The Rich Picture A Tool For Reasoning About Work Context

R33A N D R E W M O N K A N D S T E V E H O W A R Dmethods& tools Rich Picture: 10.1-P101, Root Definitions: 10.2The Rich Picture: A Tool forReasoning About Wconsider logicallyork ContextTThe Importance of ConcernsHave you ever observed the following situation? A computer system is built to satisfy well-specifiedrequirements. The requirements clearly describe the task to be supported, and the system express complete disapproval ofsatisfies them. Despite all this care and attention, the system is universally condemned by management and users. Why does this happen? Surprisingly often, the task supported is not one thatusers actually perform. More likely, the model of work underlying the computer system interfereswith other tasks the user wants to perform.i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 821terribleThe really catastrophic side effects are thoseOrigins of Rich Pictures that prevent other people from doing theirRich pictures originated in the Soft Systemswork. If the chief accountant can no longer getMethodology (SSM) [4, 3, 18]. SSM, in turn,the figures she needs, the system will never seehad its origins in sociotechnical approaches tothe light of day!system design [15]. Within this tradition,Andrew MonkA new computer system will affect the wayidentifying multiple viewpoints of a work sit-repeatedDepartment ofpeople work; if it does not there is no point inuation is a recurrent problem. SSM was devel-Psychologyintroducing it in the first place. These effectsoped during the 1960s and 1970s by Peterharmful University of Yorkwill be deleterious if the developers do notCheckland and his students at LancasterconsequenceYork, Y01 5DD consider the implications for both the system’sUniversity. At the core of SSM is a desire toUnited Kingdomusers and other people who may be affected byunderstand human activity systems in a wayAM1@york.ac.ukuse of the system. All work has numerous, andthat is meaningful to the actors in that system.sometimes competing, objectives. A singleSSM consists of seven main stages that pro-expressSteve Howarduser may have the objectives “to complete aceed from articulating the problem situation,Swinburne CHIjob well” and “to get home soon.”through building alternative systems models,Laboratory (SCHIL)Management may have the objectives “to cutto making recommendations for action.Swinburne Universitythe head count in this department” and toCheckland proposes the rich picture as a rep-of Technology“minimize the transaction times for cus-resentation to be used at the beginning of thisPO Box 218, tomers.” We call these objectives “concerns.”process.Hawthorn, 3122Concerns are the high-level objectives that sig-Rich pictures are generally constructed byAustralianificantly constrain the way work is done.interviewing people. The ideal interviewSHoward@swin.edu.auEffective systems can be designed only by tak-should take place at the workplace because thedifferenttooling into account the divergent concerns ofartifacts people use to do their work will bestakeholders. A designer may think she is tak-close at hand. They will be able to show youing an “impersonal view of the problem,” butdocuments and products, and you may evenexactthe very act of identifying the problem impliesbe able to observe them doing their work. Thea particular viewpoint.rich picture serves to organize and reasonHow is a designer therefore to reason aboutabout all the information that users provide.these divergent concerns that motivate theDrawing the picture will point to places whereway different stakeholders view the systemyou need to find out more or to apparent con-inconsistency, conflict they are designing? This paper discusses a sim-tradictions in the conclusions you have drawn.ple graphical device, called a rich picture, thatIn the latter case you will need to go back tohas been found to be useful in this respect. Ayour informants and then make changes basedrich picture is a cartoon-like representationon what they tell you. Drawing a rich picturethat identifies all the stakeholders, their con-then is an iterative process of understandingcerns, and some of the structure underlyingand refining that understanding.the work context. A rich picture is a tool forWhat does a rich picture look like? Therecording and reasoning about these aspects ofrich picture depicts the primary stakeholders,the work context, in particular, how theytheir interrelationships, and their concerns. Itshould affect the design. It is