Presentation on theme: "People in systems design"— Presentation transcript:
1 People in systems design IMD07101: Introduction to Human Computer InteractionBrian Davison 2011/12With material from Tom McEwan, Kathy Buckner, Ian Smith and David Benyon
2 Agenda Legal Social Ethical and Professional Issues (LSEPI) PACT frameworkData collection techniquesShort breakContextsActivitiesMental modelsPACT analysisScenariosPersonas
3 LSEPI Legal Social Ethical Professional Laws about data protection, processes, illegal digital material in different jurisdictions (in a global Internet)SocialIt’s counterproductive to be anti-socialIT makes new social possibilitiesEthicalIn many ways a reflection of Legal and SocialMoral issuesProfessionalAll of the above – avoid doing the wrong thingsEnsure you do (all) the right things
4 Ethics & codes of conduct “Moral principles governing or influencing conduct”Oxford English Dictionary definition of ethicsPractitioners are governed by professional codes of conduct. egBCS, The Chartered Institute for ITBritish Psychological SocietyACM: Association for Computing MachineryUsability Professionals Association
5 Research involving more than minimal risk requiring explicit ethical clearance Vulnerable groupseg children, those with a learning disability, etc.Sensitive topicseg sexual behaviour, their illegal or political behaviour, etc.Groups where permission of a gatekeeper is normally required for initial access to memberseg ethnic or cultural groups, native peoples, etc.Covert researchCarried out without participants’ consentAccess to records of personal or confidential informationeg genetic or other biological information, etcPsychological stress, anxiety or humiliation or painIntrusive interventionseg administration of drugs, vigorous physical exercise, etc.
6 Informed consent The nature of the study What the participants be required to doHow long it will takeRisks and benefitsVoluntary participation and risk free withdrawalUse of dataConfidentiality of dataCompensation/reward for participationResults of study & contact detailsContact details of researcherUse an Information Sheet and/or Consent Form
7 Legal frameworksUK: Store personal data in accordance with Data Protection Act (1998)Completely anonymised data used for research purposes ‘may’ be exemptStore data securely; use it only for the purpose for which it was gathered; destroy it when no longer neededIncludes paper as well as computer records; includes data that can be processed by computer eg CCTV recordingsCan’t be transferred outside of European Economic Area (EEA) countries without consent of individualsPurpose of use must be identified when data is collectedSome data also comes under Freedom of Information (FOI) laws – can conflict with DPA – eg civil servants’ salary data
8 What does this mean for HCI practitioners and researchers Show respect for peopleindividual, cultural and role differences, including: age, disability, education, ethnicity, gender, language, national origin, race, religion, sexual orientation, marital or family status and socio-economic status.Obtain informed consentEnsure there is no undue pressure to participate and that there are no penalties for withdrawing from studyAnonymise resultsKeep the data secure (in accordance with DPA)Assume that someone may demand, through FOI, to see your data
9 Usability Professionals Association: Code of Conduct “Act in the best interest of everyoneBe honest with everyoneDo no harm and if possible provide benefitsAct with integrityAvoid conflicts of interestRespect privacy, confidentiality, and anonymityProvide all resultant data”Compare with BCS, Chartered Institute for ITCode of conduct – 3p.Code of Good Practice – 36p.
10 Requirements analysis People & HCI (PACT)DesignRequirements analysisEvaluationFrom Benyon (2009), p ??
11 People vary Physical Ergonomics - Size, shape, reach… Cognitive characteristicsWhat motivates, pleases and engages – “affect”Experience & expectationsLanguage, culture, morality …State of healthBut not usually age, race or gender in themselves
12 The machine-centred view People areMachines areVaguePreciseDisorganisedOrderlyDistractibleUndistractibleEmotionalUnemotionalIllogicalLogical
13 The person-centred view People areMachines areCreativeDumbCompliantRigidAttentive to changeInsensitive to changeResourcefulUnimaginativeAble to make flexible decisions based on contextConstrained to make consistent decisions
14 System design exampleYou work as a support engineer for an organisation which uses a range of office software.You’ve been asked to design an application for users to report computer problems via a new online logging system.There is an old logging system but no-one uses it and so the support desk is deluged with phone calls.What questions do you need to answer?This is a very typical sort of job which you might face in the real world. One of my placement students has had to do something very similar. Although apparently a fairly mundane application, it still poses design challenges.
