+1 (208) 254-6996 essayswallet@gmail.com
  

finish it in 12 hours, no cites needed and only use the book I provide.

Do your best to use the terms and techniques that we presented in this book. Make sure you answer each question explicitly and address all parts of the question. 

Consider the following problem: The Regional Transit Authority (RTA) would like to increase ridership on the buses used to serve university and two other cities.  They have hired us to help them implement technical solutions and software to achieve their goals.

1. From the perspective of the University Community, who are the different stakeholders (including users/riders) that the RTA should consider for this project? As per the rubric, we ask for at least 4 stakeholders. 

2. For each of these stakeholders, what are the specific activities we can engage in to get a better understanding of their needs?

3. What kinds of technology could/should we be building to address these needs? Explain. (Check rubric for expectations/examples)

4. For each technology you identified in question 3, how can/shall we evaluate whether we are successful? Explain.

5. What information and resources do you believe we will require in order to do the complete project? Explain. Provide at least 2-3. 

1 Introduction

1 Why the User Interface Matters Human–computer interaction (HCI) is the study of how humans interact with computer systems. Many disciplines contribute to HCI, including computer science, psychology, ergonomics, engineering, and graphic design. HCI is a broad term that covers all aspects of the way in which people interact with computers. In their daily lives, people are coming into contact with an increasing number of computer-based technologies. Some of these computer systems, such as personal computers, we use directly. We come into contact with other systems less directly — for example, we have all seen cashiers use laser scanners and digital cash registers when we shop. And, as we are all too aware, some systems are easier to use than others.

When users interact with a computer system, they do so via a user interface (UI). This book explores how to design good user interfaces — interfaces that are easy to use and easy to understand, that meet the needs of the intended users, and that support users in the tasks they wish to undertake. In this part of the book, we intro- duce you to user interface design and evaluation. In particular, we explain why good user interface design is important and highlight the consequences of poor or bad user interface design. More important, we will get you to start thinking about users — and why and how to involve them in the design and evaluation of the user interface.

2 Computers Are Ubiquitous Technology has advanced so much that computer systems are used on an everyday basis by almost everyone. A computer system (or an interactive computer system or just a system) is the combination of hardware and software components that receive input from, and communicate output to, a user in order to support his or her performance of a task (see Figure 1.1). Computer systems may be used directly, as in the case of personal computers (PCs) in use at work or at home. Often, though, we use embedded computer systems where the technology is invisible to us. For

3 Part 1

example, computer-based mi- crochip technology can be found embedded in personal goods such as digital watches and mobile phones, in do- mestic appliances such as microwave ovens, washing machines, and video re- corders, and in the instru- ment panels of cars. Again, but less directly, computers are used when we shop; many stores use laser scan- ners that “swipe” the bar codes on goods to record both the goods we purchase and total the amounts we spend. Behind the scenes, the scanning of goods assists with automated stock control

and stock reordering. When we take money from our bank accounts using an auto- mated teller machine (ATM) or when we use ATM debit cards to buy goods elec- tronically, our bank details are accessed via the bank’s computer system. The list of everyday ways in which we use computer-based systems seems endless.

Whether we are aware of it or not, computers pervade our life. Computer applica- tions are used either by us, or for us, in some way almost every day. The user inter- face (or just interface) is that part of the computer system with which a user interacts in order to undertake his or her tasks and achieve his or her goals.

The user interface and the ways of interacting with computer-based systems are dif- ferent for each system. For example, digital watches generally have buttons that users press to set the time or use the stopwatch facility. Microwave ovens might have dials to turn or a digital display and a touchpad of buttons to set the cooking time. PCs have a screen, a keyboard, and a mouse (or sometimes a trackball or a joystick) that enable interaction to take place. So each user interface is different. Depending on the design of the interface, each of these systems will either be usable — that is, easy to learn and easy to use — or problematic for users.

Earlier we described a computer system as the combination of hardware and soft- ware components that receive input from, and communicate output to, a user to support his or her performance of a task. Although the user interface is simply the part of the computer system that enables interaction and serves as a bridge between users and the system, to users the interface often is the system (Constantine and Lockwood, 1999). The user’s view of a computer system is often limited to and based solely on his or her experience of the user interface (see Figure 1.2).

CHAPTER 1 | Introduction

4 Part 1

Figure 1.1 The interface is the part of the computer system with which the user interacts in order to use the system and achieve his or her goal.

For example, when you use the controls on the panel of a washing machine, the con- trols form the interface between you and the machine — you are not concerned with the underlying technology or the software of the washing machine itself. What is important to you is that the controls and their settings are intuitive and easy to under- stand and use so that you will achieve your goal of laundering clothes. Similarly, when you surf the Internet, the pages of a web site displayed on your PC’s monitor form the interface between you and the site. The web page UI may contain controls like scroll bars, clickable hot spots, or links in the form of text or images. These items are all part of the interface.

3 The Importance of Good User Interface Design Good user interface design is important because, as we have discussed, computer use permeates everyday life. Early computer systems were expensive and were devel- oped mainly for particular tasks, like advanced number-crunching; as such, these systems were employed only by specialist computer users. Often the systems had command-line interfaces, with obscure commands known only by these specialist users. Thus, the user had to adapt to the system, and learning how to use the system required much effort.

Computing systems, however, are no longer the province of the specialist user. As the price of PCs and computer-based technologies has fallen, the ownership of these types of goods by nonspecialists has widened. In August 2000, 51% of households in the United States had access to one or more home computers, and 42% of house- holds had access to the Internet (U.S. Census Bureau, 2001). In 2002, 54% of house- holds in the United Kingdom had access to some form of home computer, and 44% had access to the Internet (National Statistics, 2004). Therefore, the need for the design and development of user interfaces that support the tasks people want to do and that can be used easily by a variety of people with varying abilities has become

3. The Importance of Good User Interface Design

5 Part 1

User interface

User input

System output

Underlying hardware, software, interaction devices

Figure 1.2 To the user, the interface is the computer system. (From Constantine and Lockwood, 1999.)

You will learn more about command-line interfaces and other interaction styles in Chapter 11.

The design of controls, and the selection of interaction devices for input and output, will be discussed further in Chapters 12 through 14.

an important issue. Users are more comfortable with computer systems that are easy to use, easy to understand, and enable them to attain their goals with minimum frustration.

One way of demonstrating the importance of good user interface design is by showing tangible benefits that can be discussed in cash terms. For businesses, good user interfaces can lead to benefits such as higher staff productivity, lower staff turnover, higher staff morale, and higher job satisfaction. Economically, these bene- fits should translate into lower operating costs. In addition, computer systems that are easy to use and easy to understand require less training, again saving employers money. Bad user interfaces, on the other hand, may result in stress and unhappiness among staff, leading to high staff turnover, reduced productivity, and, consequently, financial losses for the business. As you will see later, it is easy to give examples of the effects of bad design, but showing the financial benefits of good user interface design can be more difficult. Invariably, many factors are involved and this makes it difficult to attribute success directly to good user interface design.