a tool in theis intended to be a broad, high-grained viewsense that a notation or representation is aof the problem situation. There is no singletool. Rich pictures have been used as an ele-best way of producing a rich picture; the samement of various methods. The next sectionanalyst will use different styles under differentbriefly explains the origin of rich pictures incircumstances. To illustrate this, Figures 1 andthe Soft Systems Methodology and how they2 present rich pictures of contrasting styles.afterwardslook. The sections thereafter sketch somedrawThey depict a pub and a Web design compa-examples of how they may be used in HCI. Any, respectively. Figure 1 is intended to cap-rich picture is a small but effective idea. It canture the viewpoints ofbeer factory be incorporated into any design process.D The brewery owning and supplying thePerhaps it would be useful in yours.pub; 22i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 8D The employees that work in it; The three most important components of aD The customers that frequent it; and rich picture are structure, process, and con-D Indirectly involved stakeholders such ascerns. the community, the police, and otherD Structure refers to aspects of the worknearby areapubs in the vicinity. context that are slow to change. TheseContrast this with Figure 2, which ismight be things such as the organizationalplaceintended to capture the internal structure ofhierarchy of a firm, geographic localities,the Web design company and viewpoints ofphysical equipment, and so on. Mostthe roles within it, as well as the viewpoints ofimportant, it includes all the people whohumanexternal bodies such as clients. Figure 1will use or could conceivably be affectedimaginably emphasizes the flow of goods and servicesby the introduction of the new system. Infrom supplier to customer, whereas Figure 2Figure 1 the structure described is a brew-emphasizes the flow of influence. So, forery, owning a pub, having a landlord andexample, the Professional Society of Webcustomers, and situated in a community.Designers influences the company throughIn Figure 2 the structure includes theexpectations and standards. The directorboundaries between the company and theinfluences the work of the analyst and theworld in general and those of a given pro-coder through strategy documents, and so on. ject within the company. The analystsFigure 1 Rich Picture of a PubProfit?ConcernAm I earningenough?Capital InvestmentThe LandlordThe EmployeeManagementGoalsThe BreweryTransformationStackholderProfitsValue for MoneyQuality of FacilitiesImageEnjoymentGoods,,,

yyyzz||Noise?CashDisturbance?Convenience?,,


yyzzz|||The CustomersThe Pubtension, conflictThe Community|||CompetitionClosing time?Drunk driving?Other PubsThe PoliceBased on Patching, 1995i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 823Table 1. Elements of an Effective Rich Pictureput on her to reduce the number of staffElementCommentin her department. Someone in that1. Include structureInclude only enough structure to allowdepartment may have a concern that hisdowngrade you to record the process and con-job may be de-skilled or that he may befirecerns. The latter requires that all the laid off. The thought bubbles coding con-people who will use or could con-cerns in Figure 1 make it clear that theceivably be affected by the introduc-brewery, the employees of the pub, andtion of the new system be included.the customers each have very differentperspectives on what the pub is for. complexdetails2. Include processDo not attempt to record all the intri-very general and without many detailsFinally, tensions between stakeholders cancacies of process; a broad brush be highlighted. The “crossed swords” iconapproach is usually all that is neededserves this purpose. In Figure 1 the pub isdrawshown to be in tension with other pubs, pre-3. Include concernsCaricature the concern in a thought seeminglysumably through their competition for a lim-bubble (see Figures 1–3 for exam-collectionited pool of customers. Identifying tensionsples). A fuller explanation may be with crossed swords is a useful preliminaryprovided in a supplementary docu-step to precisely identifying the conflictingment concerns and how they may be resolved.Table 1 lists some of the features that make4. Use the language ofThis will make the rich picture com-for an effective rich picture. The first threethe people depicted in itprehensible to your informantsserve to prevent the rich picture from becom-ing overloaded with detail. The