Science Policy Research Report: Funding Team Science

团队科学是两个或两个以上的研究人员为了共同的研究目标而相互依赖地合作。随着资助机构继续鼓励和促进团队科学，评估和综合我们所知道的团队规模的后果是很重要的。在财政资源匮乏的情况下，一个重要的科学政策问题是如何有效地将资金分配给研究团队。例如，给定一个150万美元的整体项目预算，什么时候以50万美元的价格资助一个由3名研究人员组成的团队，而不是以15万美元的价格资助一个由10名研究人员组成的团队。一方面，资助更少的研究人员（每个研究人员有更多的钱）增加了他们的动机，使协调更容易，但减少了可用的团队专业知识的数量。另一方面，用更少的钱资助更多的研究人员，减少了他们的激励，使协调更加困难，但增加了可用的团队专业知识的数量。通过对团队科学学术文献的跨学科回顾，该项目提高了我们对资助不同规模团队的收益和成本的认识。相关主题也将被探讨，包括团队复杂性的作用（例如，学科数量和机构数量），团队的生命周期，例如，团队生命周期早期的资助与团队生命周期后期的资助，以及研究发展阶段，例如，资助早期研究与资助后期研究。为了确保所确定的证据是稳健和可靠的，将使用三类来对已知的团队规模进行分类：(1)强因果推理（例如，实验室实验和现场实验），(2)中等因果推理（例如，具有适当控制的纵向研究）和(3)弱因果推理（例如，横断面调查和观察性研究）。在此基础上，将提出政策建议，指导团队科学资源的有效配置，以促进复杂问题的解决、新的研究发现和技术创新。

Team science is the collaboration of two or more researchers working interdependently towards a common research goal. As funding agencies continue to encourage and promote team science, it is important to evaluate and synthesize what we know about the consequences of team size. Given scare financial resources, an important science policy issue is how to effectively allocate funding to research teams. For example, given an overall project budget of $1.5M, when is it better to fund a team of three researchers at $500,000 each versus a team of 10 researchers at $150,000 each. On the one hand, funding fewer researchers (with more money per researcher) increases their incentives and makes coordination easier, but reduces the amount of team expertise available. On the other hand, funding more researchers, with less money per researcher, decreases their incentives and makes coordination more difficult, but increases the amount of team expertise available. Through an interdisciplinary review of the academic literature on team science, this project advances our knowledge of the benefits and costs of funding teams of varying size. Related topics will be explored as well, including the role of team complexity (e.g., number of disciplines and number of institutions), the life cycle of teams, for example, funding early in the life of the team vs funding later in the life of the team, and the stage of research development, for example, funding early-stage research vs funding late-stage research. To ensure that the evidence identified is robust and reliable, three categories will be used to classify what is known about team size: (1) strong causal inference (e.g., laboratory experiments and field experiments), (2) moderate causal inference (e.g., longitudinal studies with appropriate controls), and (3) weak causal inference (e.g., cross-sectional surveys and observational studies). Based on the evidence, policy recommendations will be made to guide the efficient allocation of resources to team science in order to facilitate complex problem solving, new research discoveries, and technological innovation.