Collaborative Research: Time-Sharing Experiments for the Social Sciences (TESS)

社会科学分时实验（英语：Time-Sharing Experiments for the Social Sciences，TESS）是一个对美国成年人进行基于概率的样本调查实验的平台。 TESS利用规模经济，使来自社会科学领域的学者能够在竞争的基础上，对具有广泛理论和实践重要性的问题进行开创性的研究。 TESS的智力价值与其成功密切相关，这些成功使社会科学家能够及时收集原始的基于人口的实验数据;促进对基本社会，政治和经济问题的更好理解;通过集中其他独立研究的费用来最大限度地提高财务效率;并提供指导和教育资源。 TESS的更广泛影响是广泛的。 TESS允许各种各样的学者进行调查实验，因为单个研究人员的成本是不可能的。 它还允许通过严格的审查程序和必要时的指导来改进这些学者的工作。 此外，在TESS框架下收集的所有数据都可以在一个存档中及时获得。TESS的当前迭代还包括在M-TURK上运行的实验以及欧洲基于概率的样本。 使用M-TURK将使学者们能够检查该数据收集系统与目前使用的基于概率的样本相比的有用性。 它将允许调查人员检查前者数据收集过程中的偏见，后者虽然便宜，但在某些目的上可能存在严重缺陷。 将欧洲基于概率的分析结果与美国进行比较，将使调查人员能够研究从美国数据得出的结果有多可靠。 在处理非政治案件的实验中尤其如此。

Time-Sharing Experiments for the Social Sciences (TESS) is a platform for conducting survey experiments fielded on probability-based samples of United States adults. TESS capitalizes on the economies of scale to enable scholars from across the social sciences, on a competitive basis, to conduct ground-breaking research on issues of broad theoretical and practical importance. The intellectual merit of TESS is closely connected to its successes in enabling social scientists collection original population-based experimental data in a timely manner; promoting better understanding of fundamental social, political, and economic questions; maximizing financial efficiency by pooling expenses for otherwise separate studies; and offering mentoring and educational resources. The broader impacts of TESS are wide-ranging. TESS has allowed a wide variety of scholars to undertake survey experiments that because of cost to single researchers would not be possible. It has also allowed for these scholar's work to be improved by a rigorous review process and mentoring where needed. In addition, all the data collected under the TESS framework is made available on a timely basis in a single archive.The current iteration of TESS also includes running experiments on M-TURK as well as probability-based samples in Europe. Using M-TURK will allow scholars to examine the usefulness of that data collection system as compared to the probability-based samples currently used. It will allow investigators to examine for biases in the data collection processes in the former, which while inexpensive may be severely flawed for some purposes. Examining the results of probability-based analysis in Europe as compared to the United States will allow investigators to study how robust results derived from United States data are. This is especially true of experiments dealing with non-political cases.