advantage of5. Use any pictorial or textual There is no correct way of drawing a device that suits your purposerich picture. There are as many styleshaving a rich picture that is comprehensible toas analysts and the same analyst will the people who have given you the informa-find different styles useful in differ-tion (Item 4 in Table 1) is that you can take itent situationsback to them for review. In this way you canobtainelicit new information and correct mistakes ofdrawing the rich picture are included ininterpretation. The discipline of using the lan-this structure to remind themselves thatguage of the work context may also help pre-they too have a separate viewpoint, con-vent the inclusion of structure, process, andcerns, and possible bias. (preference) concerns that are not real but that the analystD Process refers to the transformations thatthinks should be there. The last point in Tableoccur in the process of the work. These1 is that work context analysis requires imagi-transformations might be part of a flow ofnation and creativity, just like design itself.goods, documents, or data. In Figure 1 theExamining the examples given here and in theprocesses depicted are transformations ofreferences should provide plenty of ideas forgoods, money, and enjoyment. In Figure 2potential users of this technique.the emphasis is more on the process byThe remainder of this article illustrates thewhich different roles influence one another.role that rich pictures can play in two relatedD Concerns is the most useful component,contexts: participatory design and lightweightfor the purposes of this paper. Checklandusability engineering.calls them “issues.” We prefer the word“concern” because it captures more clearlyUses of Rich Picturesthe idea of a particular individual’s moti-Rich Pictures in Participatory DesignParticipatory design is an approach to vation for using the system. These differ-Drawing a rich picture requires that the analystdesign that attempts to actively involve ent motivations give rise to the differentwork closely with the stakeholders so that thethe end users in the design process to perspectives each person has. Each of thepictures capture the situation and related con-help ensure that the product designed meets their needs and is usable.people captured in the rich picture willcerns from the stakeholders’ points of view.have concerns. A manager might have aStakeholders participate in the process byconcern arising from the pressure beingworking with the analyst to identify structures,24i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 8==> different roles influence one anotherFigure 2 Rich Picture of Web Design ConsultancyFishy Web Inc.FISHY WEB INC.Professional Societyof Web DesignerProject TeamProfit?Long term reputation?ExpectationsStandardsboundaryDirectorResourcesNeed moreDatatimePotentialStrategyClientsDocumentsMarket ResearchAdministrationMarketingI don’t haveWorkboundaryenough timeto talk to the userIf only I hadmore powerfulWeb AnalysttoolsCompetitorConceptsCompaniesProblemsSolutionsHTML CoderMarketingGood jobdone dirt cheapFocus?Bias?preference CurrentClientsAnalystsprocesses, and concerns significant to them.that identifies the stakeholders and the workenvironmentSSM’s focus on the stakeholders’ viewpointsetting. Figures 1–3 are examples of this typeshares much with various participatory designof rich picture. | Additionally, a rich picture ofmethods [e.g., 7]. There is, however, an impor-the participatory design team itself can betant difference between participatory designused to identify the necessary managers,involving active participationand SSM: the role of the user in the designhands-on users, beneficial users, analysts,process. In participatory design the user takesdesigners, and other participants. This type ofan active role in the analysis and designrich picture can be useful in “designingprocess; in SSM this is often not the case.design,” in composing the stakeholder meet-Rich pictures can be used to record, reasonings, and in reasoning about design processes.about, communicate, and negotiate signifi-Comparing the work-context rich picturecant issues as they arise during or after partic-with the design-context rich picture provides aipatory design. Essentially the role of the richway of checking whether there is appropriatepicture is to make explicit the stakeholders,stakeholder representation on the designtheir interrelationships, and their concerns.team. Consider the use of rich pictures withInterestingly, this can be