Interactions - July/August 2009

buttons or icons), and to constantly remind the user about how the various elements of the design would eventually fit together in the envisioned product. Without human intervention, the participant is unclear about how the system should respond. On the other hand, paper prototypes do provide a quick means for testing basic interactivity that may be essential for specific design decisions [ 5]. Tools for agile prototyping and electronic sketching are also increasingly available to support quick design iteration [ 6, 7].

Conversely, fully interactive high-fidelity prototypes (created with Visual Basic, Flash, HTML, among others) are neither easily nor quickly produced.

Paper prototyping relies on the exclusive use of paper-based artifacts, which are physically detached from the use of any real device or digital interface. High-fidelity prototypes, although difficult to quickly develop, allow test participants to easily enter data and execute commands, while being much less dependent on the facilitator.

Emotion, Design, and Prototyping When it comes to prototyping and evaluating a mobile touch-screen interface, understanding the emotional response of the user is imperative to better evaluate the quality of design. In particular, not only the appearance of the interface (which can be well anticipated by current prototyping tools and practices), but also the physical holding, feeling, manipulating, and touching of the device are important factors in determining the quality of the user experience [ 8]. For this reason, standard paper prototypes, though useful and quick to gen-

erate, are very limited in delivering the depth of user experience necessary to adequately inform design iteration.

In discussing the relationship of user experience and design, Don Norman states that the “ visceral” (or physical) level is the simplest and most primitive cognitive process [ 8]. With regard to handheld devices, visceral is all about look, feel, and sound, i.e., how a device, including the interface, looks and feels in the hand of the user. The iPhone is one of the greatest examples of the visceral experience. It was designed, in great part, for the visceral level of cognition—Apple designed for visual and physical impact. Evaluating the mobile user experience without considering visceral factors is like evaluating the quality of a meal by looking at the menu. A lot can be said about the dishes, but the experience is not there. It is a challenge, of course, to anticipate this experience in early prototyping.

Beyond the visceral experience, the “behavioral” level of cognition is about designing device interaction or behavior to reflect human behavior [ 8]. In other words, device design becomes intuitive in the way it complements one’s implicit assumptions about how it might work. Last, to design for the “reflective” level of cognition is to appeal to one’s aesthetic sensi-bilities, uniqueness, and cultural preferences [ 8]. From such a design perspective, people relate to and acquire a personal adherence to a device as part of their identity and self-expression.

Understanding these three levels of cognition is extremely relevant, because emotional engagement at every level strongly

influences human-interface interaction from a physical, aesthetic, and usability perspective. Moreover, if we need to take into account these emotion-centric factors early on in the life-cycle of device design, it is clear that paper prototyping cannot deliver the necessary insight into a full visceral and behavioral experience of the interface in the context of handling the physical device. In other words, if we only use paper separate from its actual relationship to the physical device, we may bypass important elements of the user’s emotional experience. As a consequence, inadequacies of this kind may lead to highly artificial (and ultimately irrelevant) evaluation results. To date, producing high-fidelity electronic prototypes or beta-version releases appear to be the only viable way to have users test or try out mobile applications on their devices. We, however, recommend another intermediate solution.

If we could anticipate the mobile user experience (testing the interface with a device), yet spend minimal effort in implementing high-fidelity prototypes, we would have found a far more efficient way to prototype mobile applications with a much higher ROI in time, energy, and valuable user feedback.

So, how can we arrive at a prototyping technique that can deliver development speed, with both visceral and behavioral user experiences that far exceed the use of paper, but without the need to develop high-fidelity prototypes?