中文摘要：

利用2006年全国各省农村居民人均住房面积统计数据，基于空间自相关方法从空间尺度分析其空间分布特征。首先建立空间邻近网络和邻接矩阵，通过Moran′I自相关指数分析空间关联性，为确定农村人均住房面积的分布存在空间相关性而非随机性，采用蒙特卡罗方法进行多次模拟证明，并利用2011年统计数据进行实际验证。实验分析，2006与2011年农村人均住房面积Moran′I指数分别为0．5566和0．5983，表明空间聚集特点明显，基本可以分为东部发达地区、中部和西部三种聚类，并且两年之间聚集现象有显著变化。

英文摘要：

The spatial distribution characteristics of per capita housing area in Chinese mainland provinces ′rural parts are analyzed under the spatial autocorrelation method using statically data of 2006 as a case study .Firstly spatial neighbor network and neighbor matrix are built as input parameters for the Moran′s I index calculation .Then Monte Carlo simulation is conducted multi times to esti-mate the result of this spatial autocorrelation analysis , in addition taking data of 2011 as a comparison test which also shows the same distribution model.The Moran′s I index for the year 2006 and 2011 separately is 0.4450 and 0.4466 that indicates there is an overt spatial cluster condition .The result of this research shows that there are mainly three cluster parts：east-china, middle-china, west-china and some changes happened in the cluster model between 2006 and 2011 .