done at two levels. Athe following techniques seen frequently inrich picture of the work context can be drawnapproaches to participatory design.i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 825D Brainstorming: Brainstorming is oftenUsing a rich picture does not, in itself, solveskilfulused to generate ideas about the problemsany of the delicate problems encountered inand potential solutions of the work situa-participatory design: how to deal with privatetion. Because rich pictures can be drawnor confidential concerns, how to bring togeth-“on the fly” during a brainstorming ses-er different constituencies that have very dif-Unduly: more than is normal or reasonablesion, ideas can be captured without undulyferent ways of describing the work context, ordisrupting or constraining a necessarily cre-how to deal with minorities within a con-ative, unstructured process. Rich picturesstituency. However, constructing a rich pic-mamyhere present an alternative to the multitudeture with the help of the relevant stakeholdersSketch: preliminary drawingof sketches and doodles that participantswill make the concerns apparent, and identi-Doodle: to draw shapes, lines, or patterns without really thinking about what you are doingoften walk away with from brainstormingfying a problem is an important first step insessions. A rich picture helps everyonesolving it.involved in its construction to take a con-sistent view of the problem situation with-Rich Pictures in Lightweight out demanding that they all agree on whatUsability Methodsthe problem is. Multiple conflicting con-When people think of user interface designcerns can be captured in the pictures asthey usually think of large, high-profile pro-shown in Figures 1–3.jects, such as word processors or military com-D Storyboarding: Storyboarding is oftenmand and control systems. The majority ofused to describe the flow of, for example,user interface design projects are in fact verythe users’ activities so that they can besmall: perhaps, for example, someone hasreviewed and evaluated by both designersrequested a Windows 95 interface for someand users. Rich pictures can provide ansmall part of the company database. Anotherelegant adjunct to a connected series ofexample is the design of a Web page. The onestoryboards by representing, in a singleor two developers given the task probably doabstract summary, the major structuresnot even consider themselves user interfaceincreasingand flows, at an organizational level, rele-designers; yet cumulatively these small pro-vant to a work situation. Rich picturesjects significantly affect the productivity of anhere present a supplement to the floworganization. charts and procedural descriptions oftenOn a large project one can afford to recruitused to connect the separate episodes of aor train developers in specialized techniques;story.indeed, an elaborate, well-specified designMock-up: a full-size model of something large that has not yet been built, which shows how it will look or operate, D Paper-Based Prototyping: Many partici-methodology may be necessary just to manageor which is used when the real thing is not neededpatory design techniques use paper-basedthe large number of personnel involved [11].mock-ups and prototypes to repre-On a small project the techniques used mustsent design ideas early in thebe “lightweight,” that is, the costs to the orga-development process [e.g.,nization must be minimal. Nielsen [16] hasname14]. Such techniques pro-dubbed these techniques “discount.” Theyvide a way for stakehold-may only achieve 90 percent of what is possi-ers to comment on theble with more elaborate methods, but they dodetails of the design andso for very much less than 90 percent of thethe extent to which itcost. Costs here are measured in training andmeets the user’s charac-in the time it takes to apply the technique;teristics and needs. Intherefore, lightweight techniques have to becapturing the primaryeasy to learn and quick to apply. If a project isconcerns of the usersassigned only 4 person-weeks of effort, a tech-and, potentially, the major informationnique for improving some aspect of the quali-flows likely to affect the system, the richty of a user interface is unlikely to justify morepicture places the emerging design in itsthan 1 day of training and 2 or 3 days of appli-overall social and technical context.cation. Examples of lightweight techniquesi n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 8include Monk et al.’s simplified user testingdesigner can develop