Previous Table of Contents Next


2. WHAT IS THE GAP?

User-centered design can be thought of as a cognitive task that requires the designer (or group of designers) to encode and transform information in the course of application definition. For example, in general, transformation occurs when the designer combines end user work descriptions with a user interface syntax (e.g., the rules governing a graphical user interface) to produce a new means to accomplish some task. A more specific example within this general category is when the designer makes word choices to try to adhere to a maxim such as “Use terms familiar to the user”. In either case the designer is taking information in one context, transforming it, and expressing it in another. This ability to code and recode information is basic to human cognition and has been the focus of much psychological research. Psychological accounts of the efficacy of coding often rely on the concept of “representation” whereby coding results in representations more amenable to completing some particular cognitive task (e.g., Miller, 1956). Familiar elementary school rhymes used to help recollect the number of days in each month, or acronyms such as “ROY G BIV” for the colors of the spectrum, are simple instances of coding. The psychological literature abounds with other examples, including studies of chunking as a means to increase the amount of information that can be retained, studies of mental rotation (Shepard and Metzler, 1971), theoretical accounts of the structure of memory (e.g., Anderson, 1976), and accounts of how choice of representation can be critical to success in problem solving (e.g., Simon, 1981). In all cases a key part of the account given of human performance has to do with the mediating role played by internal representation or coding. Similarly, it is argued here that representation is key to defining the gap bridged in design because this gap lies between representations used in the design process.

One representational metatheory characterizes any discussion of representation as necessarily a discussion of a “representational system.”1 A representational system must specify:


1This discussion is based on a chapter by S. E. Palmer, “Fundamental Aspects of Cognitive Representation”, p. 262. Norman (1993) also discusses the notions of represented and representing worlds, but from the vantage of the end user rather than the designer.
1.  What the represented world is
2.  What the representing world is
3.  What aspects of the represented world are being modeled
4.  What aspects of the representing world are doing the modeling
5.  What are the correspondences between the two worlds

User-centered design entails the creation of a particular “representational system” through a process that begins with user data and ends with a hardware and software system for some end user task. To fulfill the necessary conditions for a design the designer will have implicitly or explicitly addressed each of these five elements. In this view, the gap between user data and concrete design is a gap between a “represented world” and a “representing world”. The represented world is the end user data, and the representing world is the application user interface.2 Bridging the gap requires the definition of items 3 through 5, although sometimes the designer’s job may begin with the basic choices of what the represented world is and what the representing world should be.


2The epistemological question of how the represented world is known, its correspondence with the “real world”, and the objectivity of this knowledge is an interesting one and affects how the representational system is defined. However, for this context it will be sidestepped with the caveat that how user-centered analysis is captured has a significant impact on the design process, as will be discussed later.

The selection of aspects of the represented world to be modeled and appropriate aspects of the representing world doing the modeling are bound up in the process for developing the correspondences between the two. This is the crux of the designer’s decision making, leading to a concrete design. Figure 3.1 illustrates the Gap between the represented and representing world and graphically shows the idea that correspondences developed through information transformation span the gap.


Figure 3.1  The representational system in design.

Artifacts containing the results of user analysis make up the represented world. In this paper the represented world comprised the results of contextual inquiries (Wixon and Ramey, 1996), rendered in scenarios and extracted use-cases that distilled many hours of meetings with end users. As will be described later, successive iterations during design elaborated the initial view of the represented world. On the other side of the gap, the representing world was a software application based on Windows NT/Windows 95™ operating systems and user interface look and feel. Two additional factors shape the development of correspondences in the representational system during design. First, the upper box in the middle of the figure shows mediating abstractions as important transformations for bridging the gap. They help manifest the represented world in the representing world by suggesting appropriate interpretive structure. Examples here include metaphors for the represented world and other conceptual models at once removed from the data of user experience, but sufficiently linked to it to serve as the basis for users’ understanding. These mediating abstractions can “have entailments through which they highlight and make coherent certain aspects of our experience” to use Lakoff and Johnson’s phrase about the role of metaphor in understanding (Lakoff and Johnson, 1980, p. 156). This understanding is shared between the designer and the end user. The designer creates the representing world around such models and metaphors, and the user comprehends the laws of that world via the same devices.


Previous Table of Contents Next