Jennifer Preece - Interaction Design.pdf

8/10/2019 Jennifer Preece - Interaction Design.pdf

1/40

Interaction Design: Beyond Human Computer Interactionby Jennifer Preece

Preece, J., Rogers, Y., & Sharp, H.. (2002).Interaction Design: Beyond Human-Computer

Interaction.

A typical undergraduate level textbook to introduce you to the field, including both scientific

background and usability design methods !ne of the fe" that ade#uately addresses affective

measures $D% & D'(

Jump To: ) * + - . / 0 1 )2 )) )* )+ ) )-

Table of Contents:

Ch ): Intro to Interaction Design

Ch *: Conceptual 3odels and Interface 3etaphors

Ch +: Cognition and 3ental 3odels

Ch : Conversational, Coordination, A"areness 3echanisms and Collaboration

Ch -: Affective Aspects of Interfaces and 4ser 5rustration

Ch .: 6he 7rocess of Interaction Design and 8ife9Cycle 3odelsCh /: 'eeds and e#uirements ;%cenarios, 4se Cases, uestionnaires, Inspections and ?alkthroughs

Ch ): 4ser 6esting, @!3%, 83, 5itts 8a"

Ch )-:


2/40

Chapter ): ?hat is Interaction DesignE

Main Goals of this Chapter:

Explain the difference between good and poor interaction design

Describe what interaction design is and how it relates to HCI and other fields

Explain what usability is

Describe what is inoled in the process of interaction design

!utline the different for"s of guidance used in interaction design

Enable you to ealuate an interactie product and explain what is good and bad

about it in ter"s of the goals and principles of interaction design

$oo% a#% Poor esig#

Central concern of interaction design: products that are 'S*+

+as to use

+ecti-e

+#oable


3/40

Building interactive versions of the designs so that they can be

communicated and assessed

+


4/40

%isibility of System Status

(atch bet!een system and real !orld

,ser control and freedom

Consistency and standards

Help users recognie. diagnose and recoer from errors

/rror preention

0ecognition rather than recall

&le1ibility and efficiency of use

)esthetic and minimalist design

Help and documentation

There are alas tra%eos ith usabilit ! ca#6t o-er co#strai# thi#gs, because it

limits ho much i#o is %isplae%

(top)

Chapter *: 4nderstanding and ConceptualiGing Interaction

have a clear understanding of hat, h a#% ho you are going to design something

before "riting any code

@oals of the chapter:


5/40

6o Develop a Conceptual 3odel:


6/40

9 ne" paradigms: ubi#uitous computing, pervasive computing, "earable computing,

augmented reality, attentive environments


7/40

extendhuman capabilities and

co"pensatefor their "eaknesses

3ain aims of chapter:

?hat cognition is and "hy it is important for I9D

3ain "ays cognition has been applied to I9D

'umber of examples from cognitive research

?4S =< PR=8+SS+S:

)ttention9 selecting things to concentrate on)

$erception 2 0ecognition9 ho" information is ac#uired from the environment viasense organs and translated into experiences ;vision is the most dominant=

*

(emory9 recalling various kno"ledge ?e filter "hat kno"ledge to process

memoriGe ;most researched area=

+

3earning 9 ho" to do something ;like learning to use a program=

0eading 2 Spea'ing 2 4riting9 using language-

$roblem Soling 2 $lanning 2 0easoning 2 Decision (a'ing9 involves reflective

cognition

.

!ften designers try to emulate the physical "orld "ith designs in the digital "orld

%ometimes this "orks "ell, other times it doesnt

Conceptual 5rame"orks for Cognition:

3ental 3odels

Information 7rocessing


8/40

9 thinks of the mind as an information processor

9 mental representations can be images, mental models, rules, other kno"ledge forms

the huma# processor mo%el;Card, et al )10+= is the best kno"n approach ;see p 1.=

model predicts "hich cognitive processes are involved "hen a user interacts "ith a

computer, allo"ing for calculations to be made on ho" long it "ill take a user to

complete a taskthis is helpful for comparing different interfaces ;efficiency=

the approach is based on modeling mental acti-ities that happe# e"clusi-el i#

the hea%. 6here are al"ays external cues in the environment so ho" truly

representative are these modelsE

$ there has been an increase in people studying cognitie actiities %in the wild% $ in the

context in which they ta&e place;ho" can things in the environment aid human cognition

and lighten the cognitive loadE=

Alternative frame"orks have been suggested: +"ter#al cog#itio# a#% istribute%

8og#itio#


9/40

humans are %!CIA8 beings

7urpose of the chapter: loo' at !ays interactie systems could be deeloped to

support and e1tend communication and collaboration bet!een peoples5

%ocial 3echanism in Communication and Collaboration:

ules, procedures and eti#uette have been established to help people kno" ho" to behave

in social groups, such as:

