Åhörarkopior Anändarcentrerad system: Föreläsning 2 Anändarcentrerad system Om i bara lägger till anändbarhet, kommer allt att bli bra då? Definition och nyckelprinciper Jan Gulan Gulliksen Adelningen för MDI/IT, Uppsala Uniersitet, Serige Jan.Gulliksen@hci.uu.se Bengt Göransson Enea Redina AB och Adelningen för MDI/IT, Uppsala Uniersitet, Serige Bengt.Goransson@enea.se Niklas Johansson Adelningen för MDI/IT, Uppsala Uniersitet, Serige Nej, i behöer en anändarcentrerad systemprocess Niklas.Johansson@hci.uu.se http://www.hci.uu.se/edu Vilka är era anändare? Anändaränlig IT Vill i att datorerna skall ara snälla Vilka bra frågor du ställer mot oss? Våra anändare är alla Våra anändare är i medeltal kinna på Anändbar IT 58 år utan eftergymnasial utbildning Våra system skall fungera för blinda, döa, icke läskunniga Vad innebär det att något är anändbart? Tillfredsställande Lätt att lära Lärorikt Förlåtande Enkelt Effektit Inspirerande Felkorrigerande Själförklarande Kräer ingen manual Intuitit Läcker Omedelbart Definition of Usability Snyggt Billigt Nyttigt Vem som helst kan anända det ISO 9241 Part 11 Guidance on Usability Ändamålsenligt The extent to which a product can be used by specified users to achiee specified goals with effectieness, efficiency and satisfaction in a specified context of use. Farmor kan anända den! Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 1
Åhörarkopior Anändarcentrerad system: Föreläsning 2 Attributes of system acceptability Anändbarhetsmål - Mätbara System acceptability Social acceptability Practical acceptability Usefulness Cost Compatibility Reliability Etc. Usability Utility Easy to learn Efficient to use Easy to remember Few errors Subjectiely pleasing Jag skall kunna registrera en ny kund på mindre än 20 sekunder. Systemet skall inte hänga sig mer än en gång om dagen. Jag skall inte göra mer än ett fel i timmen. Systemet skall få minst 7 på en tiogradig skala I anändartillfredsställelse From Nielsen (1993) Usability Engineering CHAOS-rapporten I USA spenderades 250 miljarder dollars arje år på 175 000 olika IT-projekt. 365 IT-företag med 8380 olika IT-projekt analyserades 1995. 31,1 % a företagens projekt abröts. 52,7 % utfördes med förändrade planer 16,2 % utfördes enligt plan. I snitt ökade kostnaderna för de förändrade planerna med 189 %. 81 miljarder dollars spenderas arje år på projekt som aldrig leder till några resultat. Källa: Standish Group, 1995 (www.standishgroup.com) Arbetsmiljölagen (Kap.2, 1) Arbetsmiljön skall ara tillfredsställande med hänsyn till arbetets natur och den sociala och tekniska utecklingen i samhället Arbetsförhållandena skall anpassas till människors olika förutsättningar i fysiskt och psykiskt aseende. Arbetstagaren skall ges möjlighet att mederka i utformningen a sin egen arbetssituation samt i förändrings- och utecklingsarbete som rör hans eget arbete. Teknik, arbetsorganisation och arbetsinnehåll skall utformas så att arbetstagaren inte utsätts för fysiska eller psykiska belastningar som kan medföra ohälsa eller olycksfall. Därid skall äen löneformer och förläggning a arbetstider beaktas. Starkt styrt eller bundet arbete skall undikas eller begränsas. Det skall eftersträas att arbetet ger möjlighet till ariation, social kontakt och samarbete samt sammanhang mellan enskildas arbetsuppgifter. Det skall idare eftersträas att arbetsförhållandena ger möjlighet till personlig och yrkesmässig uteckling liksom till själbestämmande och yrkesmässigt ansar. Föreskrifter för arbete id bildskärm Arbete id bildskärm som är starkt styrt eller bundet i fysisk eller psykiskt aseende eller är ensidigt upprepat får normalt inte förekomma. (AFS 1998:5 7) Forskningsproblematiken Traditionella MDI-metoder fokuserar på produkten, t ex usability engineering Vi ill i stället lägga fokus på processen för att kritiskt granska denna Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 2
Åhörarkopior Anändarcentrerad system: Föreläsning 2 Att uteckla anändbara system handlar om att tillämpa anändarcentrerad system Vad är sanning? Hur kan man eta om en modell är bättre än en annan? En gemensam terminologi Vad är en Metod implicerar ett systematiskt, repeterbart sätt att utforma, att göra något enligt en plan Modell är ett system som i själa alt just för att den abildar äsentliga egenskaper hos ett annat system Process är en strukturerad serie a händelser med ett specifikt mål att åstadkomma ett isst resultat inom en iss tid Technology-drien s. user-centered systems Hur sker systemuteckling idag? Technology-drien philosophy Technology/deeloper-drien Component focus Indiidual contribution Focus on internal architecture Quality measured by product defects and performance (system uality) Implementation prior to usability alidation Solutions are directed by functional reuirements User-centered systems philosophy User-drien Solution focus Multidisciplinary teamwork including users, deelopers, customers, usability experts Focus on usability attributes: effectieness, efficiency and satisfaction Quality defined by usability (uality in use) Implementation of user-alidated solutions only Understanding the context of use: user, work task, work enironment Tidspress, kostnadspress, resursbrist, Anändbarhet är inte något särskilt iktigt i konkurrens med andra mål, typ säkerhet, stabilitet, Det finns inte särskilt mycket kompetens om anändbarhet Anändbarheten angår alla, från managementniå till programmerarna Olika utecklingssammanhang (konsult) Anändarna identifierade Utecklarna kända A case study Utecklarna kända Anändare identifierade Kontraktsuteckling Produktuteckling In-house - uteckling Anändarna identifierade Utecklarna kända Tid Systemuteckling Projektstart Systemleerans Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 3
