Training and transfer of fault-finding skill

Abstract

The experiment investigated the effect of different training materials (a technical story versus a set of diagnostic heuristics) on the training and transfer of faultfinding skill. Two simulated chemical plants were used as the fault-finding domain. During the training session, fault-finding was measured before and after mastery of the training material and after practice in fault-finding. The latter test involved ‘old’ faults and ‘new’ faults from the same and different categories to those encountered during training. Subsequently transfer was measured to a second plant involving the same components and variables in a different layout to the plant used for training. Both conditions improved at fault-finding both after mastery of the training material and after practice when the test involved ‘old’ faults. The technical story condition performed better than the diagnostic heuristics condition on ‘new’ same category faults although neither condition was able to solve ‘new’ different category faults. Transfer to the second plant was high and positive for both training conditions. The results are discussed with regard to the role of ‘theory’ in training fault-finding. It is suggested that in order to explain such training and transfer effects, it is necessary to define terms more precisely and to attempt to describe the cognitive representations of fault-finding skill which are developed by different training methods.

Keywords:

• Fault-finding skills
• Diagnostic heuristics
• Error limitation