3.1 What Is a Good User Interface Design?

A good user interface design encourages an easy, natural, and engaging interaction between a user and a system, and it allows users to carry out their required tasks. With a good user interface, the user can forget that he or she is using a computer and get on with what he or she wants to do. Just as knowledge of the transmission mech- anism of a car is of little concern to most motorists, knowledge of the internal work- ings of a computer system should be of little consequence to its users.

Although we have used the adjectives “good,” “poor,” and “bad” to describe user interfaces, it is worth noting that each of these terms is subjective: they have differ- ent meanings for different people and their use to rate various aspects of a user inter- face will vary. You may have used the terms “good” or “bad” to describe, for example, the colors used in an interface, the pictures on the icons, or how attractive or eye- catching the interface was. These attributes describe the overall look or aesthetics of the UI. Nevertheless, they are only a part of our focus in this book. Our real concern is whether a user interface is good, bad, or poor in relation to its usability.

� What Is Usability?

Usability is defined in Part 11 of the ISO 9241 standard (BSI, 1998) as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use.” Effectiveness is the accuracy and completeness with which specified users can achieve specified goals in particular environments. Efficiency is defined as the resources expended in relation to the accuracy and completeness of the goals achieved. Satisfaction is the comfort and acceptability of the work system to its users and other people affected by its use.

Note two key aspects of this definition of usability. First, to be usable an interface should be perceived as being usable by the specified users — users for whom the system has been designed and developed. Next, the scope of focus for the design of

CHAPTER 1 | Introduction

6 Part 1

We discuss some alternative definitions of usability in Chapter 6.

the interface is extended by looking beyond the users’ immediate work environment and looking at the wider context or situation within which the system is expected to operate (i.e., the domain, tasks, and the environment that make up an organization). Thus, usability is concerned with the extent to which users of an application are able to work effectively, efficiently, and with satisfaction in their particular contexts.

A computer system that is usable in one context may be unusable in another. As a user interface designer, it is important to consider the context in which the system will be used. A UI that users find pleasurable is likely to be more acceptable than one that annoys them. Users are more likely to use a computer system that they enjoy than one that irritates them. Contented users are likely to be more productive, so usability is clearly related to user satisfaction (Constantine and Lockwood, 1999).

3.2 The Problems of Poor or Bad User Interfaces

� User Frustration and Dissatisfaction

Problems for users and the public in general arise as a result of poorly designed user interfaces. The term “computer rage” was coined in 1999 following a Market & Opinion Research International (MORI) poll conducted on behalf of Compaq Com- puter Limited, UK and Ireland. The study, Rage against the Machine (Compaq, 1999), (www.mori.com/polls/2002/bthomecomputing.shtml) found that, for one reason or another, stress and frustration levels with workplace technology are rising. Workers, it reports, have started to become both verbally and physically abusive toward the information technology (IT) in use (see Figure 1.3 and Box 1.1). Concerning mone- tary matters, the study indicates that

[t]he cost to business of this increase in stress levels of employees is not only based on sick days or under-performance, but also the working time lost through waiting for IT problems to be solved. Confederation of British Industry (CBI) statistics cur- rently evaluate this at a staggering £25,000 ($40,000) per person in lost business each year (based on one hour a day being spent sorting out IT problems). (p. 4)

In October 2002, research for British Telecom (BT) Home Computing, (www.mori.com/ polls /2002/bthome-topline. shtml) again conducted by MORI, found that 70% of per- sonal computer users suffered from “PC rage” — that is, the users surveyed admitted to shouting, swearing, or being violent to their computers when problems like crashing or virus infections arise.

You will learn more about domains, tasks, and environ- ments in Chapter 4.

3. The Importance of Good User Interface Design

7 Part 1

Figure 1.3 Computer rage: Workers have started to become physically (and verbally) abusive toward IT.

CHAPTER 1 | Introduction

8 Part 1

Box 1.1 Man Shoots Laptop

A 48-year-old man, George Doughty, was allegedly so frustrated by his laptop crashing that he took a handgun and shot it four times. According to police he apparently hung the destroyed laptop on a wall as if it were a “hunting trophy.” Lafayette police officer Rick Bashor told local newspapers, “It’s sort of funny, because everybody always threatens their computers, [but] it’s the first time someone shot a computer because he was upset with it.” The man admitted to police that he should not have shot his laptop, but that it seemed appropriate to at the time.

From http://news.bbc.co.uk/1/hi/world/americas/2826587.stm, reported March 6, 2003, downloaded June 1, 2004

Box 1.2 Survey Highlights Computer Rage

One in five Scots suffers from “Internet rage” and some feel like hurling their computers through a window, according to a survey undertaken in February 2004. Around 1000 Scots were asked to tick a box with several options about their pet hates in everyday life for the survey this month. Some 45% of those polled blamed sluggish Internet connections for making their blood boil. This was more than twice the number of people (20%) who said watching their favourite soccer team get beaten drove them mad. . . . One in 10 surfers con- fessed they sometimes felt like punching their keyboard, whacking the monitor with a hammer and even throwing their PCs out the window. A third of people quizzed said they had to walk away to cool down. Additionally, one fifth of Scots feel that slow Internet connections at work make them lose up to an hour a day.

From Jude Sheerin, PA News, as reported at news.scotland.com, http://news.scotsman.com/latest.cfm?id=2522311, February 12, 2004.

Despite more than two decades of HCI research, it remains an unfortunate fact that many computer systems do not do what users want them to do. Users often describe their difficulties in system use as “computer problems,” which is nonspecific as to the source of the problems. There could be several explanations. For example, the prob- lems could be related to buggy software or to the use of older, less efficient hardware or technology that slows the processing of information. Or maybe there was no clear understanding about the work environments in which the new computer systems were expected to operate. Box 1.3 looks at problems that occurred when the UK Passport Agency introduced a new computer system.

Equally, a poorly designed user interface could have contributed to the problems. While there is no direct evidence in any of the news reports to suggest that poor user

interface design was to blame, it is likely that user interface problems contributed to the difficulties that users had with these systems.

EXERCISE 1.1 (Allow 10 minutes)

Think for a moment about the situation outlined in Box 1.3. Suggest what the consequences of the Passport Agency’s computer problems may have been for the following groups of people:

• The general public • The workers at the Passport Agency

DISCUSSION

The consequences to the general public were enormous. People had great diffi- culty getting their passports in time for their vacations. People waiting for pass- ports may have suffered from stress and anxiety related to the possibility that they may not be able to go away. Some may have lost money as a result of being unable to travel. Business travelers would have been affected too. Both groups probably felt anger and frustration with what they would have perceived as com- puters being in control and staffed by unorganized, incompetent government administrators. Furthermore, many people had to take time off work to line up for hours at passport offices. The passport agency workers, too, suffered conse- quences. They would have been under great stress to meet the demands of the public, and they would have felt anger and frustration because the system did not enable them to do their jobs of processing passport applications and issuing passports.