8o#-ersatio#al mecha#isms9 to help the flo" of talk and to help overcome

breakdo"ns

8oor%i#atio# mecha#isms9 to allo" people to "ork interact together

are#ess mecha#isms9 to find out "hat is happening, "hat others are doing and

to let others kno" "hat is happening

8o#-ersatio#al echa#isms:

turn9taking helps coordinate conversation

Implicit cues and


10/40

3any ne" communication technologies combine the above, and try to provide ne" novel

"ays to communicate

Collaborative virtual environments, media spaces, shared dra"ing tools, tools for

collaborative document creation

7ros: support talking "hile doing a task at same time, can be efficient to have

multiple people "orking on the same thing at the same time, and greater a"areness

of "hat is going on

Cons: ?M%I?I%: "e cant al"ays see "hat people are referring to in a remote

location, and floor control ;file conflicts from multiple people "orking on the same

thing at the same time=

8oor%i#atio# echa#isms:

Collaborative activities re#uire us to coordinate "ith each other, so "e need to figure out

ho" to "ork "ith others to progress through the activities


11/40

%ince %uchmans %eminal ?ork many companies have invested in ethnographic studies to

see ho" "ork actually gets done in a range of companies ;government too=

8o#ceptual


12/40

It has been suggested that computers be designed to recogniGe and express

emotions in the same "ay that humans do

Ho" can interactive systems be designed ;both deliberately and inadvertently= to

make people respond in certain "aysE

+"pressi-e 4#teraces

Colors, icons, sounds, graphical elements and animations are used to make the look

and feel of an interface appealing

( benefit is that these e"bellish"ents proide reassuring feedbac& to the user that

can be both infor"atie and fun$ which can affect the usability of the interface

7eople are "illing to put up "ith certain aspects of an interface ;slo" do"nload rate,

etc= if the end result is very appealing and aesthetic

Aesthetics have been sho"n to have a positive effect on peoples perception of the

systems usability

%ome friendly interfaces: 3icrosofts at home "ith Bob interface, +D metaphors

;living rooms, etc=, agents in the guise of pets ;dog= that talk to the user 6hese

make users feel more at ease and comfortable4ser9created expressiveness: emotico#s9 these provide non9verbal type expression

in interfaces not originally intended to have this :9= Also, icons shorthand have

been used to add emotion to %3% texting ;I )* C4 *'I6


13/40

soda cans, dogs, cars, etc=

9 used heavily in advertising

7eople debate ho" much of this to use in system design 6hey can add a human feel to the

system, but can also get annoying

,hich is "ore preferable-

9 Hello 3attN ?elcome back Its nice to see you again 'o", "hat "ere "e doing last

timeE Ah, yes, problem five 8ets get started again

9 4ser *, commence exercise -

!r, "hen doing something "rong: 'o" 3att, thats not right, you can do better than

that 6ry again vs Incorrect, try again

#he answer:

7ros: eeves and 'aas ;)11.= found it is helpful to use praise in educational settings "hen

people do something right It increased students "illingness to continue "orking

Cons: Ho"ever, others argue this can make you feel stupid, anxious, inferior 7eople hate

"hen a computer character shakes their finger at them and says you can do better than

that, 3att, try again In this case, many prefer the impersonal message Incorrect, try

again

Airtual 8haracters

virtual characters are becoming more common 6hey can be used on the "eb, in video

games, as learning companions, "iGards, ne"sreaders, etc Ho"ever, they can be

misleading ;people confide in them=, they can be very annoying and frustrating ;Clippy

from 3% !ffice 1/, etc=

9 categoriGed by degree of anthropomorphism: synthetic characters, animated agents,

emotional agents, embodied conversational agents

Design I"plications: which one to use-

! belie-abilit: the extent to "hich users come to believe an agents intentions andpersonality

! appeara#ce: better to use cartoon9like characters, or those resembling humansE it

depends on the situation9 often the cartoon character is more believable acceptable

