Workshop: Robust Research in the Social, Behavioral, and Economic Sciences, February, 2014, Arlington, VA

该奖项支持为期两天的研讨会，以调查综合政策和程序，促进科学中强大和可靠的出版成果。近年来，人们对生物、医学、认知和社会科学中各种发现的可复制性提出了质疑。确保科学家理解并因使用产生最可靠结果的方法而获得奖励，并激励结果的复制，这是许多可能的行动之一，有助于确保已发表的结果高度可靠。本次研讨会将探讨这些和其他可能采取的措施，以提高科学发现的有效性。科学家们在基础和应用研究中建立在已发表的科学知识基础上。当这些发表的结果不可靠时，科学发现可能会受到挫折，时间和金钱会花在非生产性的研究上。此外，科学经常被用作政策和行业决策的基础，这也需要可靠性。社会和行为科学家对实验和准实验设计、方法程序（例如，随机分配盲法），和统计学（例如，元分析）技术。拟议的研讨会将召集社会，行为和统计科学家考虑：（i）科学现状，（ii）科学实践和程序的可能改进，（iii）对科学教育和培训的影响，（iv）对编辑政策和程序的影响，（v）对研究型大学政策和评估标准的影响，以及（vi）对联邦资助政策和评估标准的影响。本次研讨会将汇集这些领域的专家，制定一套连贯的建议，以确保在社会，行为和经济科学中发表的实证研究结果是可靠的和可复制的，并确定对这一主题的研究需求。研讨会将编写一份书面报告，并广泛分享介绍材料。这项工作将由复制小组委员会的成员领导，该小组委员会是国家科学基金会社会、行为和经济科学理事会咨询委员会的一部分。

This award supports a two-day workshop to investigate integrative policies and procedures that would promote robust and reliable published findings in science. Questions have been raised in recent years about the replicability of various findings in the biological, medical, cognitive, and social sciences. Making sure that scientists understand and are rewarded for using methods that yield the most robust results and incentivizing replication of results are among the many possible actions that could help ensure that published results are highly reliable. This workshop will explore these and other steps that might be taken to improve the validity of scientific findings.Scientists build upon published scientific knowledge in both basic and applied research. When those published results are not reliable, scientific discoveries can be set back and time and money spent on non-productive avenues of research. Additionally, science is often used as a basis for decisions in policy and industry, which also necessitate reliability. Social and behavioral scientists have contributed to improvements in experimental and quasi-experimental designs, methodological procedures (e.g., random assignment & blind procedures), and statistical (e.g., meta-analytic) techniques. The proposed workshop will convene social, behavioral, and statistical scientists to consider: (i) the state of the science, (ii) possible improvements in scientific practice and procedures, (iii) the implications for science education and training, (iv) the implications for editorial policies and procedures, (v) the implications for research university policies and evaluation criteria, and (vi) the implications for federal funding policies and evaluation criteria. This workshop will bring together experts in each of these domains to develop a coherent set of recommendations for ensuring that published empirical findings in the social, behavioral, and economic sciences are robust and replicable and to identify research needs on this topic. A written report from this workshop will be produced and presentation materials shared widely. The effort will be led by members of the Subcommittee on Replication, a part of the Advisory Committee of the Directorate for Social, Behavioral, and Economic Sciences at the National Science Foundation.