Usability Engineering. Usability The quality of an interface that measures if it is appropriate for the user or the task for which the system was built

Usability Engineering

Usability

The quality of an interface that measures if it is appropriate for the user or the task for which the system was built.

Usability is not a single, one-dimensional property of a user interface.

Usability has multiple components and is traditionally associated with five usability attributes i.e. Learnability, Efficiency, Memorability, Errors and Satisfaction.

Attributes of Usability

Learnability– The system should be easy to learn, so that the

user can rapidly start getting some work done with the system.

Efficiency– The system should be efficient to use, so that once

the user has learned the system, a high level of productivity is possible.

Attributes of Usability (Cont..)

Memorability– The system should easy to remember, so that the casual user

is able to return to the system after some period of not having use it, without having to learn everything all over again.

Errors– The system should have a low error rate, so that users make

few errors during the use of the system, and so that if they do make errors they can easily recover from them.

Satisfaction– The System should be pleasant to use, so that users are

subjectively satisfied when using it, they like it.

Factors Influencing Interest in Usability

Changes in hardware environments

Diversification of users

Diversification of applications

Usability Engineering Life Cycle

Usability engineering is a set of activities that ideally take place throughout the lifecycle of the product.

Usability cannot be seen in isolation from the broader corporate produce development context.

Usability applies to the development of entire product families. Planning for future versions is also a prime reason to follow up the

release of a product with field studies of its actual use. Poor usability can cause severe damage to the sales of the entire

product family Good usability will contribute to the company’s general reputation

as a quality supplier.

6

The Stages of Usability Engineering Lifecycle Model

Know the user– Individual user characteristics– The user’s current and desired tasks– Functional Analysis– The evolution of the user and the job

Competitive Analysis Setting usability goals

– Financial impact analysis

7

The Stages of Usability Engineering Lifecycle Model (cont..)

Parallel Design Participatory design Coordinated design of the total interface Apply guidelines and heuristic analysis Prototyping Empirical testing Interative design

– Capture design rationale Collect feedback from field use

8

Know the User

To study the intended users and use of the product.

At a minimum, developers should visit a customer site.

Two factors with largest impact on usability– Individual user characteristics and– Variability in tasks

9

Know the users (cont..)

The concept of “user” should be defined to include everybody whose work is affected by the product in some way.

It includes:– The users of the system’s end product or– Output if they never see a single screen

It is often difficulty for developers to get access to users.

10

Know the users (cont..)

Obstacles to access users:– Company to protect its developers from being

known to customers, since customers may bypass established technical support organizations and call developers directly.

– Fearing that the developers may offend the customers

– User organizations only make users available for a short time.

11

Individual User Characteristics

To know the class of people who will be using the system.

User may be more widely scattered. Then to visit only a few, representative customers.

By knowing the user:– Work experience– Education level– Age– Previous computer experience– Learning difficulties– Reading and language skills

12

Individual User Characteristics (cont..)

The user’s work environment and social context also need to be known.

The information needed to characterize individual users may come from:– market analysis or – from the observational studies– Through questionnaires or– interviews

13

Task Analysis

A task analysis is essential as early input to system design.

The users’ over goals should be studied as well as how they currently:– approach the task.– Their information needs and– How they deal with exceptional circumstances or

emergencies

14

Task Analysis (cont..)

The users’ model of the task should also be identified. Observe especially effective users and user strategies

and “workarounds’. Identify the weaknesses of the current situation:

– Points where users fail to achieve goals– Spend excessive time or– Made uncomfortable.

The weaknesses present opportunities for improvements in the new product.

15


The outcome of task analysis is:– A list of all things user want to accomplish with the system (the goals)– All the information they need to achieve these goals (the

preconditions)– The steps that need to be performed and the interdependencies

between these steps.– All the various outcomes and reports that need to be produced.– The criteria used to determine the quality and acceptability of these

results– And finally the communication needs of the users as they exchange

information with others while performing the task or preparing to do so.

16

Interviewing users:– To ask them to show concrete examples of their

work products rather than keeping the discussion on an abstract level.

– It is preferable to supplement such interviews with observations of some users working on real problems.