common ground byprocedure Cooperative Evaluation [13] andfocusing on actions and tasks. A rich pictureNielsen’s simplified usability inspection tech-can serve a similar communicative functionnique, Heuristic Evaluation [17]. With thesemuch earlier in design when one is thinkingtechniques, prototypes and scenarios are cru-about the general work context and the con-cial parts of communication between designerstraints this imposes.and user. Without these concrete representa-Monk [12] describes how a rich picture cantions of the design, little communication canbe used as the first step in a lightweight designoccur. With them, however, both user andprocess, to reason about the redesign of thetoo many people PFigure 3. Rich Picture of aCold Storage Warehousei n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 827work context that will be required. He sug-middle of a large sheet of paper some figureoperator : SO use and control a machine or vehicle gests that “before” and ”after” rich pictures bewho represents the primary user or operator.developed. The former records critical aspectsMonk’s lightweight technique is to encourageof the work context as it now exists, and theuser centered-design and to avoid the naturallatter illustrates how the contexttendency for developers to take a system-ori-will change when the new sys-ented view. Putting the operator at the centertem is introduced. The beforeof the picture makes her the focus of atten-picture can be presented totion. Next, the stakeholders that directlyone’s informants to checkinfluence the operator’s work can be picturedthat the analysis does notalong with the elements of structure needed tomisinterpretmisconstrue orexplain the process of work. Monk illustratesomit crucial fac-his methods with a real example of design fortors. If more thana cold storage warehouse; the rich pictureone developer isdeveloped in this process is shown in Figure 3.working on the project,The operator is given a fictitious name, Jenny.the before picture can also be invaluable inJenny’s job involves taking delivery notes fromcommunication between developers and get-the drivers of vehicles bringing goods into theting everyone to think on the same wave-warehouse (depicted as stick figures wearinglength. As the prototype design is developedcaps). Jenny enters the data from the deliverytally: a record or count of a number of itemsan after picture will emerge. The after picturenote into a computer system to provide tallycan be presented to management to alert themlists for the deliveries to be checked by theto the implications of the new computer sys-warehouse men (signified by stick figurestem. If they are unhappy it is still early enoughwearing black hats). The roles described thusfor changes to be made. If they accept thefar then are the core stakeholders in thedesign then they can make appropriate adjust-process of getting the work done. The richments, change reporting structures, organizepicture also identifies other clerks and a super-externalretraining, and so on.visor. Finally, the most peripheral stakeholdersThe procedure suggested by Monk firstare drawn in. In Figure 3 they appear at theinvolves talking to stakeholders about theirtop edge of the picture. They are the directorsjobs. A rich picture can be a useful way forand computer systems people of the two orga-developers who are not used to this sort ofnizations taking part of this operation—thework to focus their thoughts. Normally a des-owners of the cold store and the owners of theintendedignated contact in the user organization willstores supplied.be interviewed first. The people who will endWhen the major structures and processesup actually using the system should also behave been added, the concerns can beinterviewed. It is then a matter of judgmentaddressed. The thought bubble for Jenny inhow many of the additional stakeholders iden-Figure 3 codes the wide variations in workloadtified by these initial informants one alsoshe has to put up with. Other concernsneeds to talk to. It is always a good idea toincluded are the need for the drivers to getinterview people in their workplace, whereaway as soon as possible, worries about jobthey can show you documents, screens, and sosecurity, and so on. Thought bubbles may besecreton. A portable tape recorder may be useful tosomewhat cryptic to someone who was notcheck what was said, and you should alwaysinvolved in generating a rich picture, so Monkhave a prepared list of topics or interviewsuggests that an additional sheet be