The Effect of Workload on Trust Repair: Testing a New Model of Trust Repair

Given the importance of trust in human-automation interaction, there has been intense activity in understanding trust repair (Esterwood & Robert, 2021). However, despite the numerous trust repair studies, there is no coherent understanding of when and why trust repair works. A recently proposed theoretical model identified two unique categories of trust repair (central and peripheral) and the factors that might affect their success (Pak & Rovira, 2023). The current study tested a specific hypothesis arising from the theoretical model: variations in workload will affect central trust repair efficacy, but peripheral trust repair will be unaffected by workload. Participants engaged in a synthetic multitasking situation with the aid of automation that used the two trust repair strategies. Consistent with the theory, the efficacy of peripheral repair strategy did not vary between workload conditions; however, there were no significant differences between either repair strategies immediately after automation failed. The current research, which provides partial support, was the first empirical test of a prediction from the new theoretical model of trust repair. Future research may follow up our findings with more precise workload manipulations and alternate exemplars of the trust repair messages for central and peripheral conditions.