Abstract
The aim of this study is to examine some of the components involved up to and including choice of the first test in a network fault-finding task. The components investigated were identifying the set of possible faults or consistent fault set (CFS), applying the half-split rule to a specified CFS, and making the first test without the CFS being specified. Factors of network size (54 units and 24 units) and complexity, as defined by four types of CFS configuration, were manipulated. Results indicated that errors occurring in identification of the CFS are mostly omission errors. Performance of this component is degraded by increasing both network size and complexity of the CFS. The same pattern of effects is observed in selecting the first test when the CFS is not specified. Ability to select a half-split test from a given CFS is affected only by complexity of the CFS configuration. Future studies need to differentiate further the psychological demands imposed during the initial stages of faultfinding.
Get full access to this article
View all access options for this article.
