Applying a formative ecological framework to simulator design challenges

Human-in-the-loop interactive simulators are widely recognized as a cost-effective and safe way to train workers in challenging work situations, and to more tractably research human performance. Designing cheaper, lower fidelity and more tractable computer-based simulators has been proposed as a way to widen simulator adoption and extend training and human performance research benefits. There is currently no systematic method to design simplified simulations while retaining psychological fidelity. Design decisions are typically based on common tasks and simulator designers' intuition and expertise. This is difficult to replicate, complicates comparisons between simulators, and presents problems for scientific defensibility of research. This thesis demonstrates how Cognitive Work Analysis (CWA)'s phases can be applied to support designers both in systematic, defensible design of simplified simulators, and also in ongoing scenario design and development. A subset of CWA design methods are demonstrated in a petrochemical domain case study.