lecture 5. design (1) - uvic.ca

42
1 Lecture 5. Design (1) CENG 412-Human Factors in Engineering May 21 2009

Upload: others

Post on 07-Feb-2022

1 views

Category:

Documents


0 download

TRANSCRIPT

1

Lecture 5. Design (1)

CENG 412-Human Factors in EngineeringMay 21 2009

2

Outline

Rationale for user-centric design–

Cost-benefit analysis of human factors contributions

Early focus on users and their tasks: User analysis

Reading: –

Wickens

pp. 30-39

For cost-benefit analysis: Norman, “The invisible computer”

Ch. 2 posted on course site

3

What activities can be qualified as design?

Development of new products or systems•

Modification of existing products or systems

Development of environments (workstations, complex/modular environments, smart homes for independent living, etc.)

Plan and perform safety-related activities ( safety analysis, industrial safety programs, design of warning labels, etc.)

Training programs & support material

4

Rationale for HF-based design•

Many products and systems are still designed and manufactured without adequate consideration of human factors.

• As products become more technologically sophisticated they frequently become more difficult to use: many advanced features are not used at all

• Product may be completed and then given to a human factors specialist to evaluate. Problems with that?

• Human factors can save companies time and money. To get the full benefit, human factors methods must be applied early in the

design process.

This chapter is the process.

Later chapters are the basic content information necessary to carry out the design process.

5From Norman, “The Invisible Computer”, Chapter 2. http://mitpress.mit.edu/books/NORVH/chapter2.html

The needs-satisfaction curve of a technology

6

The change in customers (users) as technology matures

From Norman, “The Invisible Computer”, Chapter 2. http://mitpress.mit.edu/books/NORVH/chapter2.html

7

Cost/benefit analysis

The best way to demonstrate the value of human factors to management

HF analysis= extra cost, which has to be justified•

A good example of estimating HF-related costs and benefits is (Mantey

and Teorey, 1988)

“Cost-benefit analysis for incorporating Human Factors in the Software lifecycle”

(optional reading)

8From Mantei and Teorey, 1988

9

Eight distinct costs are added to a SE project by the human factors stages.

(1) the cost of running focus groups;(2) the cost of building product mockups;(3) the expense of the initial design of a

prototype;(4) the expense of making a prototyping

design change;(5) the expense of purchasing the

prototyping software (UIMS system);(6) the cost of running the user studies;(7) the cost of creating a user study

environment (laboratory);(8) the cost of conducting the user survey.

From Mantei and Teorey, 1988

10From Mantei and Teorey, 1988

11

In-class activity

Handout # 3•

Time: 30 minutes

12

How to estimate HF benefits

Calculate the average time (errors etc) to perform a certain task

Estimate same variables for performance evaluation if a human factors effort is conducted

Consider frequency of performed tasks and number of people performing this task (over one year)

Example: see Table 3.2 p. 34

13

HF in the product design lifecycle

Front-end analysis•

Iterative design and testing

System production •

Implementation & evaluation

System operation & maintenance •

System disposal

HF specialists must be part of multidisciplinary design teams from the beginning.

14

General approaches for user centered- design

Early focus on users and tasks•

Iterative design using prototypes

Participatory design•

Empirical measurement (questionnaires and field studies on quantitative performance data)

15

What is the First Law of Human Factors Design?

Know Thy User.The User is Not You.

16

Designers are not typical users

“Steve Wozniak, the whiz-kid co-founder of apple Computer offered the first public glimpse of CORE, his latest brainchild. “CORE, which stands for controller of remote electronics, is a single device that allows consumers to fully operate their home equipment by remote control as long as the equipment is all in one room…CORE comes with a 40 page user manual. But Wozniak says users of his gizmo won’t be daunted because initially, most will be ‘techies’”

17

The designer’s clients may not be users

in my university, copying machines are purchased by the printing and duplicating center, then dispersed to the various departments. The copiers are purchased after a formal “request for proposals”

has gone out to manufacturers and dealers of

machines. The selection is almost always based on price, plus a consideration of cost and maintenance. Usability? Not considered. The state of California requires by law that universities purchase things on a price basis; there are no legal requirements regarding usability or understandability of the product. That is one reason we get unusable copying machines and telephone systems.”

From Norman, “The design of everyday things”

18

General approaches for user centered- design

Early focus on users and tasks•

Empirical measurement (questionnaires and field studies on quantitative performance data)

Iterative design using prototypes•

Participatory design

19

User analysis

Who are your users?–

young or old

experienced computer users or novices–

level of expertise

Generic software might have different types of users–

Designing for generic software: generic user with generic skills?

20

Techniques for user (and task) analysis

Talk to your users–

Structured interviews

Semi-structured interviews–

Open-ended discussions

Participatory design•

Watch your users–

Observational techniques

21

Interviewing strategies

Goals:–

to produce a descriptive model of current work

practice that can be used to guide further design activities

To get a clear picture about the user’s domain knowledge which will be used when interacting with your application

Techniques are adaptations of methods used by ethnographers and cognitive scientists

22

Expert knowledge

Potential users are experts in the work domain which the application is intended to support

Analysts typically underestimate the complexity of expertise in a domain of knowledge different from their own.

Aspects of expertise relevant to design:–

The organization of expert knowledge

The tacit nature of some knowledge–

The potential for experts to explain their expertise to you: translation competence

23

Organization of expert knowledge

Macro-level: hierarchic organization, taxonomy with categories and subcategories

Micro-level: “chunks”

of frequently occurring patterns stored in the long-term memory

Each chunk has an attached procedure for giving a fast appropriate response in a problem-

solving situation

24

Tacit knowledge

Much of an expert’s problem solving knowledge has become automatic through extensive and frequent use

Tacit (implicit) knowledge is difficult for an expert to articulate, especially when asked to do so directly

Solution?

