NCRN-MN: Improving the Interpretability and Usability of the American Community Survey Through Hierarchical Multiscale Spatio-Temporal Statistical Models

美国社区调查(ACS)是一项持续的调查，每年发布数据，为社区提供规划资源和服务分配所需的及时信息。根据美国人口普查局的说法，调查数据为每年超过4000亿美元的联邦和州资金的分配提供了信息。2010年，人口普查发布了与ACS有关的所有标准制表地区的第一个五年期间估计数。这标志着美国历史上一个有趣的时刻，因为人口普查从十年一次的人口普查长表数据转向使用每年发布数据的持续调查。这一转变给人口普查和数据用户带来了许多方法上的挑战。本项目中的分级多尺度时空统计模型和扩展的统计方法将解决其中的许多挑战，并促进更广泛和更有效地利用可持续发展系统。特别是，该项目将制定一个有效的框架，在进行小区域估计的同时，保留因在空中交通管制系统中发现的综合结构而产生的地理和时间限制。此外，通过借用空间和时间上的多个尺度和多个结果的力量，该方法将减少ACS小区域估计及其派生结果的差异。此外，从数据用户的角度来看，该方法将同时在几个时间尺度上提供一致的估计，而不是受到公布的多年估计的阻碍，使研究人员能够比较不同地理尺度和单位的趋势。本项目将研究在广泛的应用问题上提供新解决方案的方法。这项研究将通过开发分层多尺度时空统计模型来提高ACS的可解释性和可用性。此外，该项目将提供各种独立感兴趣的方法，并可用于人口普查和其他联邦统计机构管理的许多其他调查。一些拟议的方法还将直接延伸到疾病测绘领域，从而为公共卫生提供重要工具。该项目将对统计文献作出贡献，并将对政府机构和许多专题学科的工作有价值。该项目的主要重点之一将是教育和培训研究生和博士后研究人员。预计学生将获得可能融入联邦部门的必要技能。这项活动得到了NSF-人口普查研究网络资助机会的支持。

The American Community Survey (ACS) is an ongoing survey that releases data annually, providing communities with the timely information needed to plan the distribution of resources and services. According to the U.S. Census Bureau, the data from the survey provides input into how more than $400 billion in federal and state funds are distributed annually. Census released the first five-year period estimates associated with the ACS for all standard tabulation areas in 2010. This marks an interesting time in U.S. history as Census shifts from the decennial census long-form data to using an ongoing survey that releases data annually. Making this transition presents many methodological challenges, both for Census and for data users. The hierarchical multiscale spatio-temporal statistical models and expansive statistical methodology in this project will address many of these challenges and facilitate broader and more effective utilization of the ACS. In particular, this project will develop an efficient framework for carrying out small area estimation while preserving geographical and temporal constraints that arise from the aggregate structure found in the ACS. Further, by borrowing strength across multiple scales in space and time and multiple outcomes, the approach will reduce the variance in the ACS small area estimates and its derivatives. Additionally, from a data-user perspective, the methodology will simultaneously provide coherent estimates on several temporal scales rather than being hampered by the published multiyear estimates, allowing researchers to compare trends across different geographic scales and units.This project will investigate methodology that provides novel solutions across a wide-range of applied problems. The research will improve the interpretability and usability of the ACS through the development of hierarchical multiscale spatio-temporal statistical models. In addition, the project will provide a variety of methods that are of independent interest and can be used in many other surveys administered by Census and other federal statistics agencies. Several of the proposed methods also will directly carry over to the area of disease mapping and thus provide important tools for public health. The project will contribute to the statistics literature and will be of value to the work of government agencies and many subject-matter disciplines. One of the major focuses of this project will be to educate and train graduate students and postdoctoral researchers. It is expected that students will acquire the requisite skills for possible integration into the federal sector. This activity is supported by the NSF-Census Research Network funding opportunity.