15 Typical questionsWhat do users see as the difficulties with the existing reporting system?What basic skills do users have?How often would the system typically be used?What different user groups are there?
16 Techniques (dictionary.com) The manner and ability with which an artist, writer, dancer, athlete, or the like employs the technical skills of a particular art or field of endeavourThe body of specialized procedures and methods used in any specific field, especially in an area of applied scienceMethod of performance; way of accomplishingTechnical skill; ability to apply procedures or methods so as to effect a desired result
17 Data collection techniques How to find out about users and their activitiesObservation (both visual and electronic eg “audit trail”)Questionnaire surveysInterviewsVery early prototypingMany techniques recur in evaluation
18 ObservationGets at what people really do, not what they think or tell you that they doEach person shadowed for a dayFocus: information useElectronic, e.g. mobile phones and PDAsTraditional, e.g. newspapers and sticky notesRecorded through notes and videoUnmet needs, inefficiencies and use of artefacts (things!)Users continued with a text and/or audio logGroup interpretation sessionObserving people may change their behaviour“Hawthorne Effect” (1955)
21 Gathering information from users: questionnaires Need very careful design and piloting“Closed” questionsrating scaleeasy to analyse but need to balance extremes“Open-ended” questionsharder to analysericher informationCan use statistical analysis if the sample are large enough and representative enough.We will be doing a practical questionnaire exercise in one of the tutorials.alwaysusuallysometimesrarelyneverexcellentgoodDon’t know / neutralpoorVery poor
22 Closed questions Is Expression Web easy to learn? yes no Rating Scale How easy is it to use Expression Web?veryeasyverydifficult
23 Open questionsWhat was the easiest part of learning Expression Web? Write your answer in the box below.
24 InterviewsOpportunistic information gathering vs. demands on interviewerVarying degrees of structureStructuredinterviewer can explain the questions (unlike written questionnaire)but interviewee is limited to pre-set repliesSemi-structuredTell me about your typical dayTell me three good things about……and three bad thingsWhat if you had three wishes to make the system better?What has gone wrong with the system recently? How did you cope?What else should we have asked about?Planning is essential
25 Extensions of interviews Using prototypes as part of the interviewGetting users to talk as they workavoids people having to remember how they do thingssome things are easier to demonstrate than explainFocus groupscan help to get people talkingbut also can inhibit comments
26 Artefact investigation & collection Artefacts - in HCI jargon, things used, processed or created in an activityArtefacts illustrate aspects such aswhat data must be processedhow it is currently organisedwhat’s importantwhat instructions are neededmany more subtle aspects of activities
27 Artefacts and the clues they provide Medical notes - instant clues as to the length of the patient’s history, how many different doctors have been involved, whether last consultant in a hurry…Post-it notes with instructions stuck to machines and PCs - what’s difficult to understand or remember
28 More examples Design engineers working together across different sites engineer’s ‘day book’blueprints of the designBenefit claims processing systemcopies of blank and completed claim formsstandard letters sent to claimantsinter-office memospublic information leaflet about the benefit
29 Getting the data collected for you... Records ofhelp requestsbug reportschange requestsAutomatic usage loggingUser diaries of system use
30 Techniques summary ENVISIONMENT AND EARLY PROTOTYPING interviews questionnairesobservationartefact collectionEnd product: Documented system requirementsENVISIONMENT AND EARLY PROTOTYPINGusage logshelpdesk logsbug reportschange requestsTechniques on right hand side of course only apply where there is an existing computer based system.Techniques for envisionment and early prototyping covered in later lectures.