17


Task Analysis can be decomposed in hierarchical fashion:– Starting with larger task– Goals of the organization– Breaking each of them down into smaller subtasks that can

again be further subdivided An interviewer could ask two questions:

– Why do you do it? (to relate the activity to larger goals)– How do your do it? (to decompose the activity into subtasks

that can be further studied) Other good questions to ask include:

– Why do you not do this in such and such a manner? (mentioning some alternative approach)

18



One should not analyze just the way user currently do the task, but also the underlying functional reason for the task:

– What is it that really needs to be done?– What are merely surface procedures which can, and perhaps

should, be changed.

For example:– Initial observations of people reading printed manuals could

show them frequently turning pages to move through the document.

Task Analysis (Cont..)

– A new design of online documentation might take this observation to imply really good and fast paging or scrolling mechanism.

A functional analysis would that manual users really turn pages this to find specific information, but they have a hard time locating the correct page.

Based on this analysis, one could design an online documentation interface that first allowed users to specify their search needs

Then used an outline of the document to show locations with high search scores

Finally allowed users to jump directly to these location. There is a limit to how drastically one can change the way users

currently approach their task, so the functional analysis should be coordinated with a task analysis.


The evolution of the users:– User will not stay the same.– Using the system changes the users– As they change they will use the system in new ways.

For Example:– Spreadsheets were initially invented as aids for calculation.– Later on its encourage users to integrate non-calculated data

in it. It is impossible to forecast these changes completely

as users will always discover new uses for computer system after some period of use.


Flexible design will stand a better chance of supporting these new uses.

Try to make an educated guess based on your knowledge about how other users have changed in the past.

This is one way of getting such knowledge is through the post-deployment field studies.

A typical change is that users become experts after some time.

User wants interaction shortcuts (accelerators).

Competitive Analysis

It is desirable to analyze existing products according to established usability guidelines and to perform user tests with these products.

A competing product is already fully implemented and can therefore be tested very easily.

The competing product may work fairly well. User testing with existing products can be more realistic than a

test of other prototypes. User can perform real tasks on the competing system, making it

possible to learn how well its functionality And interaction techniques support the kinds of tasks the planned

new product is expected.

Competitive Analysis (cont..)

If several competing products are available for analysis, one can furthermore perform a comparative analysis of their different approaches to the various user interface design issues for the kind of product being studied.

This will provide ideas for the new design This will also give a list of ad hoc guidelines Reviews are important techniques/approaches Reviews should be complemented with more thorough

and principled analysis and testing of a smaller number of important products.

Competitive Analysis (cont..)

Competitive analysis will involve he study of non-computer interfaces. For example:

– An electronic reference book project should first studying how people use traditional printed encyclopedia

Competitive analysis does not imply stealing other people’s copyrighted user interface design.

One would hope to be able to do better than the previous design as a result of the analysis of their strengths and weaknesses.

Setting Usability Goals

Usability is not a one-dimensional attribute of a system.

Usability comprises several components that can sometimes conflict.

Normally, not all usability aspect can be given equal weight in a given design project.

Make priorities clear on the basis of analysis of the users and their tasks.

Setting Usability Goals (cont..)

The different usability parameters can be operationalized and expressed in measurable ways.

Before starting the design of a new interface, it is important to discuss:– Usability metrics of interest to the project and– To specify the goals of the user interface in terms of

measured usability


For example:– The number of user errors per hour is counted.– When using the current system, users make an

average of 4.5 errors per hour, and the planning number of user errors is 2.0 per hour.

– The optimum is obviously to have no errors at all.– If the new interface is measured at anything

between 1.0 and 3.0 user errors per hour, it will be considered on target with respect to this usability goal.


– A performance in the interval of 3-5 would be a danger signal that usability goal was not met, even though the new interface could still be released on a temporary basis since a minimal level of usability had been achieved.

– It is necessary to develop a plan to reduce user errors in future release.

– Finally more than 5.0 users errors per hour would make this particular product sufficiently unusable to make a release unacceptable.


Usability goals are easy to set for new versions of existing system,

The minimum acceptable usability would normally be equal to the current usability level and the target usability could be derived as an improvement.

For completely new system without any competition, usability goals are much harder to set.


