Influences of Evaluation and Gaze from a Robot and Humans’ Fear of Negative Evaluation on Their Preferences of the Robot

Influences of Evaluation and Gaze from a Robot and Humans’ Fear of Negative Evaluation on Their Preferences of the Robot
复制标题

机器人的评价和目光以及人类对负面评价的恐惧对机器人偏好的影响

DOI:
10.1007/s12369-014-0270-y
复制
发表时间:
2014
影响因子:
4.7
通讯作者:
T. Kanda
T. Kanda
中科院分区:
计算机科学3区
文献类型:
--
作者:
T. Nomura;T. Kanda

文献摘要

被引文献

相似文献

为了研究机器人的评价、机器人的注视以及人类对负面评价的恐惧(FNE)对人类对机器人的偏好的影响,我们进行了一项实验,在该实验中,人类大小的机器人被用作顾问角色,提供外语教育方面的建议。控制了三个独立变量:机器人的评估(评估性与非评估性机器人)、用户的FNE(低FNE与高FNE)和机器人的凝视(抑制凝视与正常凝视),因变量是参与者对机器人的偏好作为主观衡量标准和作为客观衡量标准的话语量。实验结果表明,当机器人评估具有较低FNE的参与者时,他们更喜欢使用具有正常凝视行为的机器人,而不是具有较高FNE的机器人。虽然对于FNE较高的人来说,抑制凝视的效果尚不清楚,但对于FNE较低的人来说,当机器人评估他们时,抑制凝视会降低他们参与的意图。此外,实验结果表明,具有较高FNE的人倾向于更信任机器人,当机器人没有评估他们时,当机器人使用正常凝视时,参与者说话更多。本文讨论了潜在的评估能力的应用程序的影响,如医疗保健和教育。
To investigate influences of evaluation from robots, gazes from the robots, and humans’ fear of negative evaluation (FNE) into the humans’ preferences of the robots, we conducted an experiment in which a human-sized robot is used for an advisory role, providing suggestion in foreign-language education. There were three independent variables controlled: evaluation from robot (evaluative vs. non-evaluative robot), users’ FNE (low FNE vs. high FNE), and gaze from robot (refrained gazes vs. normal gazes), and dependent variables were participants’ preferences of the robot as subjective measures and amount of utterances as an objective measure. The experimental results suggested that when the robot evaluated participants with a lower FNE, they preferred to use the robot with normal gaze behavior more than those with a higher FNE. While the effect of refrained gaze is not clear for people with higher FNE, for people with lower FNE, refrained gaze reduced their intention to participate when the robot evaluated them. Moreover, the experimental results suggested that persons having higher FNE tend to trust the robot more, and participants spoke more when the robot did not evaluate them and when it used the normal gaze. This paper discusses the implications for applications with potential evaluation capability, e.g. for healthcare and education.