32 Different contexts of use Activities always take place in some context‘Context’ can meansthings that surround an activity, or,what glues an activity togetherPhysical environment is one type of contextATM or ticket machine versus computer at homeSocial context is importantHelp from others, acceptability of certain designsOrganisational contextPower structure, changes in lifestyle, de-skilling, etc.
40 Activities Characteristics of different activities Temporal aspects – when/how oftenTo do with timing, frequency etc.Co-operation and ComplexityWorking with others or notSafety criticalWhat problems happen if something goes wrong?ContentWhat information and media are we dealing with?How regular or infrequent are the activities?E.g. making a call on a phone vs. changing the batterySearching the Web versus working on an Excel spreadsheetBusy times versus quiet timesContinuous set of actions, or can be interrupted?Designing so that people can ‘find their place’ again after an interruptionResponse time from the system100 ms for hand-eye coordination activity1 second for cause-effect activityOver 5 seconds and people quickly get frustratedCo-operative or not? Is awareness of others and what they are doing important?Are they well-defined or vague? Browsing versus doing something clear.Safety-critical issuesData requirementsLarge amounts of alphabetic data - e.g. writingsmall amounts of static, unchanging data - e.g. swipe cardsMedia requirementsNeed for video, text, colour, sound, etc.
41 Activity: Designing a ticket machine A train station is introducing a new system of automatic barriers. Now everyone will have to buy a ticket before they travel.Write down the characteristics of this activity (5 mins in pairs)Regular/infrequent?Peaks and troughsInterruptible?Response timeCo-operation?Vague/well-defined?Safety critical?Errors?Data requirementsMedia
42 Different technologies Hardware and software to considerInputHow to enter data and commands into the system. Suitability of medium for different contexts/activitiesOutputCharacteristics of displays - ‘streamed’ media versus ‘chunked’ media. Characteristics of the content. Also feedback is importantCommunicationBetween person and technology. Bandwidth, speed. communication between devicesContentFunctional systems versus systems more focused on content
43 Ticket MachineSo, taking into consideration the contexts of use, the activities and the people:What technology will you design for the new ticket machines?(5 mins in pairs)ConsiderInputOutputCommunicationContent
44 Ticket Machine ideasInput - need to specify destination, need to provide payment, need to specify ticket typePress button (depending how many stations). Have touch screen (gets greasy). Pay by mobile phone? Contactless card (eg Oystercard)Output - need to specify options, need to provide a ticket, need to say when complete.Ticket could be electronic or paper. Printing facility needed. Options as buttons, or menu items? Need to provide change?Communication - must be simple.Could be Bluetooth. Probably button presses are easiest?Content - need to specify stations, but it could have lots of local information. Help with travel planning?
45 Mental models Also known as conceptual models… …mental models describe the ways in which we think about things - about how we conceptualize things.a key aspect of the design of technologies is to provide people with a clear model,… so that they will develop a clear mental model… but of course that depends on what they know already, their background, experiences, etc. etc.No such thing as “common sense”?
46 Ticket machine mental model Restrictions?Range of prices/times?Where?When?Price?Pay
47 Creating a mental model DesignerThe system imageTechnologyPerson using it has to work out how it works from interacting with the system image to develop the user’s mental model… has some idea about how it works… the ‘designer’s model’
48 Mental models Fill in the details that people don’t tell you “I had a haircut the other day…” – follows a “standard script”Are incomplete, in that they don’t include all the detailsYou could never know exactly what the designer knewCan be ‘run’ in that you use them in reasoning or rememberinge.g. how many windows are there in your house/flat?