! beha-ior: ho" does the agent move, gesture, refer to thingsE facial expressions can

sho" emotion ;remember "e "ant constructive feedback rather than conveying inferiority

stupidity patroniGing effect on users=

Summar:

affective aspects are concerned "ith ho" interactive systems make people respond in

emotional "ays"ell designed interfaces can elicit good feelings in users

expressive interfaces can provide reassuring feedback

badly designed interfaces make people angry and frustrated

anthropomorphism is increasingly used at the interface, through the use of agents

and virtual screen characters

(top)

Chapter .: 6he 7rocess of Interaction Design

6he ultimate goal of design is to develop a product that helps its users achieve their goals

Developing a product must begin "ith gaining understanding of "hat is re#uired of it

6he goals of this chapter are to:

nifer Preece - Interaction Design http://www.sharritt.com/CISHCIExam/preece.html

f 40 11/14/2014 6:25 PM


14/40

Consider "hat doing interaction design involves

Ask and provide ans"ers for some important #uestions about the interaction design

process

Introduce the idea of a lifecycle model to represent a set of activities and ho" they

are related

Describe some lifecycle models from soft"are engineering and HCI and discuss ho"

they relate to the process of interaction design

7resent a lifecycle model of interaction design

?hat is Interaction DesignE

dictionary: design is a plan or scheme conceived in the mind and intended for subse#uent

execution

6he plan or scheme must be informed "ith kno"ledge about its use and the target domain,

together "ith practical constraints such as materials, cost and feasibility

In Interaction Design, "e investigate the artifacts use and target domain by taking a

user9centered approach to development 6he users concerns direct the development

rather than technical concerns

Design is also about trade9offs and about balancing conflicting re#uirements @enerating

alternatives is a key principle and one that should be encouraged in interaction 6o get a

good idea, get lots of ideas ;3ark ettig=

6ypically there is a group of designers 6herefore, plans should be captured and expressed

in a "ay that allo"s for revie", such as sketches, descriptions in natural language, a series

of diagrams, and building prototypes

5our Basic Activities:

./ Identify needs and establish re0uire"ents9 ?ho is the target userE

9 ?hat kind of support "ill the interactive product provideE

1/ Deelop alternatie designs that "eet those re0uire"ents

9 suggest ideas to meet the re#uirements

9 Conceptual design: produce the conceptual model for the product

9 7hysical design: consider the details of the product ;colors, sounds, images, menu

design, icons, etc= Alternatives are considered at every point

2/ 3uild interactie ersions 4so that they can be co""unicated and assessed5

9 a soft"are version is not re#uired9 paper based prototypes are #uick and cheap to build

9 through role9playing, users can get a real sense of "hat it is like to interact "ith theproduct

6/ Ealuate the designs 4"easure their acceptability5

9 determine the usability of the product or design Criteria are: ho" appealing is itE ho"

"ell does it match the re#uirementsE Is the product fit for the purposeE

9


15/40

to choose bet"een different alternative designs

9. 4teratio#

9 allo"s for designs to be refined It is al"ays necessary to revise ideas in light of

feedback, several times Innovation rarely emerges "hole and ready to go Iteration is

inevitable because designers never get the solution right the first time

7ractical Issues in Interaction Design:

*5 4ho are the users6

9 those "ho directly interact "ith the system Ho"ever, can be any stakeholder:

purchaser, testers, people receiving products from the system

9 primary user: directly use it

9 secondary user: occasionally use it or use through intermediary

9 tertiary user: affected by the system, or "ill influence its purchase

9 stakeholders: people or organiGations affected by the system "ho influence the system

re#uirements

2. /hat %o e mea# b 1#ee%s1B

9 "e need to understand the characteristics capabilities of users, "hat they are trying to

achieve, ho" they currently achieve it, and "hether they "ould achieve their goals more

effectively if they "ere supported differently

9 characteristics that impact a products design: users physical characteristics

;ergonomics: siGe of hands, height, etc=, strength of product ;so a child cant break it=,

cultural diversity of intended users

9 representative users 34%6 be consultedN

9 users rarely kno" "hat is possible 6herefore, users cannot tell us "hat they need to

do achieve their goals

- 4e need to e1amine e1isting tas' 7)ctiity Theory68 and !hat the tas's#

conte1t. re9uirements. collaboratie nature. and procedure is5 Then !e can

enision the tas' being done in a ne! !ay 7scenarios. etc58

9. Ho %o ou ge#erate alter#ati-e %esig#sB

9 it is easy to stick "ith something that is good enough Humans stick to "hat they kno"

