Interface Design: A Cognitive Systems Engineering Perspective

Kevin B. Bennett, Ph.D.
Department of Psychology
Wright State University
Dayton, OH

Abstract

The process of creating effective interfaces will be discussed from a Cognitive Systems Engineering perspective. There are three components that must be considered in designing effective computerized decision support: 1) the work domain, 2) the agents, and 3) the interface. Each of these components contribute a set of constraints on performance; the quality of mappings between the components will ultimately determine interface effectiveness. The general characteristics of a work domain determine the category of interface design strategy that will prove effective. Work domains that are driven by the laws of nature (e.g. process control) will benefit from analogical, geometric forms that reflect the inherent constraints. Work domains that are driven by user intentions (e.g., information retrieval) will benefit from metaphorical forms (e.g., icons) that relate domain constraints to pre-existing, familiar concepts. Interface design strategies for work domains that fall in the middle of this continuum (e.g., military command and control) present a different set of challenges and successful design solutions are less well-documented. An on-going project to develop decision support for mobile Army commanders has revealed potential solutions. The key elements of this interface will be discussed in terms of three principles of effective interface design: direct perception, direct manipulation, and an intact perception-action loop.