Graphical Interfaces to Complex Systems: Separating the Wheat from the Chaff

Abstract

There seems to be a clear consensus that graphical interfaces provide an opportunity to integrate data from complex process in a way that can greatly enhance the problem solving ability of human operators in the future. However, this consensus is maske by a proliferation of terms to express this position in the basic and applied research literatures (e.g., “integrality,” “configurality,” “proximity-compatibility,” “visual momentum,” “direct manipulation,” and “ecological interface”). While the subtle nuances that distinguish among these terms are of academic interest, designers have greater concern for the general principles that might be gleaned from across the subtle distinctions. Base on a thorough review of the basic and applied literature (Bennett & Flach, In press), we argue that there is one basic characteristic of graphical representations that is critical for supporting problem solving. A good graphical display is one whose geometric (space/time) constraints reflect the functional constraints in the proess being represented. In this presentation, we will demonstrate what we mean by a “functional constraint” in a process and a “geometric constraint” in a display. We will demonstrate alternative mappings from “functional constraints” to “geometric constraints.” We will also discuss the implications of these mappings for the type of processing (cognitive versus perceptual) required of the human operator.