"orks

9 innovations arise from cross9fertiliGation from different applications9 allo"s us to break

out of the box

9 often bro"sing a collection of designs "ill inspire designers to consider alternative

perspectives and solutions Designers are trained to consider alternatives, soft"are people

are not

9 design is a process of balancing constraints and constantly trading off one set of

re#uirements "ith another, and the constraints may be such that there are fe" viable

alternatives available

9 alternatives come from looking at other, similar designs, and the process of inspiration

and creativity can be enhanced by prompting a designers o"n experience and by looking at

others ideas and solutions ;5lair and creativity : research and synthesis=

;. Ho %o ou choose amo#g alter#ati-e %esig#sB

9 there are factors that are externally visible and measurable and those that are hidden

from the users vie" +ocus on the external ) isible

9 prototypes can be used to evaluate "ith peers and users

9 fundamental user9centered design: choose bet"een alternative designs by letting users

and stakeholders interact "ith them and by discussing their experiences, preferences and

suggestions for improvement

9 technical feasibility: some are ust not possible

9 #uality thresholds: usability goals lead to criteria 6his 4%ABI8I6M CI6


16/40


17/40

#hese are per"eated with three principles:

) Involve users early in the design process and evaluation of the artifact

* Define #uantifiable & measurable usability criteria

+ Iteration is inevitable

ey characteristics of the interaction design process are explicit incorporation of user

involvement, iteration and specific usability criteria

Before you can begin to establish re#uirements, you must understand "ho the users are

and "hat their goals are in using the device

8ooking at others designs provides useful inspiration and encourages designers to consider

alternative design solutions, "hich is key to effective design

4sability criteria, technical feasibility, and users feedback on prototypes can all be used to

choose among alternatives

7rototyping is a useful techni#ue for facilitating user feedback on designs at all stages

8ifecycle models sho" ho" development activities relate to one another

6he interaction design process is complementary to lifecycle models from other fields

(top)

Chapter /: Identifying 'eeds and


18/40

proect to 5AI8

A user9centered approach to development is the "ay to solve this problem ;?hat do

users "antE ?hat do users needE=

e#uirements need clarification, refinement, completion and re9scoping

Input: re#uirements document ;maybe=

!utput: stable re#uirements

?hy establish re#uirementsE


19/40

flexibility

attitude

Data athering Techni9ues: 7see p5*< 2 Table =5* for e1cellent graph8

Duestio##aires

elicit specific information

can be M


20/40

7roblems "ith data9gathering:

identifying and involving stakeholders: users, managers, developers, customer

repsE, union repsE, shareholdersE

involving stakeholders: "orkshops, intervie"s, "orkplace studies, co9opt

stakeholders onto the development team

eal users, not managers: traditionally a problem in soft"are engineering, not so

bad no"re#uirements management: version control, o"nership

communication bet"een parties: "ithin development team, "ith customer user,

bet"een users

domain kno"ledge distributed implicit: difficult to dig up and understand

availability of key people

political problems

dominance of certain stakeholders

economic business environment changes

balancing functional and usability demands

$ui%eli#es:

focus on identifying the stakeholders needs

involve all the stakeholder groups

involve more than one stakeholder from each group

use a combination of data gathering techni#ues

start data interpretation soon after the data gathering session

do an initial interpretation before deeper analysis

use different approaches to different problems, such as class diagra"sfor obect9

oriented systems, and entity$relationship 4E$5 diagra"sfor data intensive systems

Tas' Descriptions: 7more on p5 >-?*8

Sce#arios9 an infor"al narratie story, simple, natural, personal, not generaliGable

'se 8ases9 assume interaction "ith a system, assume detailed understanding of

the interaction

+sse#tial 'se 8ases9 abstract a"ay from the details, does not have the same

assumptions as use cases

8cenariofor a shared calendar:

P6he user types in all the names of the meeting participants together "ith

some constraints such as the length of the meeting, roughly "hen the meeting

needs to take place, and possibly "here it needs to take place 6he system

then checks against the individualsJ calendars and the central departmental

calendar and presents the user "ith a series of dates on "hich everyone is free

all at the same time 6hen the meeting could be confirmed and "ritten into

peopleJs calendars %ome people, though, "ill "ant to be asked before the

calendar entry is made 7erhaps the system could email them automatically

and ask that it be confirmed before it is "ritten inQ

*se casefor a shared calendar:

