AMME: an Automatic Mental Model Evaluation to analyse user behaviour traced in a finite, discrete state space

Publisher Website
Google Scholar
Abstract
To support the human factors engineer in designing a good user interface, a method has been developed to analyse the empirical data of the interactive user behaviour traced in a finite discrete state space. The sequences of actions produced by the user contain valuable information about the mental model of this user, the individual problem solution strategies for a given task and the hierarchical structure of die task-subtasks relationships. The presented method, AMME, can analyse the action sequences and automatically generate (1) a net description of the task dependent model of die user, (2) a complete state transition matrix, and (3) various quantitative measures of the user's task solving process. The behavioural complexity of task-solving processes carried out by novices has been found to be significantly larger than the complexity of task-solving processes carried out by experts.