| The purpose of this dissertation was to understand the key success factors in adapting to change from a fundamental and integrated point of view. To assist in studying these issues, concepts from ecological psychology, in particular coordinative structures, were utilized for the domain of process control. Rasmussen's (1985) Abstraction Hierarchy (AH) was used to model the functional structure of a process control microworld (DURESS II) and to predict the resultant behaviour if a participant utilized higher-level control. Three experiments were conducted to test predictions based on the theory using different types of perturbations: global and local component perturbations, and interface form perturbations. Two displays (P and P+F) were used in the studies, motivated by different interface design principles. There were four main findings. First, the participants who used higher-level control were able to successfully adapt to the perturbations. Second, higher-level functional and invariant information were found to be important for inducing higher-level control. Third, the existence of multiple structural degrees of freedom in the work domain was found to be a key factor in making possible and providing flexibility for higher-level control. Fourth, cognitive capabilities enhanced by experience were found to be important for higher-level control. A number of contributions resulted from this research. First, this is the first set of studies to systematically investigate: (a) coordinative structures in process control, (b) stability beyond representative conditions for DURESS II, and (c) interface form perturbations for process control. Second, this research provides support for a general ecological theory of adaptation and coordination, by discovering evidence for the same phenomenon (higher-level control) in incredibly different domains, process control with DURESS II and speech motor control. Third, this research provides insight into the benefits of Ecological Interface Design in supporting adaptation, and introduces an integrated framework that helps expand the approach. Finally, a concise and novel adaptation metric (the A-index) is proposed as a tool to help predict successful adaptation. |
