4.4.10 Help and documentation “Even though it is better if the system can be used without documentation, it may be necessary to provide help and documentation. Any such information should be easy to search, focused on the user's task, list concrete steps to be carried out, and not be too large.” This principle may not be applicable to our evaluation as we are only dealing with the usability of browser and WAP sites. However, it is important to provide help and documentation separately and if possible on the sites as well. None of these sites have help functionality.
Chapter 5: Final Discussion & Conclusion: In this report we have described issues and guidelines to improve the usability of userinterface and service-interface of WAP phones. Considering the different HCI-challenges for mobile devices, as described in the introduction, one of the issues is to developer better interfaces. In this research, we spent lots of time on providing better interface design. So, here we will be able to comment on those issues. We have discussed about input layout, and found out that inputting from small key-pad of mobile phone is a difficult task so the developers have make the application accordingly. In order to do that try to minimise the input fields in the application and if it becomes inevitable to use, it then try to use the numerical input first. And at the worst case, ask for those letters that can be easily pressed from the mobile phone (e.g. the first letters from the buttons). To
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 59 of 67
prevent errors, input fields should be labelled in a way so that user can clearly understand what type of input is required, user default inputs and give proper examples to instruct users while inputting. Developers should only ask for password when the application is very critical, for otherwise there is no need for it. We also saw different presentation mechanisms. Developers should provide the links in a way that users can assume it’s destination. This can be done by better verbal usage. Getting feedback from the work is also a well-known slogan in usability. Systems should always provide feedback of its status and its progress. The feedback can be given by Iconic presentation as well as Contextual presentation. The contents that are provided on the screen should be very precise, because there is not much space. So, better summarising technique should be used. Scrolling and page-size are also very critical issues for presenting information. Scrolling will not be a bad idea if the options presented on the screen are well structured. Also having long page (i.e. more scrolling) can cause better page utilisation, as one can put more options on the screen and user can have direct access to those information. And having fewer cards needed for the navigation means there will be less reloading of new-pages each time. Here, menu architecture also comes to attention. The shallower and broader the menu design the more efficient navigation can be. Colors can be used as for better Discoverability, a usability principle that we found during our research. Better color presentation gives better recognition about the interface status. Error messages are also important here. Error messages should provide clear reason for the error and should also prompt user to overcome the error. Here, we would also like to mention that we have tried to gather all the issues affecting the WAP usability but some of them we have left out. E.g. context sensitivity is also important in task analysis that we have not mentioned here. At the system architecture level we have just talked about proxy mechanisms. One other important thing we have observed here, form its current status, that WAP can have good future in Entertainment, Sports, News, etc services as these services doesn’t need much secured or accredited information; rather then in financial services like banking, stock information or shopping. Even DoCoMo’s I-mode has gained more subscribers using entertainment applications. We can summarise the whole conclusion here in one sentence that “For better usability provision, a good user-orientation with better systems resources is inevitable.”
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Alan Dix - Janet Finlay - Gregory Abowd - Russell Beale. www.hcibook.com
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 61 of 67
