The Influence of Virtual Agents' Gender and Rapport on Enhancing Math Performance
The Influence of Virtual Agents' Gender and Rapport on Enhancing Math Performance
复制标题
虚拟代理的性别和融洽关系对提高数学成绩的影响
作者：
Bilge Karacora;Morteza Dehghani;Nicole Krämer;J. Gratch
虚拟代理的性别和关系对提高数学成绩的影响Bilge Karacora 1 (bilge.karacora@stud.uni-due.de), Morteza Dehghani 2 (morteza@ict.usc.edu), Nicole Kramer-Mertens 1 (nicole.kraemer@uni-due.de), Jonathan Gratch 2 （gratch@ict.usc.edu）杜伊斯堡-埃森大学社会心理学系，Forsthausweg 2, 47057德国杜伊斯堡，南加州大学创新技术研究所，12015 Waterfront Dr, Playa Vista, CA 90094-2536，USA Rapport已被证明是在虚拟代理中创建行为现实性的有效方法。在社会心理学中，融洽关系被描述为通过快速发现和回应对方的非语言行为在互动伙伴之间建立一种积极的关系（Gratch et al., 2007a）。这包括表现出表明积极情绪的行为（如点头和微笑），表现出相互关注（如相互凝视）和某些协调行为（如姿势模仿和同步运动）（Tickle-Degnan & Rosenthal, 1990）。Niewiadomski等人（2010）报告说，当一个代理人表现出适当的和适应社会的情感表达时，他被认为比表现出不适当的或不适应社会的人类表达的代理人更像人类。Garau等人（2005）进行的一项研究表明，只有与对其动作有反应的代理互动的参与者，才会体验到与代理的个人接触感，这种接触感影响了他们的行为，从而使他们表现得更体贴，而不是与静态或移动但无反应的代理互动。这表明，融洽关系是一个重要的特征，以使代理人被视为人类，并使任何社会影响，如社会促进，发生。先前关于社会促进/抑制的研究表明，他人的存在如何积极或消极地影响个人的任务表现（Guerin & Innes, 1982; Zanjonc, 1965; Sanders, Baron & Moore, 1978）。是否类似的促进/抑制效应发生在虚拟代理的存在已经受到一些研究。Rickenberg和Reeves（2000）发现，虚拟代理的“社交”存在会促进或抑制任务。Zanbaka等人（2004）的一项研究表明，当被要求执行一项任务时，参与者对虚拟代理的反应与他们面对另一个人的反应相似。Zanbaka等人（2007）的后续研究表明，虚拟代理的存在抑制了参与者在数学任务中的表现。本研究的局限性在于，样本中只有女性参与者面对的是与性别匹配的代理人。摘要本研究的目的是调查虚拟代理是否有助于提高参与者在数学方面的表现、努力和动机。我们假设，在人机交互中，任何社会效应的产生都需要由融洽关系的表现所引起的最低限度的行为现实性。此外，我们研究了社会促进效应是否取决于参与者的性别和相互作用的虚拟代理。在受试者之间的2x2设计中，参与者与一个男性或女性虚拟代理互动，该虚拟代理要么表现出融洽的关系，要么表现出不融洽的关系。我们的研究结果证实，在与能够建立融洽关系的虚拟代理互动时，性别起着一定的作用。当参与者与表现出融洽关系的异性代理人互动时，他们的表现和努力都得到了显著提高。我们的结果对设计教育和培训目的的代理具有启示意义。关键词：社会促进，STEM，融洽关系，虚拟代理引言对提高科学和数学成绩的因素有相当大的兴趣。最近，人们对利用社会和动机因素提高数学成绩，特别是减少性别不平等的教育技术产生了极大的兴趣（Kim, 2004; Baylor & Ryu, 2003）。这项工作建立在人们经常将计算机视为社会行动者的现象之上。因此，在传统的面对面环境中，提高人们表现的心理因素可以通过虚拟学习同伴或虚拟教师的形式被技术模拟。在本文中，我们试图解决两个相关的目标。首先，我们的目标是证明某些社会心理现象可以在人机环境中提高数学成绩。具体来说，我们展示了一种形式的社会促进可以提高标准化数学测试的表现。其次，我们寻求提供进一步的证据，证明人们确实将计算机视为社会行动者，并帮助阐明促进这种效应的设计原则。我们特别证明，虚拟代理必须具有最低水平的行为现实主义，以实现任何社会效应。
The Influence of Virtual Agents’ Gender and Rapport on Enhancing Math Performance Bilge Karacora 1 (bilge.karacora@stud.uni-due.de), Morteza Dehghani 2 (morteza@ict.usc.edu), Nicole Kramer-Mertens 1 (nicole.kraemer@uni-due.de), Jonathan Gratch 2 (gratch@ict.usc.edu) Department of Social Psychology, University of Duisburg-Essen, Forsthausweg 2, 47057 Duisburg, Germany Institute for Creative Technologies, University of Southern California, 12015 Waterfront Dr., Playa Vista, CA 90094-2536, USA Rapport has been shown as an effective way to create behavioral realism in virtual agents. In social psychology, rapport is described as the establishment of a positive relationship among interaction partners by rapidly detecting and responding to each other’s nonverbal behavior (Gratch et al., 2007a). This includes displaying behaviors that indicate positive emotions (such as head nods and smiles), showing mutual attentiveness (such as mutual gaze) and certain coordination behaviors (such as postural mimicry and synchronized movement) (Tickle-Degnan & Rosenthal, 1990). Niewiadomski et al. (2010) reports that when an agent displays appropriate and socially adapted emotional expressions, he is perceived as more human-like than an agent that shows human expressions which are inappropriate or not socially adapted. Garau et al. (2005) conducted a study showing that only participants who interacted with an agent that was responsive to their movements, experienced a sense of personal contact with the agent which influenced them to behave more