Part 3: Design and Research Intertwined

Chapter List

Chapter 6: Managing the Design of User Interfaces

Chapter 7: The One-Row Keyboard-A Case Study in Mobile Text Input

Chapter 8: Series 60 Voice Mailbox in Rome-A Case Study of a Paper Prototyping Tour

Chapter 9: 1⁄4 Samara-Hardware Prototyping a Driving Simulator

Chapter 10: Dawn on the Wireless Multimedia Highway

Chapter 11: Just-In-Time Usability Engineering

The stories you'll read in this part of the book are real-world examples of mobile UI design challenges, such as squeezing text input tools into minuscule spaces, localizing user interfaces (UIs) hand-in-glove with UI-style development, devising a way to research complex human-machine interactions for users who make phone calls and drive cars at the same time, and finding out what customers dislike most about using mobile services. The projects and the problem solving that project teams actually do will also give you a new appreciation of how many corporate functions contribute to usability engineering.

Chapters 6 to 11 reveal the range of objectives, approaches, and points of view espoused by usability and interaction design professionals at Nokia. Usability as a corporate function includes focused and comprehensive problem solving, creative design, hard decision making, leaps of faith, and thoughtful management to facilitate the work. It takes a judicious selection from the approaches listed below-and sometimes all of them-to make a mobile project cohere.

Engineered. Theory is both the starting point for new hypotheses and the framework where novel concepts can be accommodated and pursued. It is the momentum behind cumulative progress in science. Interaction design and usability engineering projects place a lot of weight on designing and engineering solutions and validating those solutions with user tests, but analyzing the phenomena behind the results gets less attention. Industrial usability research does not focus on turning data into theories, at least not on the basis of our experience. Getting something to work is our primary method of learning why it actually functions-or why it doesn't. An obvious exception to this rule occurs when the lessons learned can be immediately iterated and turned into functional solutions to be applied in the next prototype version or in the next product. Usability operates primarily in a constructive engineering mode. Why is that so? Isn't it self-evident that understanding the underlying principles of human-machine interaction would enhance our chances of success in the following rounds of design? In UI research oriented to practical product development, we encounter numerous variables related to human cognition, motivation, usage context, technology, and so on that may influence user experience. They can-in the best case-be identified, but to define each variable's role in a specific problem is beyond the scope of present understanding. Miika Silfverberg's report (Chapter 7) is a good example of the dilemma. He presents well-justified assumptions in the beginning of the project, but even so is surprised by what turns out to be the biggest issues as revealed in users' actual behavior when interacting with a prototype.

Intuitive. Usability organizations must ensure that usability and UI design knowledge is captured and catalogued, but the case-dependent nature of the discipline makes that difficult. Without cumulative theories-or professional conventions and rules, to be more modest-the people involved are the best conduits of knowledge in the lab. Their experiences with successful and failed projects turn into a kind of tacit knowledge or developed intuition. They become experts. They will gradually be able to contribute more focused studies, more rewarding new initiatives, or more helpful UI solutions. Even though it may seem somewhat contradictory to the principles of user-centered design and multidisciplinary cooperation, the knowledge of the most experienced designers and usability experts is one of the most profound strengths of a corporate usability organization. When it's impossible to test every detail, the opinion of the expert counts for more. The experts, in return, need opportunities to nurture their expertise.

Midpriced. Advocates of 'discount' usability have argued their case convincingly by proving the validity of its results-four test subjects are enough; usability tests with lo-fi (low-fidelity) prototypes will tell you what you need to know. We accept their findings, but we also recognize that discount isn't always the best value for the money. There are two main routes to improving the validity of usability research-or to slip into the midprice or even premium range: (1) improve the fidelity of the stimulus material (i.e., simulations and prototypes) and (2) construct more valid test settings with reference to sampling and context. Both will quickly outspend a true discount budget. Chapters 6 to 11 showcase a couple of cases where discount methods were rejected even though the projects were in their early phases: Pekka Ketola's contribution (Chapter 8) about internationalizing and localizing a UI style illustrates one reason; Turkka Keinonen's contribution (Chapter 9) about in-car user interfaces shows another.