[Nielsen 2000]: WAP Usability - Déjà Vu: 1994 All Over Again, Report from a Field Study in London, Fall 2000. By Marc Ramsay and Jakob Nielsen, Nielsen Norman Group, December 2000. [Bergman 2000] Information Appliances and Beyond, Interaction Design for Consumer Products, Eric Bergman (ed.), Morgan Kaufmann. [Mobile 2000] Mobile Usability. A report by Sabrina Duda, Michael Schiebl, Jan Michael Hess of Mobile Economy. [Hurst-Gellady 2000] “Building a great customer experience to develop brand, increased loyalty and grow revenues,” by Hurst, M and Gellady, E. www.creativegood.com/creativegood-whilepaper.pdf [Spiegel 2000] “Report: 70 percent of retailers lack e-commerce strategy.” Ecommerce Times. www.ecommercetimes.com/news/articles2000/000126-1shtml , Jan. 26 [Beyer-Hugh 1997] “Contextual Design: A customer-centred approach to system designs”, 1997, Morgan Kaufman Publishers, ISBN: 1558604111. By, Beyer, Huge and Holtzblatt, Karen. [Openwave 99] Developing User Friendly Applications, Application Usability Guidelines from the Phone.com Application Development Team. (August 1999). [Project Management] Essentials of Project Management. By Dick Billows. [SAP 2002] SAP Interaction Design Guide for WAP Applications, (Feb 2002) http://www.sapdesignguild.org/resources/wap_guidelines/wap_guidelines.pdf [Telenor 2000] User Interface Design Guidelines for WAP Applications, Version 1.4, 1st of October 2000. By Tone Grimstad, Harald Stegavik, and Erik Saastad. The Usability Section, Telenor Mobile-Communications. [Nielsen 10] Ten Usability Heuristics. by Jakob Nielsen. http://www.useit.com/papers/heuristic/heuristic_list.html
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 62 of 67
[Little 2002] Mobile Style Guide – Sample, Little Springs Design, a division of Little Springs, Inc http://www.littlespringsdesign.com/design/sampleguide.html [Pekka 2001] The Three Facets of Usability In Mobile Handsets. By Pekka Ketola & Mika Röykkee. Nokia, Visiokatu 1, FIN-33720 Tampere, Finland. [Brew 2002] BREW User Interface Design Guidelines, 80-D4231-1 Rev. A. March 7, 2002. http://www.qualcomm.com/brew/developer/developing/docs/80-D4231-1_A.pdf [Nielsen 94] Usability Engineering. Book by Jakob Nielsen, published by Morgan Kaufmann, San Francisco. ISBN 0-12-518406-9 (paperback) [Arizona 2001] www.cs.arizona.edu/classes/cs630/fall01/ 630-1/presentations/proxy.PDF [Dix 99] “Places to stay on the move” Software architectures for mobile user interfaces. By Alan Dix, Devina Ramduny, Tom Rodden, Nigel Davies. www.hiraeth.com/alan/topics/mobile/ [WebExpress] WebExpress: A System for Optimising Web browsing in a Wireless Environment. By Barron C. Housel and David B. Lindquist. [Rodden 98] Exploiting context in HCI design for Mobile Systems. T. Rodden, K. Cheverst, N. Davies and A. Dix (1998). Workshop on Human Computer Interaction with Mobile Devices, Glasgow, 21st & 22nd May 1998. [Nielsen 93] Usability Engineering. Book by Jakob Nielsen, published by Morgan Kaufmann. [Openwave 10] “Top 10 Usability Guidelines for WAP Applications” developer.openwave.com/resources/uiguide.html [SAP 2001] “WAP- Designing for small user interfaces”. By - Oliver Frick and Albrecht Schmidt. www.sapdesignguild.org/innovation/artikel/wap-design1.html (last accessed on 17/08/2002)
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 63 of 67
[Nokia 2002] www.nokia.com [Voice 2002] www.voicexml.org/ [RedWhale] “Adapting to Mobile Contexts with User-Interface Modelling”. By Jacob Eisenstein, Jean Vanderdonckt, and Angel Puerta. RedWhale Software Corporation, Town and Country Village Suite 273-277, Palo Alto, CA 94301. [Bhagat 94] Bhagwat, P. & Satish K. Tripathi (1994). Mobile Computing. In proceedings of Networks’94, pp. 3 – 12. [Satyanarayanan 95] “Fundamental Challenges in Mobile Computing”. By M. Satyanarayanan School of Computer Science Carnegie Mellon University [Kuutti 2002] “Challenge of Human-Computer Interfaces for mobile systems and services” By, Prof. Kari Kuutti. Deprtment of Information Processing Science, University of Oulu. http://www.ee.oulu.fi/~ato/ems2002/kuutti.pdf [Jesper 2002] “Just-in-Place” information for mobile device interfaces. By, Jesper Kjeldskov, Department of Computer Science, Aalborg University, Denmark. [Nielsen 9707b] Jakob Nielsen's Alert box for July 15, 1997: http://www.useit.com/alertbox/9707b.html [Nielsen 99] “User Interface Directions For The Web”. Communications Of The ACM. January 1999/Vol. 42, No. 1. [SAP-MiniApp] SAP MiniApp Guidelines. http://www.sapdesignguild.org/resources/ma_guidelines/ [Becker 2001] “A Usability Perspective on Wireless Internet Technology”. Shirley A. Becker Computer Science & Software Engineering, Florida Institute of Technology, 150 West University Blvd. Melbourne, Florida 32901.