addedschedule so that you cover all the criticalexplaining in slightly more detail the concernspoints. Clegg et al. [5] give useful and practi-of each stakeholder. The same sheet maycal advice on how to get the best out of yourexplain the process and specific responsibili-informants.ties not coded on the picture. When drawing a rich picture for this pur-One of the important reasons for drawingpose, you normally start by sketching in thea rich picture is to clarify one’s thoughts. For28i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 8this reason one should not be afraid to throwal interviews, “train me” sessions, work log-away an early version and start again. Richging, semistructured interviews, scenariopictures can also be presented to informantsanalysis, model building, wish lists, and(although you may need different versions forassumption challenging. They argue that noMETHODS & TOOLSa small change, improvementdifferent informants) to make amendments orsingle technique is capable of capturing fullCOLUMN EDITORSimportantradical revision. The rich picture is only thethe diversity of the work setting. Michael Mullerfirst step in Monk’s lightweight method. TheDearden and Wright draw an interestingMicrosoft Corporationnext is to identify work objectives and userdistinction between techniques that are situat-One Microsoft Wayexceptions, which are then used to developed in the work context and those that goRedmond, WA 98052scenarios that can be used to refine early pro-beyond the immediate situation. The formermullerm@acm.orgtotype designs and to make sure that thetechniques can be used only in the work place.design supports all relevant aspects of theThe latter allow the analyst and the user toFinn Kensingwork. The rich picture serves as a startingdetect issues beyond the range of the observ-Department of point and a context for all these activities.able situation, for example, the organizationalComputer ScienceReaders wishing to know more about thisand historical contexts. Dearden and WrightRoskilde Universityprocess should consult Monk [12].assert that different techniques have differentBuilding 20.2 Monk’s lightweight method is a relativelystrengths and weaknesses. Observation allowsP.O. Box 260 informal technique; that is, it is not preciselyone to separate what people say they do fromDK 4000 Roskildespecified. This has the advantage of making itwhat they really do, but it has practical limita-Denmarkrelatively easy to learn and apply. The disad-tions. With only a limited amount of time in+45-4674-2548vantage is that different people will apply it inthe workplace it may be impossible to see theFax: +45-4674-3072different ways. This is not a problem when thefull process. Infrequent, but nonethelesskensing@ruc.dkdesign team is small and coordination isimportant, problems may not crop up whilestraightforward. However, when design teamsyou are actually there. Only by using a varietyget larger a much more precisely specifiedof situated and nonsituated techniques canmethod is needed, just so that everyone knowsthe fullest account emerge, given the prevail-succeedwhat everyone else is doing [11]. Examples ofing practical constraints. The rich picture canmore tightly specified procedures that makeserve as a representation to motivate all theseuse of rich pictures aredifferent sources of information about theD TheoryBuilder [10, 19]; work. It can also serve as a representation toD Howard and Smith’s [8] use of rich pic-integrate information regarding the highertures with Johnson’s [9] Knowledgelevel work context coming fromAnalysis of Tasks; and the sources.D Multiview [2].The versatility of the richpicture arises from its sim-Some Final Commentsplicity. We suspect thatOne recurring theme in this rrepeatingeview has beenmany readers willthe value of using more than one techniquealready have seen ways ofwhen analyzing a work context. We are notincorporating rich pic-suggesting that using a rich picture will solvetures into their own meth-all your problems. It is just one of many smallods and we would encouragebut useful ideas that may be applied to anythem to do so. The foregoingdesign problem. The value of using a wideexamples of good practicevariety of techniques is eloquently discussedshould allow you to do thisby Dearden and Wright [6]. They report on aeffectively. Perhaps one daycase study that borrowed from a number ofthe rich picture will be as familiarmethodological traditions to analyze a worka diagram to see at a design meeting as thewidespreadcontext. According