� Loss of Productivity, Efficiency, and Money

Computer systems with poor user interfaces can have a financial cost. Take the crisis at the Passport Agency. It was reported that the cost of this fiasco was $20 million

3. The Importance of Good User Interface Design

9 Part 1

Box 1.3 Passport Agency Delays

In May 1999 the UK Passport Agency hit the headlines as the waiting time for a passport applied for by post lengthened from a target time of two weeks to between seven and ten weeks. While this increase in waiting times was partly due to a larger than expected increase in passport applications (because of a new requirement for children to have their own passports), a second reason cited was “computer problems” caused by the introduction of new computer systems at two of the Passport Agency’s offices. Applicants anxious to ensure that they would have their passports in time to holiday abroad queued by the hundreds in Glasgow, Liverpool, and London.

Drawn from the BBC News web site, June 15, June 24, June 28, and June 29, 1999.

(£12.6 million), which included nine million dollars (six million pounds) on staff overtime and at least $242,000 (£161,000) in compensation to the hundreds of people who missed their vacations as a result of not receiving their passports on time. The Passport Agency also spent $24,000 (£16,000) on umbrellas for people who had to wait outside of passport offices in the rain to get their passports over the counter. Subsequently the price of a passport was increased. The supplier of the computer system had agreed to pay $3.7 million (£2.45 million) of the costs, leaving the remain- der to be paid by the taxpayer. In these days of payment for productivity and effi- ciency, wages may have been lost if agency workers’ earnings were linked to a level of productivity they were unable to meet because the computer system was unfit for its purpose.

3.3 Safety and the User Interface

So far we have considered the problems of poor user interfaces in terms of user frus- tration and dissatisfaction, and the loss of productivity and efficiency to business. There is another important aspect to consider: the issue of safety, both for computer systems users and the general public.

Systems in which human or environmental safety is of paramount concern are referred to as safety-critical systems. These systems include aircraft, aircraft flight decks, air traffic control consoles, nuclear power plants, control systems, and medical devices. The Three Mile Island nuclear power plant disaster (see Figure 1.4 and Box 1.4) illustrated that safety can be severely compromised by poor user interface design, with potentially serious consequences.

CHAPTER 1 | Introduction

10 Part 1

Box 1.4 The Three Mile Island Nuclear Power Plant Disaster

One of the most discussed issues during the early 1980s was the Three Mile Island nuclear power plant disaster. The incident nearly resulted in a meltdown of the nuclear reactor. The cause of the incident was never conclusively deter- mined, but experts, official bodies, and the media all blamed a combination of operator error and bad interface design. In particular, much media attention and several official reports focused on the design of the control panels in the process plant. The incident could have been prevented if the control panels had been designed to provide the operators with the necessary information to enable them to perform their tasks efficiently and correctly. The following are just some of the interface problems that were identified:

• A light indicated that a valve had been closed when in fact it had not. • The light indicator was obscured by a caution tag attached to another valve

controller.

3. The Importance of Good User Interface Design

11 Part 1

Figure 1.4 The Three Mile Island nuclear power plant.

• The control room alarm system provided audible and visual indication for more than 1500 alarm conditions. Evidently this number of alarms was intended to facilitate control of the entire plant during normal operating conditions. However, the layout and grouping of controls on the control panel had not been well thought out and so enhanced, rather than mini- mized, operator error (Brookes, 1982; cited in Leveson, 1995).

• A single “acknowledge” button silenced all the alarms at the same time, but it was not used because the operators knew they would lose information if they silenced some of the alarms. There was simply no way for the operators to cancel the less important signals so that they could attend to the impor- tant ones.

The root of the problem, therefore, seemed to be that the control panels did not support the task of serious error and incident recovery. The control panels misinformed the operators. They did not indicate to the operators the true state of affairs in the reactor plant, and they did not provide the necessary informa- tion in a form that the operators could understand and use to rectify the situation.

3.4 Elections and the User Interface

In November 2000, the topic of user interface design suddenly became international news when the outcome of the U.S. presidential election hung on the results of one county in Florida. In this elec- tion, an apparently minor aspect of equipment design turned out to have major con- sequences. Many voters in Palm Beach County felt that their vote went to the wrong person (see Box 1.5).

State law in Florida limited the time in the voting booth (Figure 1.5) to five minutes. The design of the ballot (Figure 1.6) was considered by some to be difficult to understand. After the election, some voters said that they wanted to vote

CHAPTER 1 | Introduction

12 Part 1

Figure 1.5 The type of ballot booth used in Palm Beach County. © Steve Krug 2004, used with permission.

Figure 1.6 The problematic page of the ballot in the booth. © Steve Krug 2004, used with permission.

for Gore (which required them to punch the third hole) but they mistakenly punched the second hole down because Gore was listed second on the left.

� Small Irritations Are Also a Problem

If you have found it difficult to relate to the “catastrophic” examples we have dis- cussed so far, there are many less disastrous but still problematic examples that may be more familiar to you. Take, for instance, the process of shutting down your com- puter from Microsoft Windows. To do this, you have to press the Start button on the task bar, and find the command Shut Down on the menu. Intuitively, is that where you would expect the Shut Down command to be? What other domestic appliance, or any other type of device, is stopped by starting it? Although the Start button may not have been the obvious place to look for the Shut Down command when you first used Windows, once you have used Windows for some time you adapt to what it makes you do to shut down your computer and you just do it.

EXERCISE 1.2 (Allow five minutes)

Think about your use of the different software applications provided in the Microsoft Office Suite or in another suite of office applications that you use (e.g., StarOffice from Sun). Choose one application, and think about a particular feature that you find confusing when you use it.

DISCUSSION

Debbie writes: The application I use most often from the Microsoft Office Suite is Word. For me, a confusing feature in more recent versions of Word is tabbed

3. The Importance of Good User Interface Design

13 Part 1

Box 1.5 A Palm Beach Voter Comments on the Disputed Ballot

Tuesday at the polls, the ballot information was very confusing if one was voting for the Democrat. It was hard to know whom I was voting for, the way the ballot was printed. I did not know whether I was voting for my choice, Al Gore, or for Pat Buchanan.

That was very scary and upsetting. I had to take the ballot out a couple of times and place it back again to be sure that the arrows pointed to the right hole; even after the third try, I was not sure whom I was voting for, and that makes me very mad. Many other citizens have complained regarding this situation. I am sure this was extremely confusing for senior citizens especially.

Delia Pinto-Houbrick

From the Palm Beach Post, letters to the editor, November 10, 2000, www.palmbeachpost.com, visited July 8, 2003.

dialogs. Specifying document settings is done via the Options tabbed dialog box, which is accessed from the Tools menu.

Before my work computer was upgraded, I had been using Word 97 for a number of years. In Word 97, there are only two rows of tabs for setting document options. As there are more tabs in one row than in the other, it is easier to see and under- stand how the tabs move when a tab is clicked on.

With greater word processing functionality, the Options dialog box became more complex in Word 2002; there are eleven tabs in the Options dialog box arranged in three rows of tabs. Clicking on any tab causes a puzzling rearrangement of the tabs. In fact, each row of tabs moves as a whole; only the row positions are changed rather than the positions of the tabs themselves, so there is some reason to it (see Figure 1.7).

