首先简要介绍了两种多元统计分析方法———因子分析和典型相关分析 ,之后在对城市居住实态调查资料进行因子分析的基础上 ,运用典型相关分析方法做了进一步的空间构形分析 ,给出了变量之间的关联 ,可供寻找因果关系 ,并对今后的调查组织工作起到一定的指导和借鉴作用。 First of all, two kinds of multivariate statistical analysis methods, factor analysis and canonical correlation analysis, are briefly introduced. Then, based on the factor analysis of urban residential survey data, the paper further analyzes the spatial configuration by using the method of canonical correlation analysis, Given the relationship between variables, we can find the causal relationship, and play a guiding role in the future investigation and organization.