Diana Gromala

Critical Methodologies for the Visual Design Interface

Introduction

There is a growing recognition that Visual Design is a critical component of a successful website, as evidenced by such sites as The McKinley Group's Magellan directory, but this does not go far enough. Traditional Visual Design methodologies, while useful, have too often fallen into subjective opinion building, rather than substantive and rigorous methodologies. This paper argues for a more rigorous approach to Visual Design of html (and vrml) interfaces, one which employs a critical methodology for Visual Design that is both analytical and prescriptive. Thus, it is useful for both the analysis and design of html interfaces.

Visual Design: Analytical Methodology

The formal qualities of a work are fairly easy to examine: How legible the work is, how it appeals to and directs the eye through complex information, and how effectively it transmits its message are examples of formal qualities that can be evaluated in the light of traditional Visual Design practice. The formal qualities of a web page additionally need to include interactive aspects specific to web design: how well it is organized, interfaces among differing modalities and media, its ease of access, and how well it provides content within the hypertext web. As interactive features are added to a web page, the formal analysis must perforce deal with the quality of the interaction. Thus, this analytical methodology will additionally include emergent design methodologies derived from Interaction Design and User-Centered Visual Design studies.

The second part of an analysis of a work deals with the cultural milieux in which the work resides. This analytical method is situated in a cultural studies and critical theory perspective, which considers the appropriate contextual and epistemological questions of cultural knowledge and consensus about the language of text and image relationships. This analytical strategy provides a basis from which to question the underlying assumptions and subjective biases of the traditional design analysis outlined above to ask HOW a Visual Design Interface works and what contextual questions need to be considered. This is aspect is made especially relevant by the international nature of the World Wide Web, since pages on another continent are nominally just as easy to obtain as pages in the next building.

Finally, the analytical component derived from a cultural studies perspective takes into account larger questions of relevant perceptual shifts attendant to historical changes in technologies, as well as historical precedents. It is useful to analyze, for instance, how the development of the www parallels the development of technologies from Egyptian Stylus and Gutenberg's Press to radio and television , and how these developments influenced the production, design, and consumption, and interpretation of media.

Visual Design: Prescriptive Methodology

The prescriptive methodology is comprised of a set of considerations, rather than a much less useful "how-to" approach. It bears considerations based on traditional Visual Design and Visual Literacy principles, in combination with principles of interactivity, navigation, and multimedia from the emerging fields of Interaction Design and User-Centered Design. In addition, since the technology affords multiple modalities, time-based considerations such as those employed in film and animation come readily into play. Finally, an important aspect is the use and manipulation of 3D objects within a web page. While there is an enthusiastic embrace of VRML, many avenues still need to be explored with respect to appropriate interaction and integration with 2D web content.

Finally, user-testing for Visual Design interfaces needs to be carefully reconsidered in ways that extend traditional, scientific user-testing. User-testing doesn't map well into the more subjective visual realm, because it only asks for a comparison between a limited set of predetermined choices. There are a number of problems with this. First, in asking whether A is better than B, the subject may legitimately think that "better" also means "more familiar", and the result is that new designs are too hastily discounted. A design that is too new to be instantly appreciated, and a user test may simply confirm a preference for the familiar. Second, user-testing often serves to confirm whether correct design choices have been made, and to find mistakes that may have been overlooked. The difficulty is that user testing serves only an analytical or a diagnostic function. By its very nature, it cannot be used prescriptively. Third, user testing proceeds on the paradigm that there is some innate behavior that can be quantified. The central assumption is that everyone acts the same way, with an easily quantified variability. This is a wrong assumption in the realm of aesthetic judgement, because there is no such thing as an innate aesthetic behavior. This paper does not argue against user-testing per se. Instead, the prescriptive methodology should include user-testing as a diagnostic technique, not a design technique.

Conclusion