一种新的考虑时间和空间的相关系数及其算例  

A New Correlation Coefficient Considering Space-Time Factors and Its Illustration

作　　者：宋马林[1] 王舒鸿[2] 汝慧萍[2] 

机构地区：[1]安徽财经大学统计与应用数学学院 [2]南开大学经济学院数量经济研究所

出　　处：《数量经济技术经济研究》2010年第7期142-152,共11页Journal of Quantitative & Technological Economics

基　　金：国家自然科学基金项目(70901069);安徽省社会科学规划项目(AHSK07-08D25);山东省科技攻关项目(2009GG10010003);山东省软科学项目(2008RKA086)

摘　　要：本文从时间和空间多维的视角,通过分析全局Moran'sI相关系数与局部Morans'I相关系数的特点,提出一种新的考虑时空的相关系数分析方法。然后将中国内地各省份的专利授权数指标的多维分析作为算例,应用这种新的时空相关系数进行定量分析,其结果不仅与Moran相关系数的分析结果高度正相关,并具有后者所不具有的能够同时分析空间和时间指标变化趋势的优势。同时考虑时间和空间因素的相关系数,为社会经济时空变化的空间计量分析提供了新的工具。From the perspective of multi - dimension of space and time, this article proposes a new computational method of correlation coefficient considering both temporal and spatial factors, based on the analysis of the characteristics of Global Moran＇s I correlation coefficient and local Moran＇s I correlation coeffi- cient. Then the patent authorization quantities of Chinese mainland＇s provinces and municipalities are taken as an example of multi - dimensional analysis. The results of the new quantitative analysis show that the outcomes of this new correlation coefficient is not only highly correlated with Moran＇s, but also have advantages, analy- zing the trends of both temporal and spatial indicators simultaneously. This new correlation coefficient may provide a more scientific and effective tool for spatial econometric analysis in time and space changes of society and economy.

关 键 词：时空相关系数 空间计量经济学 Moran相关系数 

分 类 号：F224.0[经济管理—国民经济]