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 64 of 67
[Jones 99] “Improving Web Interaction on Small Displays”. By Matt Jones, Gary Marsden Norliza Mohd-Nasir, Kevin Boone, George Buchanan. Interaction Design Centre School of Computing Science, Middlesex University, UK. [Legge 85] Legge, G., Pelli, D., & Schleske, M. (1985). Psychophysics of Reading—1. Normal Vision. Vision Research, 25, 239-252. [Tulis 95] Readability of Fonts in the Windows Environment. Tullis, Thomas, Boynton, Jennifer, & Hersh, Harry. http://www.acm.org/sigchi/chi95/proceedings/intpost/tst_bdy.htm [ Duffy] T.M. Duffy al et. "Online help: Design and Evaluation". [Kearsley] G. Kearsley "Online help systems: Design and Implementation". [Marsden 2001] “Ubiquitous Computing and Cellular Handset Interfaces – are menus the best way forward?”. Gary Marsden a Matt Jones b [Ericsson 2001] http://www.nada.kth.se/~hehu/usability/WAPStudy_bw.pdf [Miller 81] The Depth/Breadth Tradeoff in Hierarchical Computer Menus. By Miller, D. P. Proceedings of the Human Factors Society 25 the Annual Meeting pp. 296-300. 1981 [Norman 88] The effect of tree structure on search in a hierarchical menu selection system. By Norman, K. L., & Chin, J. P. Behaviour and Information Technology, 7, pp. 51-65. 1988 [DoCoMo's I-mode] www.nttdocomo.com [Movie 2001] MIE 1404S: Movie Listing Service for the Mobile Phone Group 8: Final Report http://www.imedia.mie.utoronto.ca/people/takesh/html/MIE1404S/FinalReportE.htm [WebTv 97] WebTV Usability Review Jakob Nielsen's Alertbox for February 1, 1997:
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 65 of 67
http://www.useit.com/alertbox/9702a.html [Buchanan 2001] Improving Mobile Internet Usability, by George Buchanan, Sarah Farrant, Matt Jones Harold Thimbleby. May 2001. [Allnet 1] http://www.allnetdevices.com/developer/opinions/2001/08/27/usability_review.html [Allnet 2] http://www.allnetdevices.com/developer/opinions/2001/07/25/wireless_usability.html [GSM 2001] “GSM application style guide” //developer.openwave.com/ [Shneiderman 1982] “Designing Computer System Messages”. Communications of ACM 25, 1982. [Norman 83] “Design Rules Based On Analyses Of Human Errors”. Communications of ACM 26, 1983. [Brown-Gloud 1987] “Human Computer Interface Design Guidelines”. Albex, Norwood, NJ. [Shneiderman 98]: Designing the User Interface, Third Edition. Addison-Wesley Publishing Company.
Appendix A: Personal Information Appendix B: User Tasks & Results
Appendix A: Personal Information: MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 66 of 67
Name: Sex: Age: Occupation or Work
How long have you been using Internet? Less then a year
1-2 year
3-4 year
more then 5 years
How many hours a day you work on Internet? Less then 1
2-3 hours
4-5 hours
more then 6 hours
Which of the following area services you use more? Entertainment
News
Science
Government
Arts
Education
Business E-mail
Travel
Sports
Shop
Computer related
Chat
How long have you been using the Mobile phone? Less then 1
2-3 years
more then 4 years
Have you used any WAP phones before? Yes
No
Do you have any knowledge about HCI or Usability studies? Not heard
Just heard
Familiar
In WAP phones which service you think will be most important to you? Why?
Appendix B: User Tasks & Results
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.
“Investigating Usability Issues in WAP Applications”
Page 67 of 67
User Tasks: 1) wap.yahoo.com Task 1: Check the e-mail Task-2: Check out news Top-stories Task-3: Find out 8 o’ clock program on Channel 4 2) wap.hollywood.com Task 1: Find the latest movie Task-2: Find any theatre for that movie Task 3: Check out movie news.
Results:
Participant (No.)
wap.yahoo.com Task-1
Task-2
wap.hollywood.com Task-3
Task-1
Task-2
Task-3
1 2 3 4
MSc DISSERTATION for Advanced MSc in Distributed Interactive Systems, Lancaster University.