Number of Interrupting Events Influences Response Time in Multitasking, but not Trust in Automation

Abstract

ABSTRACT Objective The present study examined how the number of interrupting events (interruption load) influences the effect of task load on human-automation trust and resource allocation in a low-fidelity flight simulation environment. Background Trust is one critical factor that influences successful human-automation interaction. In the previous research, operators reported lower trust scores and made fewer fixation toward an automated system, which assisted a task, when competing task in the same workspace demanded more attention from the operator. However, it is unclear whether human-automation trust is influenced by frequent shift of attention away from a task assisted by an automated signaling system. Methods Participants concurrently performed a tracking task, a system monitoring task, and a communication task. An automated signaling system was employed to assist the system monitoring task with 70% reliability. Task load was manipulated by the difficulty of the tracking task while interruption load was manipulated by the varying the frequency of auditory messages in the communication task. Results Results demonstrated an effect of task load on human-automation trust and resource allocation, replicating previous findings. Further, participants responded faster to an auditory message that occurred less frequently when performing a tracking task at the low difficulty level but automation trust did not vary. Conclusion While operators reported higher trust levels to imperfect automation under lower task load, number of interrupting events does not influence their trust.