HOW HUMOR COULD AFFECT TRUST IN HUMAN–ROBOT INTERACTION

Guo, Xinyi

Permalink

https://hdl.handle.net/2142/125092 Copy

Description

Title

HOW HUMOR COULD AFFECT TRUST IN HUMAN–ROBOT INTERACTION

Author(s)

Guo, Xinyi

Issue Date

2020-05-01

Keyword(s)

Trust; Human Robot Interaction; Human Computer Interaction

Abstract

This thesis examines scenarios in which a person interacts with collaborative robots for certain tasks, for example when visitors are guided by a navigation robot or when doctors are assisted by surgical robotics arms. In particular, we consider robots that can initiate small talk with appropriate jokes, which is a natural human behavior when icebreaking or trying to form a bond. We ask the question: Would the user put more trust in such a robot and feel more comfortable, or less, when interacting with the robot? To answer the question, we designed an online controlled experiment that mimics a direction-guiding robot, which could give instructions to the user to reach a final goal, and we performed the controlled experiment with 22 valid participants. The results show different levels of trust among participants. Namely, 63.6% believed that robot with humor has more knowledge in the task and is more comfortable to interact with, while 77.4% prefer that robot as an actual guide in a museum setting. In addition, depending on the seriousness of the application, people have different expectations of the robot that they are working with. A friendlier and funnier robot is preferred in an exploratory setting, while a more direct and concise robot is preferred in a straightforward task completion setting. However, given the same setting, more people prefer a humorous robot and have more trust in it.

Type of Resource

text

Language

eng

Owning Collections