) 6he user chooses the option to arrange a meeting

* 6he system prompts user for the names of attendees


f 40 11/14/2014 6:25 PM


21/40

+ 6he user types in a list of names

6he system checks that the list is valid

- 6he system prompts the user for meeting constraints

. 6he user types in meeting constraints

/ 6he system searches the calendars for a date that satisfies the constraints

0 6he system displays a list of potential dates

1 6he user chooses one of the dates

)2 6he system "rites the meeting into the calendar

)) 6he system emails all the meeting participants informing them of them

appointment

Alternative courses for a shared calendar:

%ome alternative courses:

- If the list of people is invalid,

-) 6he system displays an error message

-* 6he system returns to step *

0 If no potential dates are found,

0) 6he system displays a suitable message

0* 6he system returns to step -

Exa"ple *se Case Diagra"for a shared calendar:


22/40

Hierarchical Tas' )nalysis - involves breaking do"n a task into subtasks,

then sub9sub9tasks and so on 6hese are grouped as plans "hich specify ho"

the tasks might be performed in practice

H6A focuses on physical and observable actions, and includes looking at

actions not related to soft"are or an interaction device

H6A starts "ith a user goal "hich is examined and the main tasks for

achieving it are identified

6hese tasks are then divided into sub9tasks


23/40

%cenarios, use cases and essential use cases can be used to articulate existing and

envisioned "ork practices

6ask analysis techni#ues such as H6A help to investigate existing systems and

practices

(top)

Chapter 0: Design, 7rototyping and Construction

6his chapter "ill cover: prototyping and construction ;lo" and high fidelity prototyping,

vertical and horiGontal compromises=O conceptual design ;conceptual model, using

scenarios and prototypes in conceptual design=O and physical design ;guidelines and

"idgets=

5lo" of Interaction Design:

Identify 'eeds e#uirements ;Ch /= 99R 7rototype cycles Design ;Ch 0= 99R

Construction

?hat is a 7rototypeE

in other fields, its a small scale model ;miniature car, building, etc=

in Interaction Design it can be a series of screen sketches, a storyboard, a

7o"er7oint, a video simulating use of the system, a lump of "ood ;eg 7alm7ilot=, a

cardboard mock9up, or a piece of soft"are "ith limited functionality

?hy 7rototypeE


24/40

fidelity prototypes are time consuming expensive to make, and are not effective in

re#uirements gathering

Compromises associated "ith 7rototyping:

every prototype has a compromise 9 for soft"are this may be slo" response time,

sketchy icons, limited functionality, etc

6"o types of compromise: horiGontal and vertical

horiGontal compromise: provide a "ide range of functions, but "ith

little detail

vertical compromise: provide a lot of detail for only a fe" functions

Compromises must not be ignored: products need to be engineered

Construction: taking a prototype and making it "hole by engineering a complete product

;focus on #uality: usability, reliability, robustness, maintainability, integrity, portability,

efficiency, etc=

8o#ceptual esig#:


25/40


26/40

accommodatedE

ico# %esig#: can be difficult, as icons can be cultural context sensitive 9 so

dra" on traditions standard, use concrete obects

scree# %esig#: %plit screenE Ho" much "hite spaceE Ho" to group things

;boxes lines colors=E Dra" attention to the focus point, using color, motion,

possibly animation, and use good organiGation Balance the tradeoff bet"een

overcro"ded sparse displays

4#ormatio# %ispla: sho" only relevant information, make differentmediums ;computer paper= consistent

#here is no rigid border between conceptual and physical design/// they are all iteratie

processes/ !ften in conceptual design so"e detailed issues co"e up in the iterations/ #he

i"portant part is that in the conceptual design that we don%t get tied to physical constraints

early as they will inhibit creatiity and li"it our options/

%ummary:

Different kinds of prototyping are used for different purposes and at different stages

7rototypes ans"er #uestions, so prototype appropriately

Construction: the final product must be engineered appropriately

Conceptual design ;the first step of design=

7hysical design: eg menus, icons, screen design, information display

7rototypes and scenarios are used throughout design

(top)

Chapter 1: 4ser9Centered Approaches to Interaction Design

The main aims of this chapter are to:

Explain so"e adantages of inoling users in deelop"ent/

Explain the "ain principles of a usercentered approach/

Describe so"e ethnographicbased "ethods ai"ed at understanding users% wor&/

Describe so"e participatie design techni0ues that help users ta&e an actie part in

design decisions/ 4users as co$designers will raise acceptance of product5

4hy inole users at all6

6o manage their expectations: no surprises, communicate their expectations, get a

common set of realistic expectations

)