CHAPTER 1 | Introduction

14 Part 1

The “Options” dialogue box from the “Tools” menu in Word 97

The “Options” dialogue box from the “Tools’ menu in Word 2002

Choosing the “File Locations” tab only makes the two rows of tabs swap positions. The ordering of the tabs from left to right does not change.

Choosing the “Spelling & Grammar” tab makes the three rows of tabs rotate their positions. When this is done in real time the tabs seem to scramble, and it is hard to see that the ordering of the tabs from left to right does not actually change.

Thus it is far easier to see and follow the movement of the tabs in Word ‘97 than in Word 2002, the more current version.

Options

Track Changes User Information Compatibility File Locations

Spelling & GrammarSavePrintGeneral EditView

Show

Draft font Bookmarks

Normal view options

File types: Location:

User InformationTrack changes Compatibility File Locations

Spelling & GrammarSavePrintGeneral EditView

?

Options

Track Changes

User Information Compatibility File Locations

Spelling & GrammarSecurity

SavePrintGeneral EditView

?

Documents C:\Debs_tmp\Dks\files\WORD_FLZ Clipart pictures C:\__\Microsoft_Office\Clipart

Show

Startup Task Pane Smart tags Windows in Taskbar Options

User Information Compatibility File Locations

SavePrintGeneral EditView

?

Spelling

Check spelling as you type

Track ChangesSpelling & GrammarSecurity

Figure 1.7 A comparison of tabbed dialog boxes for Word 97 / Word 2002.

4 Designing for Users With the more widespread use of computers, the knowledge, skills, and experience of computer users have become very broad. A good user interface caters to end users and supports them in the tasks they wish to undertake. A computer system that is developed without a good knowledge of the users and what they want to do with the system may be usable in that it can be used to do something, but it may not do what the users want to do in order to achieve their goals. The system will be usable, but not necessarily useful. This is not to say that all computer systems have to be designed to accommodate everyone. Computer systems should be designed for the needs and capabilities of the users for whom they are intended. Ultimately, a user should not have to think unnecessarily about the intricacies of how to use a computer unless, of course, that itself is the user’s task.

4.1 User-Centered Design

User-centered design (UCD) is an approach to user interface design and develop- ment that involves users throughout the design and development process. User- centered design not only focuses on understanding the users of a computer system under development but also requires an understanding of the tasks that users will perform with the system and of the environment (organizational, social, and phys- ical) in which they will use the system. Taking a user-centered design approach should optimize a computer system’s usability.

Earlier we provided the ISO 9241:11 definition of usability. ISO 13407, Human- Centered Design Processes for Interactive Systems (ISO, 1997), provides guidance on and lists the main principles and essential activities for human (user)-centered design, for achieving usability in systems. Briefly, the four main principles of human- centered design are (ISO, 1997 p. 7):

1. The active involvement of users 2. An appropriate allocation of function between user and system 3. The iteration of design solutions 4. Multidisciplinary design teams

The four essential human-centered design activities are (ISO, 1997 p. 10):

1. Understand and specify the context of use 2. Specify the user and organizational requirements 3. Produce design solutions (prototypes) 4. Evaluate designs with users against requirements

Adopting the approach prescribed by ISO 13407 ensures that the users’ perspectives form part of the HCI design and development process, which will positively influ- ence the usability of the final product.

4. Designing for Users

15 Part 1

4.2 The Classic Life Cycle

User-centered design and traditional software engineering take very different approaches to computer system design. Traditionally, software developers have treated each phase of the software design life cycle as an independent part of soft- ware development, which must be completely satisfied before moving on to the next phase. This is particularly so in relation to the classic life cycle (also known as the waterfall model, so named because of the cascade from one phase to another; see Figure 1.8). It prescribes a predominantly sequential transition between the succes- sive software life cycle phases, where each phase is completely satisfied before the next begins (this is represented in Figure 1.8 by the red arrows).

This view is, of course, simplistic. Software engineers readily accept that although the design is guided and regulated by this top-down somewhat linear model, in practice there are many iterations up and down between stages. Sommerville (1992), for example, has the following to say on the matter:

In practice, however, the development stages overlap and feed information to each other. During design, problems with requirements are identified; during coding, design problems are found; and so on. The software process is not a simple linear model but involves a sequence of iterations of the development activities. (p. 7)

Therefore, within the software design life cycle there is a need for the phases to feed information to each other, and for iteration, rather than the development proceed- ing from start to finish in a simple linear fashion. This iteration is represented in Figure 1.8 by the blue arrows. The essential difference between the classic life cycle

CHAPTER 1 | Introduction

16 Part 1

System and software design

Requirements definition

Operation and maintenance

Integration and system testing

Implementation and unit testing

Figure 1.8 The classic life cycle. (From Sommerville, 1995.)

and user-centered interface design is that user interface design and development is based on the premise that users should be involved throughout the design life cycle. Additionally, the process should be highly iterative, so that the design can be tested (or evaluated) with users to make sure it meets the users’ requirements. Unlike this iterative design process, the waterfall life cycle generally leaves evaluation to the end. Let us look at these aspects further. Figure 1.9 illustrates the iterative user interface design and development process.

4.3 Involving Users

The way to be user-centered is to involve users and to pay attention to their views. This can include a variety of approaches, from simply observing users’ working prac- tices as part of collecting system requirements, to using psychologically based user- modeling techniques, to including user representatives on the design team. More important, users should be involved in the testing and evaluation of the system during its design and development. But who are the users?

� Who Are the Users?

In a user interface design project, developers generally refer to several types of people as users:

• Customers, who pay for and perhaps specify the computer system under development

• Other people within the users’ organizations who have an interest in the development of the system

• Users or end users — the people who actually use the system directly to undertake tasks and achieve goals

4. Designing for Users

17 Part 1

The user interface design and development

process

Design

User testing and evaluation

Prototyping

Figure 1.9 The iterative user interface design and evaluation process. (From Greenberg, 1996.)

To avoid confusion, from this point on we will reserve the term “users” for the end users of a computer system.

The difficulty in having so many different people involved is that everyone works from his or her own perspective and has a particular agenda. For example, managers will want to ensure that their organization’s effectiveness is not impaired by the intro- duction of a new system. They will want to be sure that the final solution is cost- effective, does not threaten efficiency, and is safe and reliable. Professional bodies or trade unions will want to ensure that working agreements are not broken and that workers’ rights in relation to employment and IT-related health and safety issues are protected.

Additionally, there is the design team, which potentially includes, as well as users, design team managers, marketing personnel, software engineers and designers, pro- grammers, graphic designers, systems analysts, HCI specialists, and other research and development staff. Again, each works from his or her own perspective. A soft- ware engineer, for example, may provide advice about optimal configurations of computer system architecture and the performance of the software. A graphic designer will focus on the visual appeal of the UI and may wish to establish whether graphics are appropriate to all application areas or whether there are any potential problems in selecting iconic representations for certain situations. Marketing people want to sell products on time and as cost-effectively as possible and want the right image for the application. Users need to feel confident that the computer system will offer the right facilities to ensure that their work is carried out at least as effectively as with previous methods or computer systems.

It can be difficult to reconcile all the views, and there will be constraints and trade- offs related to decisions made for the final design.