Åhörarkopior Anändarcentrerad system: Föreläsning 2 A case study of the rise and fall of a UCSD project Initial set of principles for UCSD The pilot project was an in-house deelopment project within the Swedish National Tax Board. The purpose of the project was to deelop a new computerized casehandling tool for administrators working with national registration. The project had an explicit goal to apply a UCSD approach. Based on preious experiences simply adding usability engineering methods to existing processes did not make any difference We established a set of UCSD principles for the project to follow. Work practices of the users control the deelopment. Actie user participation throughout the project. work domain experts (continuously through the deelopment project) and actual end-users (for ealuation of arious results) Early prototyping to ealuate and deelop solutions. Continuous iteration of solutions. Multidisciplinary teams. Integrated. Actiities Problems Usability er. User and task analysis. Personas Collaboratie prototyping with users. Quality assurance plan. Little or no lifecycle perspectie. Usability ers were ignored. Use case mania. Little understanding of the documentation. Changing technical platform. Problems establishing a user-centered attitude. Implementation support. Lessons learnt User-Centered Systems Design UCSD is not commonly understood, and there is no clear definition. One needs to be ery specific about what it takes from the process to comply with UCSD. Commercial products don t gie us much help. We had to modify our principles for UCSD. All participant s and stakeholders support is crucial. UCSD is as strong as its weakest link. It s all about attitude Donald Norman wrote back in 1986: But user-centered emphasizes that the purpose of the system is to sere the user, not to use a specific technology, not to be an elegant piece of programming. The needs of the users should dominate the of the interface, and the needs of the interface should dominate the of the rest of the system. Donald Norman, Cognitie Engineering, in D. A. Norman & S. W. Draper (eds.), User Centered System Design, 1986 Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 4
Åhörarkopior Anändarcentrerad system: Föreläsning 2 Norman s modell E.g. John Karat s iew on UCD Designerns modell Designer Dokumentation System Systembild Anändarens modell Anändare For me, UCD is an iteratie process whose goal is the deelopment of usable systems, achieed through inolement of potential users of a system in system. Karat, J. (1996) User Centered Design: Quality or Quackery?, in the ACM/SIGCHI magazine, Interactions july+august 1996. I suggest we consider UCD an adeuate label under which to continue to gather our knowledge of how to deelop usable systems. It captures a commitment the usability community supports that you must inole users in system while leaing fairly open how this is accomplished. Karat, J. (1997), Eoling the Scope of User-Centered Design, in Communications of the ACM, Vol. 40, No 7, pp. 33-38. Lack of common definition of UCSD User-Centered Systems Design is a process Usability engineering (Nielsen, Mayhew) Human-centered (ISO 13407) Goal-directed (Cooper) Usage-centered (Constantine & Lockwood) Contextual (Wixon & Holtzblatt) Customer-centered (Beyer & Holtzblatt) Cooperatie (Scandinaian School) (Greenbaum & Kyng) Participatory (Muller, Haslwanter & Dayton) Feedback with suggestions for changes. Ealuation in context. Measure against usability goals and criteria for. Actie users, usability focus and iteratie Analysis of users, actiities, tasks and context of use. Design suggestions using prototypes, an iteratie and creatie process in itself. Usability People tend to beliee that usability is something that can be added on. That is not true! The usability of a system is defined as: Principles / Best practices / Success factors A principle is a commonly accepted fundamental rule or law from which other principles can be deried. We hae identified a set of key principles or best practices for UCSD. The extent to which a product can be used by specified users to achiee specified goals with effectieness, efficiency and satisfaction in a specified context of use. ISO 9241-11 Guidance on usability Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 5
Systemförutsättningar Planer och mjukarukra System Inkrement 1 Detaljerad Kodning Integration Införande Drift och underhåll Inkrement 2 Detaljerad Inkrement n Kodning Detaljerad Integration Kodning Införande Integration Drift och underhåll Införande Drift och underhåll Systemet i drift Inkrement 1 Inkrement 2 Inkrement 3 Inkrement n Åhörarkopior Anändarcentrerad system: Föreläsning 2 Background to the principles for UCSD ISO 13407 Human-centered processes for interactie systems IBM research: Gould, Boies & Ukelson (Lewis) 1983, 1985, 1988, 1997 Early and continual focus on users. Empirical measurement. Iteratie. Integrated wherein all aspects of usability eole together. identify need for human-centred understand & specify the context of use ISO 13407 ISO 13407, Human-centered processes for interactie systems. Actie inolement of users and a clear understanding of user and task reuirements. Appropriate allocation of function between users and technology. Iteration of solutions. Multi-disciplinary. ealuate s against reuirements system meets specified functional, user & organisational reuirements produce solutions specify user & organisational reuirements Definition and Key Principles for User-Centered Systems Design User-centered systems is a process focusing on usability throughout the entire deelopment process and further throughout the system life cycle. It is based on the following key principles. User focus the goals of the actiity, the work domain or context of use, the user s goals, tasks and needs should early guide the deelopment. Actie user inolement representatie users should actiely participate, early and continuous throughout the entire deelopment process and throughout the system lifecycle. Eolutionary systems deelopment the systems deelopment should be both iteratie and incremental. Simple representations the must be represented in such ways that they can be easily understood by users and all other stakeholders. Prototyping early and continuously, prototypes should be used to isualize and ealuate ideas and solutions in cooperation with the end users. Ealuate use in context baselined usability goals and criteria should control the deelopment. Explicit and conscious actiities the deelopment process should contain dedicated actiities. A professional attitude the deelopment process should be performed by effectie multidisciplinary teams. Usability champion usability experts should be inoled early and continuously throughout the deelopment lifecycle. Holistic all aspects that influence the future use situation should be deeloped in parallel. Processes customization the user-centered systems process must be specified, adapted and/or implemented locally in each organization. A user-centered attitude should always be established. User focus The goals of the actiity, the work domain or context of use, the users goals, tasks and needs should early guide the deelopment. All members of a project must understand the goals of the actiity, the context of use, who the users are, their situation, goals and tasks, why and how they perform their tasks, how they communicate, cooperate and interact, etc. This helps in creating and maintaining a focus on the users needs instead of a technical focus. Actiities, such as identifying user profiles, contextual inuiries and task analysis, must be a natural part of the deelopment process. Make sure that all project members hae met real or potential users, for instance, by isiting the workplace. Descriptions of typical users, tasks and scenarios could, for instance, be put up on the walls of the project room/area to maintain the user focus. Actie user inolement Eolutionary systems deelopment Representatie users should actiely participate, early and continuously throughout the entire deelopment process and throughout the system lifecycle. The users should be directly inoled, both in the deelopment project and in related actiities, such as, organizational deelopment and the deelopment of new work practices. The users must be representatie of the intended user groups. Specify where, when and how users should participate in the deelopment. Identify appropriate phases for user participation and specify their characteristics. Emphasize the importance of meeting the users in context, for instance, at their workplace. Note the important difference between domain experts (ery familiar with the actiity, but not necessarily potential users) and real users. Domain experts can be members of the deelopment team but real users should be inoled on a temporary basis, in single actiities during the analysis, and ealuations of solutions. The systems deelopment should be both iteratie and incremental. It is impossible to know exactly what to build from the outset. Hence, UCSD reuires an approach which allows continuous iterations with users and incremental delieries so that suggestions and solutions can be ealuated by the users before they are made permanent. An iteration should contain a proper analysis of the users needs and the context of use, a phase, a documented ealuation with concrete suggestions for modifications and a re in accordance with the results of the ealuation. These actiities do not hae to be formal. An iteration could be as short as half an hour, as long as it contains all three steps. Incremental deelopment means that, based on an oerall picture of the System Under Deelopment (SUD), priorities are set and the system is diided into parts that can be deliered for real use. Each increment is iterated until the specified goals hae been met. Ealuations of the increments in real use should influence the of the subseuent increments. Let the software grow into the final product. Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 6
Åhörarkopior Anändarcentrerad system: Föreläsning 2 Simple representations The must be represented in such ways that it can be easily understood by users and all other stakeholders. ß Use representations and terminology that are easily understood by all users and stakeholders so that they can fully appreciate the conseuences of the on their future use situation. Use, for instance, prototypes (sketches and mock-ups) and simulations. Prototyping Early and continuously, prototypes should be used to isualize and ealuate ideas and solutions in cooperation with the end users. ß Use multiple paper sketches, mock-ups and prototypes to support the creatie process, elicit reuirements and isualize ideas and solutions. The prototypes should be ealuated with real users in context. ß Abstract notations, such as use cases, UML diagrams or reuirements specifications are not sufficient to gie the