Presentation
Towards Understanding the Role of Perceived Situational and Relational Risk in Human-Automation Trust Under Multitasking Environment: A Pilot Study
DescriptionPrevious study indicated that subjective trust elevated when users experienced high perceived risk in a cognitively demanding environment. Yet, it remains unclear whether this effect was driven by the user’s beliefs about the consequences of interacting the automation (perceived relational risk) and/or completing the task itself (perceived situational risk). Theoretically, perceived relational risk is expected to directly influence subjective trust, whereas perceived situational risk is not (Stuck & Walker, 2021). The present study extended Sato et al.’s (2020) work to examine whether the two types of perceived risk mediated the effect of risk on subjective trust across task load conditions. Sixty participants concurrently performed a tracking task, a system monitoring task aided by a 70%-reliable automation, and a resource management task. Task load was manipulated by altering the difficulty of the tracking task, while risk was manipulated via deceptive instructions. Participants in the high-risk group were told they would repeat the experiment if they performed poorly, whereas those in the low-risk group received no such instruction. Bayesian mediation analysis indicated that neither type of perceived risk mediated the effect of risk on subjective trust across task load conditions. Nevertheless, perceived relational risk influenced trust as measured by Jian et al.’s (2000) scale across task load conditions. In contrast, perceived situational risk influenced purpose-based trust as measured by Chancey et al.’s (2017) scale, but only under low task load condition. Results partially support theoretical predictions suggesting that perceived situational and relational risk may independently influence different facets of subjective trust.
Author
Event Type
Late Breaking Results
TimeThursday, October 16th5:10pm - 5:30pm CDT
LocationGrand Hall G/H