EXERCISE 1.3 (Allow 10 minutes)

Think about a recent project you have been involved with at work or an impor- tant meeting that you have attended (for example, a meeting at a club you belong to or a local community meeting). Various people were there, all expressing their own views. In the project or at the meeting, was it easy/necessary/appropriate to accommodate all the views expressed when arriving at a decision?

DISCUSSION

Many people have been involved in the development of this book. The follow- ing is a partial list:

• The academic members of the editorial team • The authors and others involved in the Open University course that this book

has been drawn from • Academic and professional reviewers • Our publisher and its editing staff

You will learn more about constraints and trade-offs in Part 2.

You will learn more about persuading others to make changes in Chapters 28 and 29.

CHAPTER 1 | Introduction

18 Part 1

It has sometimes been challenging to work as a team as we are scattered across different time zones and some people have moved on to other work since they made their original contributions.

4.4 Making the Design Process Iterative

Making the user interface design and development process iterative is a way of ensur- ing that users can get involved in design and that different kinds of knowledge and expertise related to the computer system can be brought into play as needed. We shall use an approach adapted from a model proposed by Hix and Hartson (1993). It is known as the star life cycle for obvious reasons, as you can see from its appearance in Figure 1.10.

The star life cycle encourages iteration. First, you will notice that the central point of the star is evaluation, which is viewed as being relevant at all stages in the life cycle and not just at the end of product development as the classic life cycle tends to suggest. Evaluation is concerned with gathering data about the usability of a design or product by a specified group of users for a particular activity within a specified environment or work context. Not surprisingly, a host of different evaluation tech- niques are needed to support the different stages of design and to reflect the design needs of different kinds of products. These include interviews with users and others, observing users in their workplace, and getting users’ opinions from questionnaires or other types of surveys. Second, the star life cycle is “intended to be equally sup- portive of both top-down and bottom-up development, plus inside-out and outside- in development” (Hix and Hartson, 1993, p. 101). Thus, a design can start with any process in the star life cycle.

4. Designing for Users

19 Part 1

Implementation

Prototyping Requirements specification

Task analysis/ functional analysis

Evaluation

Conceptual design/ formal design

Figure 1.10 The star life cycle. (From Hix and Hartson, 1993.)

You will learn more about evaluation in Part 4.

� When and How to Involve Users

Users should be involved in every part of the user interface design and development life cycle.

• Early in the design process when the requirements are being specified. Users could help in defining the requirements for the system by contributing a specification or by testing early mockups. Users can get involved by allowing themselves to be observed and giving feedback about the problems of the current system.

• During prototyping, to test designs and options. Users could test versions of the interface to provide feedback and make suggestions to the designers.

• Just before delivery of the product. Users again could test the product by using it or completing surveys about the various features. At this point, however, only minimal changes would be allowable.

• During training/after delivery of the system. Again, users would use the product and give their opinions and detail any problems. Revisions at this stage would be included in the next version.

EXERCISE 1.4 (Allow 10 minutes)

Think about the approach you take to software development at work. If this is not applicable to you, think about how you might approach a task like prepar- ing a meal for friends. Think about where you would involve users and where you might work iteratively (that is, where you might repeat steps).

DISCUSSION

If I were to prepare a meal for friends, I would certainly involve them by first asking when they could come — there’s no point in cooking a great meal and then asking them whether they are available! I would also consult them in advance over particular food likes and dislikes or perhaps what they would want on the menu. In cooking, there are dozens of things you do repeatedly (or itera- tively): lots of tasting goes on, and we tinker with ingredients until the taste is how we want it. My grandmother owned a restaurant and my father was a chef, so when I make, for example, a sauce for spaghetti or lasagne, it’s an all-day affair. I start with tomatoes and vegetable stock, I add some spices, and then let it simmer (stirring occasionally) before tasting it. Usually it will need more salt or more stock, and once this is added I let it simmer a while longer. Somewhere along the way I add vegetables and meat (if there are no vegetarians dining with us). The sauce is usually just right about five hours later, after lots of iterations to the cooking sauce! Essentially, my approach is no different to that in Figure 1.10 — I design the sauce from my recipe, I test prototypes of the sauce as it simmers, and then I evaluate how good (or not!) it is.

CHAPTER 1 | Introduction

20 Part 1

5 The Two Types of Knowledge Needed for UI Design In addition to the need for knowing about the intended users of a system and for designing iteratively, there is other information you need to know about and take into account when designing a user interface. This information comes from two sources:

• Information-gathering activities and analyses that form part of the user interface design and development process

• User interface design knowledge — for example, design principles and design rules.

6 Evaluation This book is text based. The limitation of it being a static text-based volume of work is that the knowledge we are trying to impart can only be presented sequentially: chapter by chapter, topic by topic, in a given order. However, while you are studying this book, it is important to keep in mind the star model and to remember that there is no one single-user interface design approach or life cycle: it can start with any of the processes detailed in the star life cycle, and it should be highly iterative. Iterative design provides the opportunity for evaluation and user testing, the reviewing of the results of user testing, and the consideration of how that feeds into the cycle. In addition, it allows you to consider what design alternatives would improve the user interface design.

Don’t be intimidated by the thought of doing evaluation. For most of us, evaluation is something we actually already do, but without formally saying, “I’ll just do an eval- uation here.” We undertake evaluation every day, in many ways. For example, we shop, we buy different foods, we eat them, and then we decide whether they are good enough to buy again. If you drive a car, then you may be thinking about things like how many miles per gallon you are getting from your car or how well it is running. And if your car is running badly, then you may be skeptical about how well the mechanic has done his job — you are evaluating the mechanic’s skill in servicing your car in relation to how well the car is performing. When we buy consumer goods, we may go into a retail outlet and play with (or evaluate) several different manufacturer’s models of the item we are interested in purchasing; we press buttons, turn knobs or dials, and generally try to decide which model seems to do best what we want to purchase the item for.

EXERCISE 1.5 (Allow 10 minutes)

Think about a recent purchase you have made. It might be a new PC, a new car, or a domestic appliance for the home; or perhaps you were involved in making the decision about buying some equipment at work. Think about how you made the decision. Did you buy the first item you saw, or did you look for specific features to judge alternatives by and to guide your choice?

6. Evaluation

21 Part 1

You will look at the information- gathering activities in Part 2 and the UI design knowledge in Part 3.

DISCUSSION

There is no right or wrong answer here, as decisions in relation to purchases are often personal to the individual and specific to the item being purchased. For example, I recently bought a microwave oven. I was looking for an oven without a turntable. This feature was important to me, as my previous model had a rotat- ing microwave antenna under the floor of the oven (rather than an internal turntable). I had become used to being able to put any size or shape of cookware into my now defunct 15-plus-year-old model without worrying about cookware getting wedged, as often happens in ovens with turntables. Although there were other features to consider, like cost, color, design style, power levels, cooking pro- grams, and the type of controls (dials and knobs versus touchpad and digital display), these were less important to me personally. In the end, as I was unable to find a microwave oven without a turntable, I evaluated the dozen or so models available on the other features before deciding which model to purchase.

