Tuncer Sylvaine, Gillet Sarah, Leite Iolanda
Royal Institute of Technology (KTH), Stockholm, Sweden.
Front Robot AI. 2022 Mar 3;9:729146. doi: 10.3389/frobt.2022.729146. eCollection 2022.
Our work is motivated by the idea that social robots can help inclusive processes in groups of children, focusing on the case of children who have newly arrived from a foreign country and their peers at school. Building on an initial study where we tested different robot behaviours and recorded children's interactions mediated by a robot in a game, we present in this paper the findings from a subsequent analysis of the same video data drawing from ethnomethodology and conversation analysis. We describe how this approach differs from predominantly quantitative video analysis in HRI; how mutual gaze appeared as a challenging interactional accomplishment between unacquainted children, and why we focused on this phenomenon. We identify two situations and trajectories in which children make eye contact: asking for or giving instructions, and sharing an emotional reaction. Based on detailed analyses of a selection of extracts in the empirical section, we describe patterns and discuss the links between the different situations and trajectories, and relationship building. Our findings inform HRI and robot design by identifying complex interactional accomplishments between two children, as well as group dynamics which support these interactions. We argue that social robots should be able to perceive such phenomena in order to better support inclusion of outgroup children. Lastly, by explaining how we combined approaches and showing how they build on each other, we also hope to demonstrate the value of interdisciplinary research, and encourage it.
我们的工作基于这样一种理念,即社交机器人可以帮助促进儿童群体中的包容过程,重点关注刚从国外来到的儿童及其在学校的同龄人。基于一项初步研究,我们在该研究中测试了不同的机器人行为,并记录了儿童在游戏中通过机器人介导的互动。在本文中,我们展示了从同一视频数据的后续分析中得出的结果,该分析借鉴了民族方法学和对话分析。我们描述了这种方法与HRI中主要的定量视频分析有何不同;相互凝视是如何作为不熟悉的儿童之间具有挑战性的互动成就出现的,以及我们为什么关注这一现象。我们识别出儿童进行眼神接触的两种情况和轨迹:请求或给予指示,以及分享情感反应。基于实证部分对一系列节选内容的详细分析,我们描述了模式,并讨论了不同情况和轨迹之间的联系以及关系建立。我们的研究结果通过识别两个儿童之间复杂的互动成就以及支持这些互动的群体动态,为HRI和机器人设计提供了信息。我们认为社交机器人应该能够感知到此类现象,以便更好地支持外来儿童的融入。最后,通过解释我们如何结合不同方法并展示它们如何相互补充,我们也希望展示跨学科研究的价值,并鼓励开展此类研究。