%o users feel ownership: by making users active stakeholders, they are more likely

to forgive accept problems, and are more likely to accept the final product

*

Degrees of user inolement:

member o the %esig# team ;part ti"e s/ full ti"e: degree of input time and

contactO short ter" s/ long ter": degree of consistency across proect life 8ong

term members might loose contact "ith users=

#esletters 5 e!mail 5 etc ;disseminate information to a large selection of users,

but re#uires *9"ay communication, not )9"ay=

Actual user involvement may be a combination of the above t"o "ays

3icrosoft involves users by activity based planning ;studying users doing tasks=,

usability tests, internal developer usage of products, and customer support lines


f 40 11/14/2014 6:25 PM


27/40

4hat is a user-centered approach6 4ser9centered approach is based on:

Early focus on users and tas&s:directly studying cognitive, behavioral,

anthropomorphic & attitudinal characteristics

users tasks and goals are the driving force behind the development

users behavior and context of use are studied and the product is designed to

support them

users characteristics are captured and designed forusers are consulted throughout development, from earliest phases to the

latest, and their input is seriously taken into account

all design decisions are taken "ithin the context of the user, their "ork, and

their environment

E"pirical "easure"ent: usersJ reactions and performance to scenarios, manuals,

simulations & prototypes are observed, recorded and analyGed

Iteratie design:"hen problems are found in user testing, fix them and carry out

more tests

,nderstanding ,sers# 4or': /thnography

/thnographystems from anthropology, and literally means "riting the culture 9 a

form of participant observation Ho"ever, it is difficult to use the output of

ethnography in design Design is concerned "ith abstraction and rationaliGation,

"hile ethnography is concerned "ith minute details, so it is difficult to harness the

data gathered from ethnography so that it can be used in design

+ra"ewor& for using ethnography in design:

distributed coordination: distributed nature of the tasks activities and the

means mechanisms by "hich they are coordinatedplans and procedures: organiGational support for the "ork, such as "orkflo"

models and organiGational charts, and ho" these are used to support the "ork

awareness of wor&: ho" people keep themselves a"are of others "ork

Coherence: a method offering #uestions to address these dimensions ;above= by

presenting the ethnographic study data as a set of ;iewpoints;and ;concerns;

+"amples:

Distributed coordination: Ho" is the division of labor manifested through

the "ork of individuals and its coordination "ith othersE

Plans and procedures: Ho" do plans and procedures function in the"orkplaceE

(wareness of wor&: Ho" does the spatial organiGation of the "orkplace

facilitate interaction bet"een "orkers and "ith the obects they useE

Conte1tual Design: developed to handle data collection and analysis from field"ork for

developing a soft"are9based product ;used commercially #uite "idely= 6here are seven

parts to Contextual Design:

7. 8o#te"tual 4#Cuir

2. /or3 o%eli#g

9. 8o#soli%atio#

;. /or3 Re!%esig#E. 'ser +#-iro#me#t esig#

F. oc3!up a#% test ith customers

G. Putti#g it i#to Practice


f 40 11/14/2014 6:25 PM


28/40

*5 Conte1tual In9uiry: an approach to ethnographic study "here the user is

an expert, and the designer is an apprentice It is a form of intervie"ing, but

takes place at the users "orkplace "orkstation, and is often *9+ hours long

5our main principles of contextual in#uiry are:

) Context: see "orkplace and "hat happens

* Partnership: user and developer collaborate

+ Interpretation: observations interpreted by user and developer

together

+ocus: proect focus to help understand "hat to look for

5 4or' (odeling: In interpretation sessions, models are dra"n from the

observations 5ive models are:

/or3 lo mo%el: the people, communication and coordination

SeCue#ce mo%el: detailed "ork steps to achieve a goal

rtiact mo%el: the physical things created to do the "ork

8ultural mo%el: constraints on the system from organiGational culture

Phsical mo%el: physical structure of the "ork, eg office layout

?5 Consolidation: each contextual in#uiry ;one for each user developer

pair= results in a set of models, "hich need to be consolidated into one vie" of

the "ork

i#it iagram: organiGes interpretation session notes into common

structures and themes

Categories arise from the data

Diagram is built through induction

?ork models consolidated into one of each type

$articipatory Design: ;%candinavian background= emphasiGes social and organiGational

aspects 9 based on study, model9building and analysis of ne" and potential future systems

Aspects to user involvement include:

