HSD: An Exploratory Space-Time Data Analysis Toolkit for Spatial Social Science Research

对区域经济增长、不平等、差异和趋同的研究引起了多个社会科学领域的极大兴趣。根据定义，这些分析依赖于空间引用的数据。然而，只有少数最近的研究注意到空间依赖性和空间异质性在区域经济演变的实证分析中的作用。地理信息系统（GIS）、空间统计学和空间计量经济学领域的研究已经产生了新的方法来处理这些空间效应，但这些方法没有解决区域经济变化的动态维度。一门真正整合的社会科学需要一个集社会经济现象的空间和时间维度于一体的工具包。本研究项目的目标是通过(1)研究空间聚类和空间异质性对动态环境下探索性数据分析（EDA）技术应用的影响来开发这样一个工具包；(2)为探索性时空数据分析（ESTDA）开发新的统计方法；(3)将这些方法实现在一个开源包中，用于社会过程的探索性时空分析。所使用的方法包括探索性空间数据分析（ESDA）、探索性时间数据分析（ETDA）、新的ESTDA方法的经验特性蒙特卡罗模拟研究、面向对象编程和动态地理可视化。该项目将对空间社会科学的实践、人类动力学的建模以及对区域增长和不平等本质的基本理解做出重大贡献。将空间和时间纳入区域不平等和增长的模型将提供对人类社会和经济行为更全面和准确的描述。该项目将开发一个探索性时空工具包，该工具包将是开源的，可供广泛的社会科学研究人员使用，以增强对人类和社会动态的分析。因此，这项研究预计将对城市隔离模式、时空流行病学和公共卫生、犯罪学、住房市场动态、社会经济不平等等领域的研究产生影响。从政策角度来看，制定新的空间明确措施将使规划人员和分析人员有能力设计针对个别关键地理区域的政策干预措施。通过考虑空间溢出效应，这一以空间为重点的战略将使此类政策方案的影响跨越单一地区的边界，从而提高政策的有效性。该项目得到了2004财年美国国家科学基金会人类与社会动力学（HSD）竞赛的支持。HSD竞赛和HSD奖项组合的协调管理涉及所有NSF董事和办公室。

The study of regional economic growth, inequality, divergence and convergence attracts considerable interest across multiple social sciences. By definition, these analyses rely on data that are spatially referenced. Only a few very recent studies, however, have given attention to the role of spatial dependence and spatial heterogeneity in the empirical analysis of regional economic evolutions. Research in the fields of geographical information systems (GIS), spatial statistics, and spatial econometrics has generated new methods designed to treat these spatial effects, but these methods do not address the dynamic dimensions of regional economic change. A truly integrated social science requires a toolkit that integrates both the spatial and temporal dimensions of socioeconomic phenomena. The objectives of this research project are to develop such a toolkit by (1) examining the implications of spatial clustering and spatial heterogeneity for the application of exploratory data analysis (EDA) techniques in a dynamic context; (2) developing new statistical methods for exploratory space-time data analysis (ESTDA); and (3) implementing these methods in an Open Source package for exploratory space-time analysis of social processes. The methods to be used include exploratory spatial data analysis (ESDA), exploratory temporal data analysis (ETDA), Monte Carlo simulation studies of the empirical properties of the new ESTDA methods, object-oriented programming, and dynamic geovisualization.This project will make significant contributions to the practice of spatial social science, the modeling of human dynamics, and to basic understanding of the nature of regional growth and inequality. The incorporation of space and time into models of regional inequality and growth will provide more comprehensive and accurate descriptions of human social and economic behavior. The project will develop an exploratory space-time toolkit that will be Open Source and accessible to a broad array of social science researchers to enhance the analysis of human and social dynamics. As such, this research is expected to have implications in areas such as studies of urban segregation patterns, space-time epidemiology and public health, criminology, housing market dynamics, socioeconomic inequalities, among others. From a policy perspective, the development of new spatially explicit measures will provide planners and analysts with capabilities to design policy interventions targeted at key individual geographical areas. By taking spatial spillovers into account, this spatially focused strategy will leverage the impact of such policy programs across the boundaries of a single area, thereby increasing the effectiveness of the policy. This project is supported by an award resulting from the FY 2004 NSF-wide competition on Human and Social Dynamics (HSD). Coordinated management of the HSD competition and the portfolio of HSD awards involves all NSF directorates and offices.