Multidimensional Scaling as a Method for Probing the Conceptual Structure of State Categories: An Individual Differences Analysis
Identifying the underlying decision criteria used by operators to classify system state, and revealing the way in which that information is internally represented is one of the challenges facing designers of control systems. This paper describes the use of multidimensional scaling (MDS) to probe the structure and composition of the internal conceptual models used by operators to identify system state. Specifically, the issue of individual differences in mental model is addressed, as well as the impact of these differences on individual performance in a classification task. Twenty subjects were trained as operators to classify instances of system data into one of four system state categories. After training, subjects were asked to rate the similarity between instances of system state. Results showed that the dominant dimensions used by an individual are related to his/her performance on the classification task.