One approach is to define a set of sample tasks and as several usability specialists how long it “ought” to take users to perform them.

One can also get an idea of he minimum acceptable level by asking he users.

User are unfortunately are always reluctant in this respect.

Financial Impact Analysis

Financial Impact analysis involves – Estimating the number of users– Who will be using the system– User’s loaded salary and other cost – The approximate time they will be using the system

Financial impact analyses are easiest to make for software that is being developed in house or under contract directly from the user organization.

32

Financial Impact Analysis (Cont..)

Example:– Assume the development of a word processor that is expected to sell

one million copies. About half of the users are expected to be secretaries who will be using the word processor about half of their working day, and the other half of the users are expected to be business professionals who will be using the word processor about 10% of their working day.

– The load cost of a secretary is $20 per hour and the loaded cost of a business professionals is $100 per hours.

– This means that the amount of money spent by users while using the word processor is about$19,000,000,000 annually (Calculated at 8 hours per day and 236 working days per year).

33

Financial Impact Analysis (Cont..)

The financial impact analysis should have tow components:– An estimate of the impact on the development

organization (to help determine the magnitude of the usability budget)

– An estimate of the impact on the user organizations (to help prioritize the focus of the available usability resources)

34

Parallel Design

In Parallel Design Process several different designers work out preliminary designs.

The goal of parallel design is to explore different design alternatives before one settle on a single approach that can then be developed in further detail and subjected to more detailed usability activities.

One can have three or four designers involved in parallel design.

It is normally better to have designers work individually rather than in larger teams.

35

Parallel Design (cont..)

The only aims of Parallel Design are to generate rough drafts of the basic design ideas.

It is specially important to employ parallel design for novel systems

For more traditional systems, where competitive products are available, the competitive analysis can serve as initial parallel design.

It might still be advantageous to have a few designers create additional parallel designs to explore further possibilities.

36

Parallel Design (cont..)

Parallel design is a very cheap way of exploring the design space, exactly because most of the ideas will not need to be implemented.

The main financial benefit of parallel design is its parallel nature, which allows several design approaches to be explored at the same time.

37

Participatory Design

In this approach, the designers should have access to a pool of representative users after the start of the design phase.

It is important to have access to the people who will actually be using the system, and not just their managers or union representatives.

Users often raise questions that the development team has not even dreamed of asking.

38

Participatory Design (cont..)

Users should be involved in the design process through regular meetings between designers and users.

Users are not designers, so it is not reasonable to expect them to come up with design ideas from scratch.

Users are very at reacting to concrete designs they do not like or that will not work in practice.

39

Participatory Design (cont..)

It is necessary to present design ideas as paper mock-ups or simply a few screen designs/prototypes in front of users.

It is important to realize that participatory design should not just consist of asking users what they want.

Users are different, so it is dangerous to rely too much on small set of users.

40

Coordinating the Total Interface

Consistency is one of the most important usability characteristics.

Consistency should apply across the different media which form the total user interface, including not just the application screens but also the documentation, the online help system, and any online videotaped tutorials.

Consistency is not just measured at a single point in time but should apply over successive releases of a product so that new releases are consistent with their predecessors.

41

Coordinating the Total Interface (cont..)

Effort should be made to promote consistency across entire product families.

It is also necessary to maintain some flexibility so that bad design is not forced upon users for the sake of consistency alone.

To achieve consistency of the total interface it is necessary to have some centralized authority for each development project to coordinate the various aspects of the interface.

42

Coordinating the Total Interface (cont..)

Prototyping also helps achieve consistency, sine the prototype is an early statement of the kind of interface toward which the project is aiming.

Consistency can also be increased through technological means such as:– Code sharing– A constraining development environment.

43

Guidelines and Heuristic Evaluation

Guidelines list well-known principles for user interface design which should be followed in the development project.

– General guidelines: applicable to all user interfaces.– Category-specific guidelines: for the kind of system being

developed (e.g. for window-based administrative data processing or for voice interfce accessed through telephone keypads.

– Product specific guidelines: for the individual product.

All these guidelines can be used as background for heuristic evaluation.

44

Guidelines and Heuristic Evaluation (cont..)

The difference between standards and guidelines is that

– A standard specifies how the interface should appear to the users.

