Cognitive Analysis of Expert and Novice Troubleshooting Performance

Abstract

The purpose of this study was to investigate the influence of mental models on the performance differences between expert and novice service technicians who troubleshoot faulty technical equipment. This study involved the use of behavioral and cognitive methods of analysis to identify and develop the ideal mental models that are used by expert technical troubleshooters. These two methods of analysis involved the identification of the specific tasks used during troubleshooting and the collection and analysis of the verbal protocols of expert and novice technicians as they attempted to identify faults in technical equipment. The results of the study showed definite differences between experts and novices. The experts were able to acquire and correctly interpret the initial symptoms of the problem, and from the interpretation of the initial symptoms the experts were able to represent the problem through the development of an accurate problem space. The sequences used by the experts to find faults in technical systems involved the generation of hypotheses that were used to reduce the size of the problem space and, therefore, to narrow the number of potential locations for the fault. In contrast, the novice subjects were unable to accurately interpret the initial symptoms of the problem. They were also unable to represent the problem through the development of an accurate problem space. The lack of an accurate problem representation resulted in the apparent random search for faults in the system.