多元统计分实验报告.docVIP

例5—2续 对例5—2的数据（数据略）作如下分析： 分析该组数据是否适合做因子分析。 若适合，做因子分析并做正交旋转，写出因子模型，并作适当解释（各因子个代表什么），计算各因子以及综合得分，作出得分图，进行评价。 与聚类分析结果进行比较。 具体分析如下： （1）输出结果1 Correlation Matrix x1 x2 x3 x4 x5 x6 x7 x8 x9 Correlation x1 1.000 .869 .770 -.053 .211 .920 .899 .795 .896 x2 .869 1.000 .978 .387 .472 .886 .804 .814 .849 x3 .770 .978 1.000 .523 .531 .797 .736 .740 .811 x4 -.053 .387 .523 1.000 .323 .115 -.023 .125 .051 x5 .211 .472 .531 .323 1.000 .175 .260 .371 .317 x6 .920 .886 .797 .115 .175 1.000 .877 .815 .768 x7 .899 .804 .736 -.023 .260 .877 1.000 .757 .818 x8 .795 .814 .740 .125 .371 .815 .757 1.000 .715 x9 .896 .849 .811 .051 .317 .768 .818 .715 1.000 有上面结果可知，原始变量之间相关系数多大于0.300，说明有较强的相关性，所以进行因子分析是合适的。 （2）输出结果2 Component Matrixa Component 1 2 3 x1 .931 -.315 -.012 x2 .976 .163 -.071 x3 .931 .322 -.090 x4 .232 .863 -.436 x5 .433 .596 .670 x6 .923 -.200 -.188 x7 .897 -.274 .035 x8 .871 -.064 .069 x9 .899 -.154 .062 Extraction Method: Principal Component Analysis. a. 3 components extracted. Communalities Extraction x1 .967 x2 .983 x3 .978 x4 .989 x5 .992 x6 .927 x7 .880 x8 .768 x9 .836 Extraction Method: Principal Component Analysis. Total Variance Explained Component Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % 1 6.150 68.332 68.332 5.679 63.096 63.096 2 1.473 16.365 84.698 1.402 15.577 78.673 3 .697 7.749 92.447 1.240 13.774 92.447 Extraction Method: Principal Component Analysis. Component Score Coefficient Matrix Component 1 2 3 x1 .151 -.214 -.017 x2 .159 .111 -.102 x3 .151 .219 -.129 x4 .038 .586 -.625 x5 .070 .405 .961 x6 .150 -.136 -.269 x7 .146 -.186 .051 x8 .142 -.043 .099 x9 .146 -.104 .089 Extraction Method: Principal Component Analysis. 根据输出结果2中Component Matrix 是因子载荷矩阵，用标准化的主成分（公共因子）近似表示标准化原始变量的系数矩阵，用fac1,fac2,fac3表示各公因子，x1（100元工业总产值实现利税（%））为例，即有 标准化的 得到初始载荷矩阵与公因子后，为了解释方便往往需要对因子分析进行旋转，用SPSS分析后得输出结果3 输出结果3 Rotated Component Matrixa Component 1 2 3 x1 .980 -.063 .056 x2 .884 .362 .267 x3 .