socially considerate as opposed to interacting with a static or moving but unresponsive agent. This indicates that rapport is an important feature in order for the agent to be perceived as human-like and for any social effects, such as social facilitation, to occur. Previous research on social facilitation/inhibition illustrates how the presence of others affects an individuals’ task performance either positively or negatively (Guerin & Innes, 1982; Zanjonc, 1965; Sanders, Baron & Moore, 1978). Whether or not similar facilitation/inhibition effects occur in presence of virtual agents has been subject to several studies. Rickenberg and Reeves (2000) found that tasks are facilitated or inhibited by the “social” presence of a virtual agent. A study by Zanbaka et al. (2004) indicates that when asked to perform a task, participants reacted similarly to the presence of a virtual agent as they would have in the presence of another human. A follow-up study by Zanbaka et al. (2007) demonstrates that the presence of a virtual agent inhibits the performance of participants on a mathematical task. The limitation of this study was that the sample consisted of only female participants being confronted with an agent of matching gender. Hayes et al. (2010) found a similar decrease in performance with regard Abstract The purpose of the present research is to investigate whether virtual agents can help enhance participants’ performance, effort and motivation in mathematics. We hypothesize that a minimal amount behavioral realism induced by display of rapport is necessary for any social effects to occur in human- computer interaction. Further, we examine whether social facilitation effects occur depending on the gender of the participants and the interacting virtual agents. In a 2x2 between subjects design, participants interacted with a male or female virtual agent that either displayed rapport or no rapport. Our results confirm that gender plays a role when interacting with virtual agents that are capable of establishing rapport. Participants’ performance and effort were significantly enhanced when interacting with an agent of opposite gender that displayed rapport . Our results have implications on designing agents for education and training purposes. Keywords: social facilitation, STEM, rapport, virtual agents Introduction There is considerable interest in factors that enhance science and math performance. Recently, there has been an upsurge of interest in educational technology that exploits social and motivational factors that enhance math performance in general, and reduce gender inequality in particular (Kim, 2004; Baylor & Ryu, 2003). This work builds on the phenomena that people often treat computers as social actors. Therefore, psychological factors that improve people's performance in traditional face-to-face settings can be simulated by technologies in form of virtual learning companions or virtual instructors. In this paper, we seek to address two related goals. First, we aim to show that certain social psychological phenomena can enhance math performance in a human-computer setting. Specifically, we show that a form of social facilitation can improve performance on standardized math tests. Second, we seek to provide further evidence that people do treat computers as social actors and help elucidate the design principles that foster this effect. We specifically demonstrate that virtual agents must possess a minimum level of behavioral realism to achieve any social effects.