to this approach an SSM-now ubiquitous data flow diagrams and flowstyle rich picture is just one of the techniquescharts. When that day arrives we will haveused. Dearden and Wright also used contextu-moved much further toward removing ouri n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 829blinders and making a genuine attempt to seeLondon, 1992.the other person’s point of view. System design10. Khushalani, A., Smith, R., and Howard, S. Whatcan only benefit from such a change.happens when designers don’t play by the rules: towardsa model of opportunistic behaviour in design.AcknowledgmentsAustralian Journal of Information Systems 1, 2 (1994),Dr. Monk was supported by the ESRCpp. 13–31.Cognitive Engineering Program. Dr. Howard11. Kraut, R. E., and Streeter, L. A. Coordination inwas supported by a grant from Swinburnesoftware development. Communications of the ACM 38,University of Technology, Australia. We3 (1995), pp. 69–81.would like to thank the editors of this section12. Monk, A. F. Lightweight techniques to encouragefor valuable comments on an earlier manu-innovative user interface design. In L. Wood & R.script. Zeno, eds., Bridging the Gap: Transforming UserRequirements into User Interface Design. CRC Press,References Boca Raton, 1997.1. Avison, D. and Fitzgerald, G. Information Systems13. Monk, A. F., Wright, P., Haber, J.,Development: Methodologies, Techniques andand Davenport, L. ImprovingTools. Blackwell Scientific Publishers, Oxford,your human-computer1988.interface: a practical tech-2. Avison, D. and Wood-nique. BCS PractitionerHarper, T. MultiviewSeries. Prentice-Hall,Methodology. BlackwellHemel Hempstead, 1993. Scientific Publishers,14. Muller, M. J. PICTIVE—An exploration in partici-Oxford, 1990.patory design. In S. P. Robertson, G. M. Olson, and J.3. Checkland, P. SystemsS. Olson, eds., CHI’91 Human Factors in ComputerThinking, Systems Practice. JohnSystems (New Orleans, 1991). ACM, New York, pp.Wiley and Sons, Chichester, 1981.225–231.4. Checkland, P. and Scholes, J. Soft15. Mumford, E. Sociotechnical system design: evolvingSystems Methodology in Action. John Wiley and Sons,theory and practice. In G. Bjernes, P. Ehn, and M.Chichester, 1990.Kyng, eds., Computers and Democracy: A Scandinavian5. Clegg, C., Warr, P., Green, T., Monk, A., Kemp, N.,Challenge. Avebury, Aldershot, UK, 1987.Allison, G., and Lansdale, M. People and Computers:16. Nielsen, J. Usability Engineering at a Discount. InPERMISSION TO MAKE DIGITAL/ HARDHow to Evaluate Your Company’s New Technology. EllisG. Salvendy and M. J. Smith, eds., Proceedings of theCOPY OF PART OR ALL OF THIS WORKHorwood, Chichester, UK, 1988. Third International Conference on Human-ComputerFOR PERSONAL OR CLASSROOM USE IS6. Dearden, A. and Wright, P. Experiences UsingInteraction, HCI-International ‘89, Boston, September,GRANTED WITHOUT FEE PROVIDEDSituated and Non-situated Techniques for StudyingElsevier Science, 1989, pp. 394–401.THAT COPIES ARE NOT MADE ORWork in Context. In S. Howard, J. Hammond, and G.17. Nielsen, J. and Mohlich, R. Heuristic evaluation ofDISTRIBUTED FOR PROFIT OR COM-Lindgaard, eds., Human Computer Interaction—INTER-user interfaces. In J. C. Chew and J. Whiteside, eds.,MERCIAL ADVANTAGE, THE COPYRIGHTACT’97. Chapman and Hall, London, 1997.Human Factors in Computer Systems, CHI’90. CHI ’90,NOTICE, THE TITLE OF THE PUBLICA-7. Greenbaum, J. and Kyng, M. Design at Work:Seattle, April, ACM, New York, 1990, pp. 249–256.TION AND ITS DATE APPEAR, ANDCooperative Design of Computer Systems. Lawrence18. Patching, D. Practical Soft Systems Analysis. London,NOTICE IS GIVEN THAT COPYING IS BYErlbaum Associates, Hillsdale, NJ, 1991.Pitman Publishing, 1990.PERMISSION OF ACM, INC. TO COPY8. Howard, S. and Smith, R. Using the Soft Systems19. Smith, R., Howard, S., Sutherland, T., andOTHERWISE, TO REPUBLISH, TO POSTMethodology to Front-end Task Analysis. In HCI: AKhushalani, A. TheoryBuilder: A Behavioural PerspectiveON SERVERS, OR TO REDISTRIBUTE TOLight Into the Future, Proceedings of OZCHI’95on Modelling and Improving Systems Development.LISTS REQUIRES PRIOR SPECIFIC PERMIS-(Wollongong, Australia, November 1995), pp. 88–94. Proceedings of First Australian Seminar on Modelling andSION AND/OR A FEE. 9. Johnson, P. Human Computer Interaction: Psychology,Improving Systems Development. School of Information© ACM 1072-5220/98/0300 $3.50 Tasks Analysis and Software Engineering. McGraw-Hill,Technology, Swinburne University, 1994.30i n t e r a c t i o n s . . . m a r c h + a p r i l 1 9 9 8