– A set of guidelines provides advice about the usability characteristics of the interface.

For example a guideline may state that user should always be able to back out from any undesired system state

One standard might instantiate that general guideline by specifying that an undo command should always be available and it should be shown as an icon at the top of the screen.

45


Another standard might follow the same guideline by returning to the previous system state whenever the user hits the escape key.

Guidelines:– [brown 1988] with 302 guidelines– [Marshal et al. 1987] with 162 guidelines– [Mayhew 1992] with 288 guidelines– [Smith and Mosier 1986] with 944 guidelines

46


Heuristic evaluation is a discount usability engineering method for quick, cheap, and easy evaluation of a user interface design.

Heuristic evaluation is the most popular of the usability inspection methods.

Heuristic evaluation is done as a systematic inspection of a user interface design for usability.

The goal of heuristic evaluation is to find the usability problems in the design so that they can be attended to as part of an iterative design process.

Heuristic evaluation involves having a small set of evaluators examine the interface and judge its compliance with recognized usability principles (the "heuristics").

47

Prototyping

User interface (UI) prototyping is an iterative development technique in which users are actively involved in the mocking-up of the UI for a system. UI prototypes have several purposes:

– As an analysis artifact that enables you to explore the problem space with your stakeholders.

– As a design artifact that enables you to explore the solution space of your system.

– A basis from which to explore the usability of your system. – A vehicle for you to communicate the possible UI design(s) of your

system. – A potential foundation from which to continue developing the system

48

Prototyping (cont..)

Tips and Techniques:– Work with the real users. The best people to get involved in prototyping are

the ones who will actually use the application when it is done. – Get your stakeholders to work with the prototype. Working with the

prototype hands-on, they can quickly determine whether the system will meet their needs. A good approach is to ask them to work through some use case scenarios using the prototype as if it were the real system.

– Understand the underlying business. You need to understand the underlying business before you can develop a prototype that will support it. In other words, you need to base your UI prototype on your requirements. The more you know about the business, the more likely it is you can build a prototype that supports it. Once again, active stakeholder participation is critical to your success.

49


You should only prototype features that you can actually build. If you cannot possibly deliver the functionality, do not prototype it.

You cannot make everything simple. Sometimes your software will be difficult to use because the problem it addresses is inherently difficult. Your goal is to make your user interface as easy as possible to use, not simplistic.

Get an interface expert to help you design it. User interface experts understand how to develop easy-to-use interfaces.

Explain what a prototype is. The biggest complaint developers have about UI prototyping is their users say “That’s great. Install it this afternoon.”

Consistency is critical. Inconsistent user interfaces lead to less usable software, more programming, and greater support and training costs.

50


51


One should not start full-scale implementation efforts based on early user interface design.

Instead, early usability evaluation can be based on prototypes of the final systems that can be developed much faster and much more cheap.

52


Essential user interface:

53


Screen sketches

54


Concrete UI prototype (HTML page)

55

Interface Evaluation

The most basic advice with respect to interface evaluation is simply to do it, and especially to conduct some user testing.

Every body expect best results from testing real users and real systems, but doing so may not always feasible.

The prototyping methods provide means of performing evaluation early enough to influence a project while it can still change direction

The heuristic evaluation method allows you to assess usability without the expense of a user test.

56


The major result of evaluation will be a list of the usability problems in the interface as well as hints for features to support successful user strategies.

Severity ratings are usually gathered by sending a group of usability specialists a list of the usability problems discovered in the interface and asking them to rate the severity of each problem.

Usability specialists are given access to use the system while making their estimates, and sometimes they are asked to judge the problems based only on written description.

57


Severity ratings derived purely by subjective judgment from usability specialists are not very reliable.

Never rely on severity ratings from any single usability specialists.

Collect rating from several independent evaluators. Three to four evaluators, the mean of their ratings is

much better than the ratings from any single one of them.

58


A single rating scale for the severity of usability problems might be:

0 = this is not a usability problem at all

1 = cosmetic problem only --- need not be fixed

unless extra time is available on project.

2 = minor usability problem – fixing this should be given low

priority

3 = major usability problem – important to fix, so should be

given high priority

4 = usability catastrophe ---- imperative to fix this before

product can be released.