?ho "ill represent the user communityE Interaction may need to be assisted by a

facilitator

%hared representations

Co9design using simple tools such as paper or video scenarios

Designers and users communicate about proposed designs

Cooperative evaluation such as assessment of prototypes

Benefits of 7articipatory Design: PComputer9based systems that are poorly suited to ho"

people actually "ork impose cost not only on the organiGation in terms of lo" productivity

but also on the people "ho "ork "ith them %tudies of "ork in computer9intensive

"orkplaces have pointed to a host of serious problems that can be caused by ob design

that is insensitive to the nature of the "ork being performed, or to the needs of human

beings in an automated "orkplaceQ $uhn, % in Bringing Design to %oft"are, )11.(

7IC6IF


29/40


30/40

discuss ho" developers cope "ith real9"orld constraints

explain the concepts and terms used to discuss evaluation

examine ho" different techni#ues are used at different stages of development

?hat to evaluate:

iterative design and evaluation is a continuous process that examines: early ideas for a

conceptual modelO early prototypes of the ne" systemO later, more complete prototypes

Designers need to check that they understand users re#uirements

?hy evaluate:

9 because user experience can be extremely important for a products success

9 because the cycle of %esig#and testi#gis the only validated methodology in existence

that "ill consistently produce successful results

&ie good reasons for inesting in user testing 7Tognaini8:

9 fixed problems before the product is shipped

9 the team can concentrate on real problems9


31/40

uestionnaires

Asking experts their opinions is inexpensive and #uick

6esting users performance ;ch )=

"ays to measure user performance to compare * or more designs

modeling users task performance to predict the efficacy and problems of a user

interface


f 40 11/14/2014 6:25 PM


32/40

some of these techni#ues: @!3% and the keystroke model

D


33/40

@uic' A Dirty obseration: can occur any"here, anytime @ood for immediate

feedback


34/40

revie" data and identify key themes)

record themes in a coherent flexible form*

recode the data and time of data analysis session+

check your understanding "ith the people you observe

iterate process until your story represents "hat you observed-

report findings to development team ;oral or "ritten report=.

analyGing and reporting ethnographic data ;from participant observation, intervie"s,artifacts:

look for key events)

look for patterns of behavior ;in various situations and players=*

compare sources of data+

report findings

Cualitati-e a#alsis or categori@atio#

look for incident patterns)

analyGe data into categories ;content analysis=*

determine the content categories reliability and inter9research reliability ratings+

analyGe discourse ;conversation analysis=

*

Cua#titati-e %ata a#alsis

video data collected in usability labs)

annotated*

recording to calculate performance times+

analyGe statistically

+

uestionnaires are used in #uick and dirty evaluation, in usability testing,

and in field studies to ask about facts9 behaior9 beliefs and attitudes/


f 40 11/14/2014 6:25 PM


35/40

Intervie"s

types: can be ope#!e#%e% (u#structure%), structure%, semi!structure%, or

group i#ter-ies

unstructured: no script, not replicable

structured: tightly scripted like a #uestionnaire

semi-structured: guided by script but open for deeper exploration if desired

group: often a focus group of +9)2 people, consensus reached on #uestionsdata analysis on structured interiews is li&e a 0uestionnaire9 while

unstructured li&e participant obseration

* types of #uestions: open and closed Closed re#uire intervie"ee to choose

bet"een options, open allo" for a free9range response

>uestionnaires

can use: yes noO 8ikert scaleO semantic scaleO open9ended responses on #uestions

to make a good #uestionnaire, provide a clear purpose statement, plan the #uestions,

decide if phrases have a positive or negative connotation, pilot test your #uestions,

and decide ho" the data "ill be analyGed

to reach a large amount of people: guarantee anonymity, offer online ;large base

instant results & often instant data analysis=

Be careful9 online #uestionnaires dont prevent people from ans"ering multiple times,

and can have a lo" response rate

data analysis includes identifying trends9 using si"ple statistics9 "a&ing use of

percentages < bar graphs

Heuristic


36/40

Cons: can be hard to find experts, important problems can get missed, often trivial

problems get identified

'ote: different combinations and types of heuristics are needed to evaluate different

types of applications and products

?alkthroughs

"alkthroughs are an alternative to heuristic evaluation and involve "alking through

a task "ith the system an noting problematic usability features often these dont

involve users

6"o main types: 8og#iti-e /al3throughsand Pluralistic /al3throughs

8og#iti-e /al3throughssimulate a users problem solving process at each step in

the human9computer dialog, checking to see if the users goals and memory for

actions can be assumed to lead to the next correct action

focus on ease of learning

designer presents an aspect of the design and usage scenarios

one of many experts "alk through the design prototype "ith the scenario

expert is told the assumptions about user population, context of use, task

details

uestionnaires are a cheap and easy "ay to reach large numbers of people


f 40 11/14/2014 6:25 PM


37/40

9 typically - experts can find /-U of the usability problems

9 heuristic evaluation is cheaper and more flexible than user testing

9 4ser testing and heuristic evaluation often reveal different usability problems