Although I didn’t find the microwave oven I wanted, in the end I made an okay choice: the oven looks good, it is only one-third the size of my old oven (which freed up some worktop space), and it was reasonably priced. However, the biggest flaw in my decision making was that I did not (or rather could not) assess (evaluate) how usable it was. It is a limitation that you cannot test certain domes- tic appliances in showrooms before you buy them. Now I own a microwave where I need to have the manual in front of me to set a cooking program, as the sequence of button presses is complicated and there are no reminders anywhere on the interface to tell me what the steps are (i.e., the visibility is poor).

At one meeting we discussed how each of us decided which washing machine to buy. One person was clear that he would only buy from a particular manufacturer. Another person’s concern was style, another person’s was spin rpm. Not one person mentioned the interface, interface controls, or usability as a prime determiner in their choice. The point here is that choices are made for many reasons, which often have nothing to do with the actual interface. Nevertheless, in designing your inter- faces you should ensure that you follow an iterative user-centered design process both to increase the usability of your interfaces and to make it less of a gamble for the users.

Consciously or unconsciously we “do evaluation” in everyday life. We gather infor- mation, compare alternatives, reflect on outcomes, and make choices based on the information gathered and how well our requirements could be met. Evaluation undertaken for UI design is no different in essence. What differs is when the evalua- tion is done, and how the information (or knowledge) gathered is used to inform the design and development process.

6.1 When and How Do You Evaluate?

Evaluation is a way of finding out whether your systems work, and it is an ongoing activity throughout the life cycle. You can look for as many problems as possible

CHAPTER 1 | Introduction

22 Part 1

(diagnostic evaluation) or you can try to measure the performance of the system (measurement evaluation). The role of the evaluation is always to inform the design and to improve it at all stages. Measurement evaluation frequently contributes to subsequent versions of a product. Different kinds of evaluation may be carried out at different design stages and for different reasons.

� Evaluation Early in the Life Cycle

Evaluation during the early design stages is undertaken to validate the users’ require- ments, predict the usability of the product or the usability of an aspect of the product, and to assess how well the interface meets the users’ needs.

The earliest user evaluations may be best done using paper-based prototypes and mockups. These are very low cost but can yield a great amount of information and feedback about a design. Findings from early evaluations can be taken on board and fed back into the design process before the design is set.

� Evaluation Later in the Life Cycle

Evaluation later in the design cycle is also undertaken to assess how well the user interface meets the users’ needs. It is carried out to check the usability of the nearly completed system and to ensure that the system meets the specified usability requirements. At this point, findings are unlikely to be fed into the UI design and development process, as generally by this point the system is more or less ready. These findings might be used for the next version or release of a system rather than for changing the near-finished product.

� How Do You Evaluate?

Choosing what to do will depend not only on the questions that you want to answer but also on logistical factors, such as the time available to do the evaluation, the avail- ability of suitable expertise and equipment, access to users, and so on. Often the choice comes down to money: How much will it cost and how will we benefit from doing it?

There are many different techniques for evaluation. Here are just a few to get you started. The first two will be discussed further in Chapter 2, while the process of evaluation will be covered in depth in Part 4.

• Observing the organization and how people work. Several kinds of evaluation depend on some form of observation or monitoring of the way in which users interact with a product or prototype. The observation may take place informally in the field or in a laboratory as part of more formal usability testing.

• Interviewing, talking, and asking questions. As well as examining users’ performance, it is important to find out what they think about using the technology. No matter how good users’ performance scores are when using technology, if for some reason they do not like using it, then it will not be used. Surveys using questionnaires and interviews provide ways of collecting users’ attitudes to the system.

6. Evaluation

23 Part 1

• Making predictions. The aim of this kind of evaluation is to predict the types of problems that users will encounter without actually testing the system with them.

As a general rule, any kind of user testing is better than none. You will learn some- thing valuable from even the most informal evaluations — watching a single user interacting with a UI under design will provide interesting and informative feedback about the design.

7 Summary In this chapter, we introduced you to user interface design, discussed why it is impor- tant, and explored what can happen when UIs are badly designed. We then intro- duced you to user-centered design, emphasized the importance of user involvement throughout design and development, and stressed the need for an iterative approach to design and frequent evaluations of the design in progress. We then discussed when to evaluate user interface designs and provided a short introduction to evaluation techniques.

CHAPTER 1 | Introduction

24 Part 1

252 Requirements

1 Overview Good user interface (UI) design involves understanding, so far as possible, the requirements for the system under development. Whether you are redesigning an existing computer-based system or designing a system that will computerize tasks currently being performed manually, you will find it a lot easier if you gather the nec- essary information in order to gain an understanding of the requirements.

Part 2 of this book concentrates on requirements. We will consider the following questions:

• What area of expertise, or domain, will the application be developed for? • Who are the users? • What do they want to do with the system? • Where will it be used?

We then help you to analyze and make sense of the information for your UI design.

2 What You Will Find in the Chapters Chapter 2 describes some techniques to use for finding out the requirements for the system. We introduce you to several techniques — observation, interviews, and ques- tionnaires/surveys — that you can use in your requirements-gathering activities. Chapter 3 gets more specific. We detail the investigative activities for finding out about the the users of the system and the domain for the system. Chapter 4 explores the work or other tasks the users will perform with the system and the environments within which they will work.

In Chapter 5, we ask you to think about the information you might have gathered and how to analyze it. Analysis involves looking at the information gathered and decid- ing how it can inform the design of your UI. This will prepare you for the discussion of conceptual design that takes place in Part 3.

Chapter 6 describes creating usability requirements, the section of the overall requirements that specifically relates to the usability of the interface. We discuss con- straints and trade-offs in creating requirements, and we describe prototyping and consider how it can be used for requirements gathering and for working toward an effective design with users and stakeholders. Prototyping is an important part of the iterative, user-centered design life cycle. You will meet it again throughout the course as you learn about designing the UI and about user evaluation of UI designs under development.

Finally, in Chapter 7, we take a break from theoretical material to look at a practical example. The first part of our case study illustrates how requirements were gathered in practice by Tokairo, UK, for the design and development of a system to collect worksheet information from truck drivers distributing oil-based products.

PART 2 | Requirements

26 Part 2

3 Learning Outcomes After studying Part 2, you will be able to:

• Employ several techniques that can be used to gather the requirements for the design of a UI

• Describe the activities involved in gathering the requirements for the design of a UI

• Understand the role of prototyping in requirements gathering

4 Theoretical Influences The chapters on requirements-gathering techniques, domain, users, tasks, and mental models draw from the fields of psychology and cognitive psychology. The section on environments draws from psychology and social/organizational psychology. The chapter on prototyping draws from computer science/software engineering.

4. Theoretical Influences

27 Part 2

2 How to gather requirements: some techniques to use

1 Introduction There are many techniques you can use to find out about the application domain, the users, their tasks, and the environment within which the system will be used. These techniques include observing users, interviewing users, and obtaining infor- mation from users via questionnaires or surveys. Our discussion here focuses on the use of these techniques for requirements gathering, but you will meet these tech- niques again later, when we discuss how they can be employed to evaluate user inter- faces (UIs).

