Mitigating Judgmental Fallacies with Social Robot Advisors

The role of social robots as advisors for decision making is investigated. It has been consistently shown that when asked to rank options, people often make fallacious judgements. Furthermore, such fallacies can be sensitive to presentation mode. We study whether having social robot advisors presenting options can mitigate and reduce the fallacy rates of participants. For this purpose we explored a novel presentation mode of options with conjunction judgmental fallacy, namely, choosing among different rank-orders, as opposed to rank the options themselves. We first show that the mere presentation mode has a significant mitigating effect on the fallacy rates. We then further show that when social robot advisors present the rank-orders, the fallacy rates of participants significantly decrease even further. Moreover, participants perceive the fallacious robot as more likeable and intelligent, but assign the non-fallacious robot to trustworthy roles, such as jury and analyst. These results suggest that social robot advisors may be used to influence and mitigate human fallacious judgmental decision making.