users and stakeholders a concrete understanding of the future use situation. ß It is essential to start with low-fidelity materials, for instance, uick sketches, before implementing anything in code. ß The representations must also be usable and effectie. The goal is that all parties inoled share the same understanding of what is being built. ß Start with the conceptual on a high leel and do not moe on to detail too uickly. If possible produce seeral prototypes in parallel. ß Work with prototypes in cooperation with the users in context (contextual prototyping). Ealuate use in context Baselined usability goals and criteria should control the deelopment. ß Critical usability goals should be specified and the should be based on specific criteria. Ealuate the against the goals and criteria in cooperation with the users. ß Early in the deelopment project, one should obsere and analyze the users reactions to paper sketches and mock-ups. Later in the project, users should perform real tasks with simulations or prototype tasks and their behaior and reactions should be obsered, recorded and analyzed. ß Be sure to specify goals for aspects that are crucial for the usability and that coer critical actiities as well as the oerall use situation. Explicit and conscious actiities The deelopment process should contain dedicated actiities. ß The user interface and the interaction are of undisputed importance for the success of the system. Far too often, the UI and interaction happens as a result of somebody doing a bit of coding or modeling rather than being the result of professional interaction as a structured and prioritized actiity. ß Remember that to users the user interface is the system. The of the SUD as regards the user interaction and usability should be the result of dedicated and conscious actiities. The construction of the SUD should adhere to that. A professional attitude The deelopment process should be performed by effectie multidisciplinary teams. Usability champion Usability experts should be inoled early and continuously throughout the deelopment lifecycle. ß Different aspects and parts of the system and deelopment process reuire different sets of skills and expertise. ß An experienced usability expert (usability er) or possibly a usability group, should be on the deelopment team. ß The analysis, and deelopment work should be performed by empowered multi-disciplinary teams of, for instance, system architects, programmers, usability ers and interaction ers. ß The usability er should be deoted to the project as an engine for the UCSD process from the beginning of the project and throughout the lifecycle. ß A professional attitude is reuired and so are tools that facilitate the cooperation and efficiency of the team. ß The usability er must be gien the authority to decide on matters affecting the usability of the system and the future use situation. Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 7
Conceptual Interaction [ More iterations needed ] [ Goals not met? ] Plan the usability process Contextual usability analysis Detailed Formal usability ealuation User feedback Actie user participation User assistance Åhörarkopior Anändarcentrerad system: Föreläsning 2 Holistic Processes customization All aspects that influence the future use situation should be deeloped in parallel. Software does not exist in isolation from other parts of, for instance, a work situation. When deeloping software for the support of work actiities, the work organization, work practices, roles, etc, must be modified. All aspects should be deeloped in parallel. This includes work/task practices and work/task organization, user interface and interaction; on-line help; manuals; user training, work enironment, health and safety, etc. Other parts of the context of use such as: hardware euipment, social and physical enironments, etc. must also be considered in the integrated process. The responsibility should be placed with one person or a team. The UCSD process must be specified, adapted and/or implemented locally in each organization. UCSD and usability cannot be achieed without a usercentered process. There is, howeer, no one-size-fits-all process. Thus the actual contents of the UCSD process, the methods used, the order of actiities, etc, must be customized and adapted to the particular organization and project based on their particular needs. A UCSD process can be based on a commercial or in-house software deelopment process, where actiities are added, remoed or modified. Existing methods and techniues may well be re-used, if they comply with the aboe factors/practices. A user-centered attitude User-Centered Systems Design is a process A user-centered attitude should always be established. UCSD reuires a user-centered attitude throughout the project team, the deelopment organization and the client organization. All people inoled in the project must be aware of and committed to the importance of usability, but the degree of knowledge on usability may differ depending on role and project phase. The key principles defined in this paper can sere as a common ground. Feedback with suggestions for changes. Ealuation in context. Measure against usability goals and criteria for. Actie users, usability focus and iteratie Analysis of users, actiities, tasks and context of use. Design suggestions using prototypes, an iteratie and creatie process in itself. Definition and Key Principles for User-Centered Systems Design User-centered systems is a process focusing on usability throughout the entire deelopment process