25

Observational techniques

Observe users work in their typical work environment•

Main difference from interviews: analyst does not interact with user during observation (passive role)

Indirect observational techniques: surveys, questionnaires–

may replace direct observation (and interviews) for a first-

round user analysis under special circumstances.

After viewing the movie about the observational techniques used in the design of Office 7, list some major findings that were obtained by these observational techniques.

26

Building interviews

Structured interviews•

Centered on tasks (when, how and why a task is performed)

Semi-structured interviewing–

Mixture of specific and open-ended questions

Contextual interviewing–

Semi-structured interviewing

Ethnographic style observations -

can give you an idea about the cultural constraints of the interaction (i.e. general attitude about a specific type of user interface)

27

Contextual interviewing

Key issues:–

Deciding who to interview

Deciding how to ask and who to ask–

Analyzing the data

Planning the interview process

28

Who to interview?

People that are representative for your target users

Get a good mix•

Ask participants who else to talk to

How many? 6-10 according to usability experts

29

Developing user profiles

Iterative•

Develop an initial profile

Use that profile to determine an initial round of people to interview

Interview this round and update your profile•

Contact remaining participants and interview (iterate as necessary)

30

Developing user profiles

Think about relevant characteristics–

Those that all users might share

Those that might make a difference among users–

Example: for an on-line flight booking system, relevant user characteristics are: …

31

How to plan the interview

Identify who to interview–

Use initial user profiles

Make interview appointments•

Decide who will make the visit

Have a procedure•

Buddy system: interviewer/scribe

Consider pros/cons of taping•

Always summarize immediately afterwards together

32

What to ask? How to ask? Dos•

First thing: always explain to participants why you

do this interview.•

Semi-structured interviewingDevelop a set of questions that allow for open-ended

responses and elaborationsAsk specific questions about how people approach the task or

activity•

Understanding context

Have people work through the actual task throughthe meeting•

Critical incidents

Ask them to tell you about the best, the worst, andthe strangest time they did X.The Last Question

33

What to ask? How to ask it? Don’ts

questions with yes/no answers•

leading questions

Linear thinking•

Speculative questions–

Ex: Do you think you might use an electronic diary?

• system-oriented questions–

Ex: Do you think you'd use feature XYZ?

• general questions–

Ex: tell me how you do your job

34

What to ask? How to ask?

In-class critique of filmed interviews

35

Post-interview data analysis

Identify emerging patterns indicating:–

Important design priorities

Important design issues–

Develop and update user profiles and personas

• Capture priorities and issues as tasks

• Identify other information that you need

for building your first usability prototype:sorts and amounts of data, data sources,special equipment etc.

36

Personas

Hypothetical user•

What are personas used for?–

To represent a user (a real one) throughout the design process

To guard against designers and programmers making unreasonable assumptions about user needs, desires, skills

To provide context for discussing differences in design opinions

To make it possible to consider user skills in a rich, differential way

37

What does a persona look like?

A specific, detailed, and precise description of a person, with a name and picture, within the context of home and work

How many personas?–

Cooper: cast of characters and primary persona

38

An example persona for a banking application

Frances Miller•

Sixty-seven year-old Frances is the mother of four children and the grandmother of twelve.

She lives in her own home and has two cats.

She likes to knit and do needlework, which she either gives away as presents to her family or donates to the annual sale to raise money for the church she belongs to.

Every morning she goes for a one hour walk along the lake front when the weather is good.

On bad days she’ll go with her neighbor to the local mall where a group of senior citizens “Mall Stroll”

each morning before sitting down at one of the restaurants for coffee.

She is a middle-class retiree living on a fixed income.

Her mortgage has been paid off and she has one credit card which she seldom uses.

She has been a customer of the bank for 57 years although has never used an automated teller machine (ATM) and never intends to.

She has no patience for phone banking and does not own a computer.

Every Monday at 10:30 am she will visit her local bank branch to withdraw enough cash for the week.

She prefers to talk with Selma the branch manager or with Robert, a CSR who was

a high-school friend of her oldest son.

from http://www.agilemodeling.com/artifacts/personas.htm

39

Why identify a primary persona?

From Alan Cooper: “Design for pleasure”

40

Scenarios

Scenario design describes the persona’s process in achieving a goal It highlights opportunities for features and functionality developmentYou will be able to create a scenario after task analysis (next lecture)

41

Examples of exam questions•

You have just been put in charge of a design team in the early stages of developing a concept for an onboard automobile navigation system. The leading approach employs an LCD screen in the center console that shows location obtained through an onboard GPS (global positioning satellite) overlaid on an area map. Technologically, the most feasible candidates for input mechanism seem to be voice and touch screen. Your marketing people tell you your target customer is aged 55-70, male and in a middle-income bracket. This is a new demographic for the company, which has always focused on the low-end market. Before you came, the team had never heard of “user-centered design”.

One of the three basic principles of usability engineering is the early focus on users and their tasks. According to this principle, you

are

asked to design an interview with representative users in order to gather data for the first design iteration.–

Which type of interview (structured, semi-structured, contextual) is more suitable for this project? Justify your answer.

Design at least three interview questions and explain what information you expect to obtain by asking these three questions.

42

Examples of exam questions

Design an experiment that measures the benefits of usability testing for :

1) data entry using a conventional keyboard (baseline)•

2) data entry using an ergonomic keyboard

You will have to specify and justify the following:-

Between-user design versus within-user design

-

Independent variables-

Dependent variables

-

Confusing variables-

Statistical test