2 Observing Your Users Going to observe users in their natural setting — observing them while they are doing real work in their real working environment or using a home system in their homes — is an essential part of user-centered design. In addition to finding out what users do, you can also discover what aspects of the current system they like and dislike. Observation of users in their workplace or home can be either direct or indirect.

2.1 Direct Observation

Direct observation is a straightforward activity and will rapidly provide you with an insight into the users, their tasks, and the environment for a computer system. Direct observation can be undertaken in many ways, but generally direct observation studies are classified as either field studies or controlled studies. Field studies directly observe users in their usual work or home environment, doing their normal work or home tasks, with the observer making notes about interesting behaviors. Controlled studies directly observe users in a place other than their normal environment (for example, in a usability laboratory), performing specially devised tasks, with the observer recording their performance in some way, such as by timing tasks or par- ticular sequences of actions.

29 Part 2

You will learn more about field studies and controlled studies in Part 4.

Direct observation is always worth doing, as it is an easy activity to undertake and always yields interesting data, but it does have some limitations. For example, it only allows a single pass at the information gathering (or data collection), and although the observer may take notes, it is hard to get a full record of user activity in one obser- vation session. The observer has to make decisions about what is important to record, and there is no opportunity to review that decision and look at alternative data later on. Furthermore, direct observation is considered to be obtrusive and can alter a user’s behavior and performance.

EXERCISE 2.1 (Allow 10 minutes)

Suppose you have been asked to go into a school where a prototype of a new multimedia teaching application is being tried out by groups of 10-year-olds for the first time. The developers have asked you not to interfere with the children’s activities, but to note the kinds of things they do and what difficulties they encounter. What problems might you experience using direct observation?

DISCUSSION

A lot will be going on: children talking at once, getting excited, changing groups, and maybe taking turns using the keyboard. Some children may not have lis- tened to the instructions and may be more interested in disrupting the activities of others than in joining in the lesson. You will not be able to write down every- thing you see or hear. You will have to decide what is important and focus on that, which may mean that you miss some interesting interactions. You may also get distracted yourself and thus miss things. When you try to make sense of your notes later, you may not understand your own cryptic comments or you may not be able to read your own writing. However, despite these problems, having gone into the school you will undoubtedly have a better idea of how the teaching application can be used. So even this kind of observation is better than none at all.

Direct observation is useful early in the life cycle as part of gathering the require- ments for a computer system. If you want a permanent record of your observations, then some sort of recording equipment (such as video, audio, or interaction logging) should be used.

2.2 Indirect Observation: Video Recording

Video recording on its own is an alternative to direct observation as it provides a per- manent record of the observation session. Sometimes video recording may be used with software that automatically logs users’ keystrokes or interactions. Although col- lecting several kinds of information can be beneficial, it means that there is more data to analyze, which can be time consuming. Because indirect observation creates more distance between observers and users, it is considered to be more objective than direct observation. Although specially mounted recording equipment (the facil-

CHAPTER 2 | How to gather requirements: some techniques to use

30 Part 2

ities typically found in a usability lab, for example) is extremely useful, you may be surprised by just how much valuable data you can collect using ordinary consumer video equipment, especially now that small digital video recorders are available at a reasonable price. However, there are some important issues to consider. You need to plan the observation, which means thinking about what you want to find out and what kind of data you need. For example, in a study of the way that people use a UI in the context of their own workplace, it may be useful to record samples of them using the UI every day over a period of several days or weeks. These interaction samples could then be analyzed; categorizing the activities, for example, will tell you what the UI is used for, what work it helps the users to do, or how often a particular task is done. A study with quite a different and much finer focus might involve an in- depth examination of two users interacting with the UI over a period of just five minutes.

There can be practical problems associated with setting up video equipment. For instance, no matter how unobtrusive you try to be, users are likely to be aware that they are being filmed. One way of reducing the impact that the equipment has on their behavior is to leave it in place for several days before recording starts, so that the users grow accustomed to it. You will also need to decide how and when you will start and stop your recording, how you will label the recording so that you can catalog it, who will change the cassette, where the equipment will be physically located, and so on.

2.3 Points to Consider in Relation to Observation

Both direct and indirect observation will require you to make trade-offs. If you record data using video or logging software, then you can go back and look at it later. However, you may end up with an overwhelming amount of data to analyze, which can be a problem unless you have a clear idea of what you are looking for and what you want to find out. It takes many times longer to fully analyze video than it does to film it in the first place. If you record those things of interest by hand, however, your recording will probably be incomplete because you will miss things. You will thus have a less complete picture to review later.

Direct observation is the cheapest and most straightforward way of recording observations. Automatic indirect recording provides a permanent record that you can return to later and as often as necessary. The two techniques are not mutually exclusive, since you may use direct observation to initially plan your automatic recording.

EXERCISE 2.2 (Allow 10 minutes)

Figure 2.1 shows a machine for purchasing tickets to travel on the Prague under- ground. It is a standard UI, much like the machines you might use to purchase travel tickets for train journeys in other countries. But do you notice anything unusual?

2. Observing Your Users

31 Part 2

DISCUSSION

As noted previously, the UI shown in Figure 2.1 is pretty much what you would expect of a machine that lets you purchase travel tickets. However, did you notice the areas on the right-hand sides of the machines, near the coin slots, where the paint had been scraped off? This wear on the machines was caused by users who thought that the coins would drop more effectively if they were rubbed against the machine before being pushed into the slot. The reasons for this par- ticular wear pattern on the machines would have remained unknown if someone had not gone out and observed the ticket machine being used by real commuters.

What may not be immediately apparent is how this informs the design of the UI. Basically it implies that the finish on these machines needs to be more robust. These machines are already vulnerable, as they are situated in an outside envi- ronment and exposed to all types of weather and extremes of temperature. Any damage to their finish, beyond expected wear and tear, will shorten the working life of the machine.

CHAPTER 2 | How to gather requirements: some techniques to use

32 Part 2

Figure 2.1 A Prague ticket machine.

3 Interviewing Your Users Interviewing involves talking to or questioning users. It enables the gathering of information in a fast and friendly way. You will need to plan interviews carefully, deciding in advance who to interview, what questions to ask to gather the relevant information, and how long the interviews should be. There are two main kinds of interview: structured and unstructured (flexible). A structured interview has prede- termined questions that are asked in a set way; there is little, if any, scope for explor- ing additional topics that might arise during the interview. In contrast, a flexible interview generally has some set topics for discussion and exploration, but no set sequence: the interviewer is free to follow up the interviewees’ replies, and to find out more about anything that is said.