49 PACT AnalysisUnderstanding the People, their Activities, Context and TechnologiesUndertaking a PACT analysis is a useful starting point for design…Useful for both analysis and designUnderstanding the current situationSeeing where possible improvements can be madeEnvisioning future situationsTo do a PACT analysis, scope the variety of the P’s, A’s, C’s and T’s in the particular domain
50 Doing a PACT analysisIdentifying the range of PACT elements in a domainBrainstorming (mindshower)Envisioning ideasDraw pictures, sketches, cartoons, cut out pictures from magazines and stick them on a board, etc.Work with relevant peopleWorkshops, interviews, observationsWrite up as scenarios (stories)
51 ScenariosScenarios are stories, about people undertaking activities using technologies in contextsDevelop conceptual scenarios that cover the main activities that the technology has to supportDevelop concrete versions of these for specific designs of the technology (implementation)For example - a conceptual scenario might say‘Pete logs onto the computerAnd a concrete version might be‘Pete clicks on the ‘log on’ icon’
52 PersonasA Persona is a profile of an archetypical person in the domainPersonas are synthesized from knowledge of real people in the domainPersonas need to have goals - describe what they are trying to achieveLike scenarios, conceptual personas are abstract typesstudents, lecturers, etc.For design it is best to develop a few concrete personaswho have hard characteristics such as age, interests, a name, etc.Try to bring the character alive - perhaps include a picture or two
53 Example PersonaWilliam is 70 yrs old retired bank manager. He is quite conversant with computers and latest technology and often uses them for and other uses. He has lived alone with his pet dog 'Tommie' for 4 years after his wife died of cancer. He had a good salary and owns many electronic gadgets and appliances. He is a good cook and normally cooks himself, although he orders take-aways occasionally. He is a Manchester United fan and likes to keep himself updated about the news in sports and politics. He is good natured and likes to watch comedy serials and chat with people. Recently he was diagnosed with diabetes and so has to visit his doctor regularly for checkups. Also he has to take precautions in his diet and take medicines on time. He loves his dog and look after it himself. He also keeps his garden green and tidy (which is also a hobby). He wants to enjoy his life and pursue his interests in his old age, but due to mental and physical degradation, he faces problems tackling everything himself and at times feels lonely and rather helpless.
54 Persona + Context setting Jan and Pat are a couple in their mid thirties. Pat is a university lecturer in Cultural Studies and Jan is an accounts manager at Standard Life insurance.They live in the Stockbridge area of Edinburgh, Scotland in a two-bedroom flat overlooking the river. It is noon on August 15th. Jan and Pat are sitting in their large, airy kitchen/dining room. The remains of pizza and mixed salad mingles with a pile of newspapers on the kitchen table.Jan and Pat have recently returned from a holiday on the island of Zante and, apart from checking their , have not gone back to work. They decide that they would like to go to see one of the events that is happening as part of the Edinburgh Arts festival.
55 PACT analysis example - A system for controlling access to laboratories at a university Peoplestudents, lecturers, technicians, people in wheel chairs, visitors, other ‘stakeholders’ e,g, cleaning staff, securityActivitiesget security clearance; one step; well-defined; no co-operation; not safety-criticalContextsindoors, people carrying books, etc. Socially - alone or in crowd. Political issues?Technologiessuitable for a small amount of data entered quickly. Very simple to use. Clear output that clearance is successful
56 PACT summaryPeople demonstrate a wide range of knowledge, abilities and other characteristics. Their mental models of things are criticalActivities have different characteristics which affect how we design to support themContexts of use affect the suitability of different designsTechnologies provide many opportunities for doing things differentlyActivities and the contexts in which they happen set requirements for the technologies which in turn provide new opportunitiesDoing a PACT analysis is a good way of understanding a situation; scope the Ps, As, Cs and Ts looking at the variety of each.
57 What’s next? Tutorial: Using PACT in the design process Practical: Web authoring with Expression WebNext week:Applying design techniquesBring one laptop per group if possibleLook in WebCT for details of the activities this weekCheck ahead to see what’s coming up