Identifying multimodal dynamics of coordination to understand joint performance in diverse tasks

人们需要协调他们的行动才能进行广泛的日常活动。 例如，一起做饭、搬一张大沙发、在运动或商业中作为团队合作，都需要参与者之间的协调。尽管有证据表明，这些联合任务中的合作伙伴会实时调整彼此的行为，但尚不清楚哪些协调特征会带来最佳结果。该项目的目的是开发一个框架来预测不同环境和角色中人们之间的最佳协调。通过结合实证研究和动态计算模型，研究人员将解决围绕对话、观点采择和协作在成功执行联合任务中的作用的问题。了解参与联合任务的人员之间成功协调的原则在各种环境中具有潜在的商业和社会影响，包括医疗保健、教育、航空、军事行动和搜救场景。 该项目还将通过增加两个学术机构的本科生和研究生参与认知科学研究的机会，为跨学科劳动力的发展做出贡献。 关于人际协调如何影响任务绩效的一个突出观点是，当任务伙伴调整或匹配他们的行为时，他们的联合任务绩效就会提高。 例如，发音、单词选择、句子结构或动作的一致性已记录在需要合作伙伴监控彼此观点的任务中。然而，尚不清楚不同的任务如何增加或减少成功绩效的一致性。在这个项目中，研究人员将操纵许多任务特征：任务的目标、合作伙伴角色的对称性以及合作伙伴必须采用的观点数量。将分析眼动测量和对话记录，以量化合作伙伴在注意力和语言使用方面的一致性。目的是评估和建模如何通过任务特征和人际协调程度来预测任务绩效。例如，在需要合作伙伴相互密切监督的任务中（例如，当合作伙伴因为持有不同的信息而具有不对称的角色时），人际一致性可能会增加并且更有用。在需要较少监控的任务中，过度调整可能会不利于任务结果。协作绩效的动态模型具有任务约束和人际协调的参数，将根据实验数据进行评估，并将用于生成有关新任务绩效的预测。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

People need to coordinate their actions in order to carry out a wide range of daily activities. For example, cooking a meal together, moving a large couch, and working as a team in sport or business, all require coordination among the participants. Despite evidence that collaborating partners in these joint tasks adapt their behavior to one another in real time, it is unclear what features of coordination lead to the best outcomes. The aim of this project is to develop a framework for predicting optimal coordination among people across diverse settings and roles. Using a combination of empirical studies and dynamical computational modeling, the investigators will unpack issues surrounding the roles of dialogue, perspective-taking, and collaboration in successfully performing joint tasks. Understanding the principles that underlie successful coordination among people engaged in a joint task has potential commercial and societal impact in a wide variety of settings, including health care, education, aviation, military operations, and search-and-rescue scenarios. The project will also contribute to the development of an interdisciplinary workforce by increasing opportunities to engage undergraduate and graduate students in cognitive science research at two academic institutions.A prominent view of how interpersonal coordination influences task performance is that when task partners align or match their behavior, their joint task performance improves. Alignment of, for example, pronunciation, word choices, sentence structure, or movement has been documented in tasks that require partners to monitor each other’s perspective. However, it is unknown how different tasks may increase or decrease alignment for successful performance. In this project, the investigators will manipulate a number of task features: the goals of the task, the symmetry of the partners’ roles, and the number of perspectives that partners have to adopt. Eye movement measurements and dialogue transcripts will be analyzed in order to quantify how aligned partners are in their attention and language use. The aim is to evaluate and model how task performance is predicted by task features and by the degree of interpersonal coordination. For example, interpersonal alignment might increase and be more useful in tasks that require partners to monitor one another closely (e.g., when partners have asymmetrical roles because they hold distinct information). In tasks that require less monitoring, excessive alignment might be detrimental to task outcomes. A dynamical model of collaborative performance, which has parameters for task constraints and interpersonal coordination, will be evaluated against the experimental data and will be used to generate predictions about performance in new tasks.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.