59

Iterative Design

Based on the usability problems, one can produce a new version of the interface.

Thinking aloud technique provide sufficient insight into the nature of the problems to suggest specific changes to the interface.

Log files of user interaction sequences often help by showing where the user paused or other wise wasted time and what errors were encountered most frequently.

60

Iterative Design

Solutions need to be designed based solely on knowledge of usability guidelines, and it may be necessary to test several possible solutions before making decisions.

During the iterative design process it may not be feasible to test each successive version with actual users. The iterations can be considered a good way to evaluate design ideas simply by trying them out in a concrete design

61

Follow-up Studies of Installed System

The main objective of usability work after the release of a product to gather usability data for the next version and for new, future product.

A newly released product can be viewed as a prototype of future product.

Follow-up Studies assess how real users use the interface for occurring tasks in their real world working environment.

It will provide much insight that would not be easily available from laboratory studies.

62

Follow-up Studies of Installed System (cont..)

Sometimes, follow-up studies feedback can be gathered as part of standard marketing studies.

The user satisfaction survey confirm the laboratory results.

One may have to conduct specific studies to gather follow-up information about the use of released product. The following methods can be used:– Interviews– Questionnaires– Observational studies

63


Follow-up studies are addressing the usability of an existing systems.

Follow-up studies will also indicate the ability of the software product that how it is being used across a variety of tasks.

In addition to field studies, information can also be gained from the more passive technique of analyzing user complaint, modification requests and calls to help lines

64


Economic data on the impact of the system on the quality and cost of the user’s work product and work life are very important and can be gathered through:– Survey– Supervisors’ opinions– Statistics for absenteeism

These data should be compared with similar data collected before the introduction of the system.

65

Meta Methods

To ensure the successful application of the usability engineering methods, it is important to supplement each of them with the following meta methods (methods that apply to methods):– Write down an explicit plan for what to do when

using the method. How many users to test? What kind of users to test What test tasks these user would be asked to perform And time schedule for the study.

66

Meta Methods (cont..)

– Plan to be reviewed by a person who is not otherwise on usability team and who can critique it from a fresh perspective. This person should preferably be experienced with respect to usability engineering.

– Perform a pilot activity by investing about 10-15% of the total resources budgeted for the use of the method. Then revise your plan for the remaining 85-90% to fix the difficulties that invariably will be found during the pilot activity.

67

Meta Methods (cont..)

As early as possible in the project, an overall usability plan should be established list the usability activities to be performed throughout the lifecycle.

Not all projects can afford to use all the methods, and the exact methods to use will depend on the characteristics of the project.

The meta-methods may involve a little extra work, but they save work in the long term and ensure that your efforts are on the right track to increase usability, thereby reducing the risk of truly wasting the main effort.

68

Prioritizing Usability Activities

It is not always possible to perform all the recommended usability activities in any given project. Nielsen Discount Usability Engineering suggests:– Visit the user sites– Prototyping through scenarios– Simplified thinking aloud– Heuristic evaluation

69

Prioritizing Usability Activities (cont..)

The top six methods as suggested by Nielsen:1-2 Iterative design and task analysis of the

user’s current task

3 Empirical tests with real users

4 Participatory design

5-6 Visit to customer location before start of design and field study to find out how system is actually used after installation.

70

Be Prepared

It is preferable to use usability engineering methods throughout the software lifecycle.

Better results may be expected if the usability specialists have prepared to handle emergency.

The following precautions can be taken during any less hectic periods that may be available between urgent projects:

– Get a good user interface prototyping tool and acquire proficiency in using it.

– Learn appropriate techniques for usability inspection and heuristic evaluation.

71

Be Prepared (cont..)

– Build up an understanding of the types of users, tasks, applications, and computer platforms that are typical for your organization.

– Setup procedures that will allow you to recruit test users easily when they needed.

– Find an train a usability champion in each project group that does not have its own full-time usability specialists

– Read more usability books and articles– Also try out a lot of different systems with different kinds of

interfaces to get experience with alternative interaction styles.

72

Documents

Usability Engineering. Usability The quality of an interface that measures if it is appropriate for the user or the task for which the system was built