Abstract
Owing to technological advancements, individuals can increasingly automate and delegate private decisions. However, prior research shows that decision-makers tend to prefer human decision support despite the superiority of algorithms. Further, individuals prefer retaining control of decisions despite increased effort. We propose a model explaining underlying considerations of these paradoxes in decision support acceptance. In a vignette-based experiment in the context of investment management, we test the model to explain trade-offs in varying levels of automation and user control and analyze effects on the intention to utilize decision support. Results provide support for a positive effect of automation on performance expectancy and a negative effect of user control on perceived risk. These findings support the idea of increased automation in decision-making while letting users retain control over the process. This study extends our understanding of decision support in private contexts and holds implications for providers of decision support systems, particularly robo-advisors.
Recommended Citation
Rühr, Alexander; Berger, Benedikt; and Hess, Thomas, "Can I Control My Robo-Advisor? Trade-Offs in Automation and User Control in (Digital) Investment Management" (2019). AMCIS 2019 Proceedings. 2.
https://aisel.aisnet.org/amcis2019/cognitive_in_is/cognitive_in_is/2
Can I Control My Robo-Advisor? Trade-Offs in Automation and User Control in (Digital) Investment Management
Owing to technological advancements, individuals can increasingly automate and delegate private decisions. However, prior research shows that decision-makers tend to prefer human decision support despite the superiority of algorithms. Further, individuals prefer retaining control of decisions despite increased effort. We propose a model explaining underlying considerations of these paradoxes in decision support acceptance. In a vignette-based experiment in the context of investment management, we test the model to explain trade-offs in varying levels of automation and user control and analyze effects on the intention to utilize decision support. Results provide support for a positive effect of automation on performance expectancy and a negative effect of user control on perceived risk. These findings support the idea of increased automation in decision-making while letting users retain control over the process. This study extends our understanding of decision support in private contexts and holds implications for providers of decision support systems, particularly robo-advisors.