Evaluating models in systems ergonomics with a taxonomy of model attributes

Abstract

A model, as the term is used here, is a way of representing knowledge for the purpose of thinking, communicating to others, or implementing decisions as in system analysis, design or operations. It can be said that to the extent that we can model some aspect of nature we understand it. Models can range from fleeting mental images to highly refined mathematical equations of computer algorithms that precisely predict physical events. In constructing and evaluating models of ergonomic systems it is important that we consider the attributes of our models in relation to our objectives and what we can reasonably aspire to. To that end this paper proposes a taxonomy of models in terms of six independent attributes: applicability to observables, dimensionality, metricity, robustness, social penetration and conciseness. Each of these attributes is defined along with the meaning of different levels of each. The attribute taxonomy may be used to evaluate the quality of a model. Examples of system ergonomics models having different combinations of attributes at different levels are provided. Philosophical caveats regarding models in system ergonomics are discussed, as well as the relation to scientific method.