A flexible interview is less formal than a structured interview and is very useful early in the design process for gathering requirements and gauging users’ opinions about a particular idea. If you intend to undertake a flexible interview, however, you will find it useful to have a rough plan of the topics you want to cover, particularly if you are inexperienced at interviewing. This rough plan can be either in your head or dis- creetly written on paper and kept out of view. As you gain experience, you will find that interviewing becomes easier. Another factor you will need to consider is how to make the interviewee feel comfortable so that rapport is established between you. This is particularly important if you are trying to gain information that the inter- viewee may feel embarrassed or concerned about telling you. For example, some people feel embarrassed about criticizing a system, particularly if it involves describ- ing their own difficulties in using it. In general, people who lack confidence tend to assume that the mistakes they make are due to their own stupidity rather than to poor design. Alternatively, they may think that their opinions are trivial and of no interest to you, or that what they say is of no importance. If you want to obtain this kind of information, then you will need to create a friendly, unthreatening atmos- phere by being casual yourself while keeping sufficient control to direct and channel the discussion so that you obtain the information you want. This requires practice and experience.

3.1 Points to Consider in Relation to Interviewing

In general, the more structured the interview, the easier it is for the interviewer (Welbank, 1990). The less structured the interview, the more scope there is for picking up relevant issues, but the harder it is for the interviewer. You will need to make a judgment about the right balance to strike. Another issue to consider is how you intend to avoid asking leading questions that provoke a particular response. You will gain a lot from doing a small pilot study: either try out your interview questions and practice your interviewing skills on one or two users who will not take part in the real study, or, if you have too few users, try it out on colleagues. Data analysis is, of course, more difficult with flexible or less structured interviews, but in general such interviews provide much richer information. It is standard practice to record inter-

3. Interviewing Your Users

33 Part 2

views with users; you should seek permission to record, and users rarely object. As with video logging, the advantage of audio recording is that you have a permanent record. Audio recordings of interviews should be transcribed so that you can examine what has been said in detail. Subtle comments can be easily missed if you rely solely on notes taken during the interview, as your notes are likely to be incomplete. A dis- advantage to audio or video recording interviews is that initially the technique may change users’ behavior.

4 Questionnaires and Surveys Questionnaires and surveys take a different approach to interviews for the purpose of gathering information. The focus shifts from the flexible and friendly approach provided by interviewing to the preparation of unambiguous questions and state- ments for the gathering of more precise information.

4.1 Types of Question Structure

Broadly speaking, there are two question structures for questionnaires: closed ques- tions and open questions.

� Closed Questions

A closed question asks the respondent to select an answer from a choice of alterna- tive replies. Closed questions may require just “yes” or “no” responses, or they may have some form of rating scale associated with them. Which type you use will depend on whether you need simple or detailed information, as you will see from the exam- ples below. The simplest rating scales are just checklists consisting of basic alterna- tive responses to a very specific question. For example, a three-point scale that allows respondents to choose “yes,” “no,” or “don’t know” is often used (see Figure 2.2). These questions are easy to analyze because all you need to do is to count the number of responses in each category.

More complex rating scales increase the number of points (or responses) to produce a multipoint rating scale called a semantic differential. The meanings of just the end points are given, as shown in Figure 2.3. Users are asked to select the point along the scale that most closely matches their feelings or opinion about what is being rated.

CHAPTER 2 | How to gather requirements: some techniques to use

34 Part 2

Can you use the following text editing commands?

Don’t knowYes No

COPY

PASTE

Figure 2.2 An example of a simple checklist.

You can explore a variety of views about the system such as whether the users think it is easy or hard to do certain tasks or whether it makes them feel good or bad.

Semantic differentials are often created with seven points, but five-point or even three-point scales can be just as effective and are quicker to analyze. You will get better results from semantic differentials if you make sure that the two adjectives at the end points are definitely opposed and are meaningful for each of the aspects that you are asking the users to rate. A series of specifically designed pairs of adjectives will often give you better results than asking users to rate a variety of aspects of the system from “poor” to “excellent.”

A Likert scale is a selection of statements, each similar to a semantic differential, that when analyzed together portray a user’s attitude. The construction of Likert scales requires statistical analysis that is outside the scope of this book.

Once a semantic differential has been completed by the selected population of users, then you can get a feel for the strength of opinion in the respondents by counting up the number of responses at each point in the scale. Although it is tempting to try to calculate a numeric value by adding up the plus and minus points score and divid- ing by the number of respondents, this can be misleading as some people rarely or never choose the outside values in the scale even though they have strong opinions, while others will choose extreme values to represent milder opinions.

� Open Questions

An open question allows respondents to say whatever they like in response, and they are used where there are no predetermined answers. Open questions typically start with phrases such as “What do you . . . ,” “How do you . . . ,” or “What ways . . . .” Lim- iting the amount of space on the form for the answer can encourage respondents to prioritize their points (Rubin, 1994). Open questions provide richer data than do closed questions, although the responses will be more time consuming to analyze as you you need to read each one and decide on some way of grouping and classifying them. If you have a fairly small sample, say up to 100 respondents, it may be quicker and just as effective to create a simple list of all the responses to each open question.

4.2 Points to Consider When Designing Questionnaires

Potentially, questionnaires can reach a large number of people, so it is important to ensure that they are well designed. A boring questionnaire that asks impertinent or

For more information on questionnaire design and attitude measurement, see Oppenheim (1999), a much easier book to read and use than its initial appearance would suggest.

4. Questionnaires and Surveys

35 Part 2

Rate the usefulness of the COPY command on the following scale.

Very useful Of no use

Figure 2.3 An example of a semantic differential.

complicated questions will get a low response rate and may alienate your users. On the other hand, a carefully designed questionnaire that you have piloted can be a speedy way of getting data from a lot of users. There are several points to keep in mind when designing a questionnaire:

• Make the process easy for the person who is answering by keeping the questions simple, and ask as few questions as possible; unless absolutely necessary, aim for no more than two sides of letter paper or A4 paper (an ISO standard size of paper slightly longer and slightly less wide than letter paper).

• Make sure the questions are clear and unambiguous, as you will not be there to address any difficulties that the people completing the questionnaire may have.

• Make sure the questions will gather the information you need. • Provide oppportunities for your respondents to offer information that you may

not have thought about; for example, you might include some open questions or an “any other comments” box.

As with interviews, it is important to test your questionnaire by doing a pilot study, either with a small sample of users who will not be part of the survey or with some work colleagues.

If you need to survey a large number of users — to find out about their opinions and difficulties in relation to a system being redesigned, say — then closed questions will enable a large amount of information to be collected and analyzed relatively easily. Open questions provide a rich source of data, but making sense of this data requires more time. Generally, effective questionnaires contain a mix of both closed and open questions.

If you think you will need a more complex statistical analysis, then we advise you to consult a statistician while planning your survey. Many statistical packages are avail- able to support data analysis; a good example is the Statistics Package for Social Sciences (SPSS). If you choose to consult a statistician, make sure you do so before designing your questionnaire. Many inexperienced evaluators fall into the trap of col- lecting data and then trying to decide what statistics they should apply after the fact.

5 Summary This chapter explored several investigative techniques: observation, interviews, and questionnaires/surveys. Any or all of these techniques can be used in the requirements-gathering phase of UI design. Complementary investigative tech- niques are often used in combination; for example, you might use a questionnaire and also undertake some interviews, or you might use a questionnaire and also undertake some observation studies. This enables the strengths and weaknesses of the various techniques to be balanced.

CHAPTER 2 | How to gather requirements: some techniques to use

36 Part 2