and further throughout the system life cycle. It is based on the following key principles. User focus the goals of the actiity, the work domain or context of use, the user s goals, tasks and needs should early guide the deelopment. Actie user inolement representatie users should actiely participate, early and continuous throughout the entire deelopment process and throughout the system lifecycle. Eolutionary systems deelopment the systems deelopment should be both iteratie and incremental. Simple representations the must be represented in such ways that they can be easily understood by users and all other stakeholders. Prototyping early and continuously, prototypes should be used to isualize and ealuate ideas and solutions in cooperation with the end users. Ealuate use in context baselined usability goals and criteria should control the deelopment. Explicit and conscious actiities the deelopment process should contain dedicated actiities. A professional attitude the deelopment process should be performed by effectie multidisciplinary teams. Usability champion usability experts should be inoled early and continuously throughout the deelopment lifecycle. Holistic all aspects that influence the future use situation should be deeloped in parallel. Processes customization the user-centered systems process must be specified, adapted and/or implemented locally in each organization. A user-centered attitude should always be established. Actiity lists The purpose of the actiity list that accompanies each principle is to elaborate on what it takes to apply a principle. The actiity list suggests actiities of a general nature alongside appropriate methods or tools for the achieement of the principle. The principles are general but the actiity list should be deeloped specifically to fit each organization. The actiity list seres as both a To-do list and a checklist, where each item can be ticked off. There are three options for each actiity: No = we decided to not perform this actiity. We gae rationales for this decision and had a general agreement on the moties. Yes = we performed this actiity, in full or to the extent that the project team and management, found appropriate. N/A = we found that this actiity was not applicable. The rationales for this were clearly stated and agreed on. We hae conducted other acties to compensate for this. Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 8
Åhörarkopior Anändarcentrerad system: Föreläsning 2 Example of actiity list for User focus Vision, purpose goal and constraints of the target actiity analyzed and understood by all project members. Tools and methods: Goals analysis, Focus groups Identification, description and prioritization of all user groups. Tools and methods: User analysis Visualization and characteristics of target user groups made aailable to eeryone in the project. Tools and methods: Decorate a project room with artifacts, etc. that illustrates the users work situation, enironment and characteristics. Potential limitations and restrictions in the users capabilities (for instance ision impairments or language problems) are clear to eeryone in the project. Target user groups hae guided the deelopment.. continues. example of actiity list for User focus. continued. The users hae expressed their impressions of current system and expectations on future system. Tools and methods: Users asked about good things and bad things in their current work situation, Think-out loud. Users obsered as they were performing their tasks in context. Tools and methods: Analysis of information utilization, Context-of-use analysis, Field studies, Contextual inuiry. Use situation documented Tools and methods: Video and still camera, scenarios, personas Tasks analyzed. Tools and methods: Task analysis Copies of artifacts (forms, documents archies, notebooks, etc.) used by the users collected. Application of the principles Explanation model to analyze and communicate why organizations, projects or processes did not meet their goals as regards usability Process deelopment for defining a UCSD process Process / organization customization to customize or adapt an organization, project or deelopment process to UCSD Process / organization assessment to assess the user-centeredness of an organization, project or process Knowledge transfer to teach and transfer knowledge about UCSD, Procurement support support for procurers as a basis for specifying reuirements on the process as such Using the principles in practice These 12 principles ensure a truly user-centered approach and generate seeral benefits. They facilitate communicating, assessing and deeloping processes for the analysis,, ealuation, construction and implementation of an interactie system. They help in maintaining the focus on the users and the usability throughout the entire deelopment process. We fully appreciate that it will be more or less impossible to start applying the principles in one strategic shift. Adopting them gradually is probably more feasible and practicable. It is, howeer, important to comply with the principles to as large an extent as possible, at any point in time. What you can do Decide on how important usability is. You specifically hae to address usability aspects to achiee a usable system. A specific and explicit approach and philosophy is needed user-centered. Include user-centered actiities and foster a common understanding among all stakeholders about the importance of usability. Ensure actie user participation. Jan Gulliksen & Bengt Göransson, 2002 http://www.hci.uu.se/acsd/ 9