Team Dynamics, Networks, and Assembly (Team DNA)

描述（由申请人提供）：在医学实践、医学和公共卫生研究以及更普遍的科学领域，相对于单独的科学家，团队发挥着越来越重要的作用。本研究的目的是了解成功团队的特征、即将发生的团队失败的主要指标，以及提高团队科学和问题解决效率的潜在政策。为了实现这些目标，我们将开发一个团队动态、网络和装配的计算模型，这些模型由团队协作的真实世界数据参数化。我们的研究将推进对团队科学的理解，包括整个过程，从团队的组建开始，继续通过团队沟通和解决问题的动态，并以团队产出或团队过早解散结束。与之前的模型不同，在这些模型中，团队网络是使用风格化的数学结构（如优先依恋或Watts-Strogatz小世界网络模型）外源性地指定的，我们模型中的团队将从潜在的合作者池中内源性地自我组装。我们的团队组合模型将由来自NSF支持的网络平台Nano HUB的独特数据集参数化，该数据集包括数千名在纳米技术领域工作的科学家的互动和科学成果。此外，我们的模型将不仅包括团队成员之间的通信网络结构，还包括这些通信的质量。由于我们可以访问Nano HUB团队成员之间的完整通信日志，因此我们将能够应用包括语言风格匹配（LSM）和情感激活和情感分析在内的通信质量度量来参数化我们计算模型中的模拟通信。除了使用团队互动的真实数据输入参数化我们的模型外，我们还将使用两个额外的团队协作数据集从外部验证计算模型，一个来自跨学科烟草使用研究中心（TTURC），另一个来自基于web的软件开发主机GitHub。由此产生的经过验证的计算模型不仅可以作为团队动力学、网络和集合理论的测试，还可以作为试验潜在政策干预的试验台，以鼓励成功团队的发展，并防止或减轻团队失败造成的损害。

DESCRIPTION (provided by applicant): Teams play an increasingly dominant role relative to solo scientists in medical practice, medical and public health research, and science more generally. The aim of this research is to understand the characteristics of successful teams, the leading indicators of impending team failure, and potential policies for increasing the productivit of team science and problem solving. To accomplish these aims we will develop a computational model of team dynamics, networks, and assembly parameterized by real world data on team collaboration. Our research will advance the understanding of team science by including the entire process that begins with the assembly of a team, continues through the dynamics of team communication and problem solving, and ends with either the production of team output or the premature dissolution of the team. Unlike previous models in which team networks are specified exogenously using stylized mathematical structures (such as preferential attachment or the Watts-Strogatz small world network model), teams in our model will endogenously self assemble from a pool of potential collaborators. Our model of team assembly will be parameterized by a unique dataset from the NSF supported web platform Nano HUB, which includes the interactions and scientific output of thousands of scientists working in the field of nanotechnology. Moreover, our model will incorporate not only the network structure of communication among team members, but also the quality of those communications. Because we have access to the full log of communications among Nano HUB team members, we will be able to apply measures of communication quality including Linguistic Style Matching (LSM) and Emotional Activation and Sentiment Analysis to parameterize the simulated communications in our computational model. In addition to parameterizing our model with input from real word data on team interactions, we will externally validate the computational model using two additional datasets on team collaboration, one from the Trans disciplinary Tobacco Use Research Center (TTURC) and a second from the web based software development host GitHub. The resulting validated computational model will serve not only as a test of theories of team dynamics, networks, and assembly, but also as a test bed for experimenting with potential policy interventions to encourage the development of successful teams and to prevent or mitigate the damage from team failures.