The Impact of Automation on Error Detection: Some Results from a Visual Discrimination Task

Abstract

A dynamic decision-making task was designed for microcomputer that allowed subjects to operate in either a manual mode, or an automatic mode. Nine subjects performed in a repeated measures design that presented identical errors for detection in the two modes. Results showed that sensitivity was higher in the automatic mode; the manual mode elicited a conservative response bias. NASA bipolar rating scales presented a clear picture of higher workload in the manual mode.