Abstract
This study sought to determine whether the type of display that is most beneficial for performing supervisory tasks would change depending on the relationships between system components. Eight subjects detected and diagnosed failures for systems whose variables were related either by correlation or by causality. One of four dynamic displays represented the three system components: a set of bar graphs, a triangle, a pentagon, or a schematic face. Subjects supervising the correlational system diagnosed the system by choosing which component had become less correlated with the others. Subjects diagnosed the causal system by determining which way the causal structure relating the components had changed. For both the correlational and causal systems, and for both detection and diagnosis there was a cost to using object displays or a face display rather than bar graphs.
Get full access to this article
View all access options for this article.