9 pluralistic and cognitive "alkthroughs are focused and good for evaluating a small part of

the interface

(top)

Chapter ): 6esting and 3odeling 4sers

A central part of Interaction design is user testing

4sability testing uses a combination of techni#ues, including user testing and user

satisfaction #uestionnaires 4ser testing is of central concern

6he end of this chapter talks about @!3% ;@oals, !perators, 3ethods, %election rules=,

83 ;eystroke 8evel 3odel= and 5itts 8a"

4ser 6esting

4ser testing is applied experimentation in "hich developers check that the system

being developed is usable by the intended user population for their tasks

4ser testing tests typical users, measuring their typical task time, and the number

and type of errors are recorded

Can consist of completion time, observational data, ans"ers to #uestionnaires,

ans"ers from Intervie"s, and keystroke logs

4ser testing is a systematic approach to evaluating user performance to inform or

improve usability design

4sually there are fe" participants ;-9)2=, but can be )9* in #uick and dirty for

#uick feedback

6ypically "e record: task completion timeO task completion after being a"ay from

the productO number types of errorsO errors per unit of timeO number of navigations

to help manualsO number of users making a particular errorO number of users

completing a task successfully

4sing the D


38/40

Design %-a#tages isa%-a#tages

Different

Participants'o order effects

3any participants needed Individual

differences bet"een participants is a

problem Can be offset to some extent

by randomly assigning to groups

8a"e

participants


39/40

relatively easily

outcome: counter9intuitive, help make decisions about the effectiveness of ne"

products

Cons: not often used for evaluation purposes because of its highl limite% scope

only good for predicting e"pert perorma#ce, and error is #ot mo%ele%;average

users not predicted=

many unpredictable factors come into play

A Con of 7redictive 3odels: they can make predictions about predictable behavior, but it is

difficult to use them as a "ay of evaluating ho" systems "ill be used in the real "orld

6hey are only useful for comparing the efficiency of different methods in completing a

short, simple task

9.


40/40

"hich combination of methods to use "hen designing and evaluating your productE

"hat happens "hen the product being developed is confidential and there are no

users available to test itE

ho" many users should be involved in testsE

"hat should "e do "ith the evaluation findingsE

ho" much should "e expect from usersE

Case study: 'okias mobile communicator

9 used ethnographic research and did scenarios and task models to get re#uirements

9 ;method= follo"ed participatory design9 involved users throughout

9 ;method= used interface metaphors

9 ;method= follo"ed "ith fre#uent lo" fidelity prototypes based on alternative designs

immediate evaluation ;yielded invaluable insight to designers=

9 "rote usage scenarios ;high level descriptions of device in use=

9 user testing ;usability tests=: did summative testing before release ;at end, not

throughout $formative(= and #uestionnaires after release

Case study: redesign of a telephone response information system ;6I%=

9 current system "as very hard to use9 a deep menu system over the phone9 users cant

remember it "ithout cues

9 @!3% 83 used to sho" ho" interface supported users tasks

9 heuristic evaluation used as an alternative method for sho"ing usability problems ;expert

revie" of system=

9 methods complemented each other and sho"ed benefit of doing a re9design

ey points:

esig# i#-ol-es tra%e!os that ca# limit choices but ca# also result i#

e"citi#g %esig# challe#ges

7rototypes can be used for a variety of purposes throughout development, including

for marketing presentations and evaluations

6he design space for making upgrades to existing systems is limited by the design

decisions previous system 6he design space for ne" products is much greater

Cycles of rapid prototyping and evaluation allo" designers to examine alternatives in

a short time

%imulations are useful "hen evaluating systems used by large numbers of people

"hen it is not feasible for them to "ork on the system directly

Pieci#g together e-i%e#ce rom %ata rom a -ariet o sources ca# pro-i%e a

rich picture o usabilit problems, h the occur, a#% possible as o

i"i#g them.

(top)

Jump To: ) * + - . / 0 1 )2 )) )* )+ ) )-


Documents

Jennifer Preece - Interaction Design.pdf