International. Global perspective on usability is a recurring topic in this book, and it is also one of the main challenges in making reasonable tradeoffs between comprehensive engineering approaches and discount usability. Simple methods can grow into remarkably large-scale efforts as the evaluations are carried out in several countries on separate continents. Translating the stimulus material, organizing travel arrangements, and teaching local moderators-it all takes time. The discount usability mindset starts to turn into something else when your goal is to test the understandability of terms and navigation sequences in several language areas. Researchers must heed the relationship between the cost of their efforts and the usefulness of the results, as Pekka Ketola does. The tasks that require world tours must be distinguished from those for which testing with some colleagues in a neighboring department suffices.

Real-time. Time is a crucial factor in UI design. It is even more critical when designing user interfaces to be manipulated while driving. Drivers' attention-sharing behavior quickly becomes the main usability challenge. A user interface for a car system itself must adhere to severe limits on how much of the driver's visual attention it requires. Thus, designing for attention sharing calls for trials and test tools that deliver inputs at the pace of real products, and that provide realistic tactile and auditory responses. Although driving is an extreme case underlining the importance of a users' attention-sharing performance, the issue is common in mobile UI design. Mobile UIs must be robust in the face of continuous interruptions and usable in recurring dual-task situations. Consequently, your usability evaluation should be able to assess the robustness of proposed solutions, which puts tough requirements on the prototypes.

Behavioral. Contextualizing usability approaches in user studies was emphasized in Part 2. This part (Chaps. 6 to 10) deals with design and concept creation, and here we often have to accept less context and more isolation of the human-machine interaction in a research setting. Behavioral usability research can focus on a specific aspect of human performance: perhaps the ability to learn a new input technique or to divide attention between primary and secondary tasks. When this is the case, the projects may rely on laboratory measurements and company-internal subjects-or at least such reliance can be seen as an acceptable tradeoff for efficiency. Typically the aim in these test settings is to compare several new design solutions, or a new solution to an existing one, with quantitative criteria. The tradeoff pays the researcher back by providing quantitative results relatively soon.

Contextual. Another category of design and research approaches aims at integrating solutions for a coherent user experience. Screen designs and interaction conventions must be combined to form a unified concept for a mobile phone UI, and various technologies and services must be conjoined to deliver mobile Internet services, as Pekkarinen and Salo demonstrate (Chapter 10). An acceptable or interesting solution is judged not only by reference to an existing product or an alternative concept but also, above all, by the users' expectations for utility and pleasure. Test settings accordingly have to be more contextual than the isolated behavioral measurements applied to the study of specific aspects of interaction. Assessments are carried out in the field with external subjects and with open-ended research approaches.

Responsive. Usability activities in industrial settings are typically related to the product creation process. The activities may provide support to product programs or may actually be a part of design-related decision making. In these cases usability functions in a responsive mode, reacting to the challenges posed by emerging technologies and the design of new products. Product programs set the schedules and determine the necessary level of detail for the results. It's essential to come up with concrete proposals or definitive answers as to the superiority of one of the alternatives. In Chapter 11 John Rieman tells us about one way of doing responsive usability engineering.

Proactive. The remaining examples concern concept creation exercises or interaction research preparatory to making implementation decisions. UI designers and usability specialists may meet the project objectives and approaches relatively independently, as they see appropriate. They are given an opportunity to take the initiative and propose solutions on the basis of their understanding of optimal user experience. Usability research of this sort functions in a proactive mode.

Managed. Constructive, midprice, behavioral, and proactive mode-the whole spectrum of usability approaches is needed to balance the speed, costs, and reliability requirements of usability engineering. Corporate usability groups must have enough flexibility in its toolkit of methods to cover the range. Management must understand the business, the user interfaces, and the importance of usability to make the right choices among possible modes of operation. The corporate usability function will not be organized effectively and function smoothly unless it has facilitators and leaders. User interface management ensures that expertise in a usability organization is utilized to its full potential. To realize this vision, the UI design manager must have the skills to manage the team and the project, quality, documentation, and the product itself, as Christian Lindholm and Turkka Keinonen show in Chapter 6.