As the roles of social robots are increasingly diverse, it is important to design the robots according to their application fields and end-users needs. A robot’s voice is a strong social cue, the design of which can influence people’s affective evaluation and acceptance toward robots. The aim of this study is to investigate the affective evaluation of different robot voices in various application fields, and obtained suitable voice types for robots in different application fields. In particular, this study focused on the three applications (i.e., shopping reception, home companion, and education), investigated the effect of voice types (i.e., male, female, child, and synthetic) of social robots on the affective evaluation of users. Principal component analysis identified three latent influencing factors (i.e., social skills, competence and the state of the interaction relationships). Multivariable analysis proved that for overall acceptance, significant interaction effects existed between robots’ voice types and their application fields. For shopping reception robots, the most acceptable voice type is adult male voice and child voice. For home companion robots, the most acceptable robot voice types are adult male and child voices. For education robots, the most acceptable voice types are adult female and male voices. The results of this study are expected to construct design principles for robot voice design in various applications.

随着社交机器人的角色越来越多样化，根据机器人的应用领域和最终用户需求设计机器人非常重要。机器人的声音是一种强烈的社交线索，其设计可以影响人们对机器人的情感评价和接受度。这项研究的目的是调查不同应用领域中不同机器人语音的情感评价，并获得适合不同应用领域中机器人的语音类型。特别是，本研究着重于三种应用（即购物接待，家庭陪伴和教育），调查了社交机器人的语音类型（即男性，女性，儿童和人工合成）对用户的情感评价的影响。主成分分析确定了三个潜在的影响因素（即社交技能，能力和互动关系的状态）。多变量分析证明，对于总体认可，机器人的语音类型与其应用领域之间存在显着的交互作用。对于购物接待机器人，最可接受的语音类型是成年男性语音和儿童语音。对于家庭伴侣机器人，最可接受的机器人语音类型是成年男性和儿童语音。对于教育机器人，最可接受的语音类型是成年女性和男性语音。这项研究的结果有望为各种应用中的机器人语音设计构建设计原则。最可接受的语音类型是成年男性语音和儿童语音。对于家庭伴侣机器人，最可接受的机器人语音类型是成年男性和儿童语音。对于教育机器人，最可接受的语音类型是成年女性和男性语音。这项研究的结果有望为各种应用中的机器人语音设计构建设计原则。最可接受的语音类型是成年男性语音和儿童语音。对于家庭伴侣机器人，最可接受的机器人语音类型是成年男性和儿童语音。对于教育机器人，最可接受的语音类型是成年女性和男性语音。这项研究的结果有望为各种应用中的机器人语音设计构建设计原则。