用SPSS中19的logistic 回归希望能得出各个自变量的参数值等，但是呈现出却是这一大堆估值，请指教

显示全部楼层 · 2011-12-9 07:30:57

参数估计Exp(B)的置信区间95%社会满意度二分变量aB标准误Walddf显著水平Exp(B)下限上限0截距-38.2131832.376.0001.983[网络事件关注=1]4.1871.4698.1281.00465.8143.7001170.527[网络事件关注=2]1.820.8694.3921.0366.1741.12533.881[网络事件关注=3]2.151.8795.9921.0148.5981.53548.143[网络事件关注=4]0b..0....[校园满意度汇总=1.86]36.1303669.889.0001.9924.909E15.000.c[校园满意度汇总=2.00]-1.0243303.464.00011.000.359.000.c[校园满意度汇总=2.29]36.1302900.490.0001.9904.909E15.000.c[校园满意度汇总=2.43]18.4171832.376.0001.99299660118.576.000.c[校园满意度汇总=2.57]37.3642836.036.0001.9891.686E16.000.c[校园满意度汇总=2.71]36.1303669.889.0001.9924.909E15.000.c[校园满意度汇总=2.86]20.2911832.376.0001.9916.487E8.000.c[校园满意度汇总=3.00]20.4491832.376.0001.9917.601E8.000.c[校园满意度汇总=3.14]20.2821832.376.0001.9916.431E8.000.c[校园满意度汇总=3.29]19.3761832.375.0001.9922.600E8.000.c[校园满意度汇总=3.43]19.4881832.375.0001.9922.907E8.000.c[校园满意度汇总=3.57]18.9551832.375.0001.9921.707E8.000.c[校园满意度汇总=3.71]18.8431832.376.0001.9921.526E8.000.c[校园满意度汇总=3.86]2.3402342.721.0001.99910.377.000.c[校园满意度汇总=4.00]17.7141832.375.0001.99249325299.713.000.c[校园满意度汇总=4.14]16.9621832.375.0001.99323251602.810.000.c[校园满意度汇总=4.43]1.6783664.873.00011.0005.356.000.c[校园满意度汇总=4.57]-.6883303.464.00011.000.502.000.c[校园满意度汇总=5.00]0b..0....[家庭状况=2]-1.515.8353.2921.070.220.0431.129[家庭状况=3].213.556.1461.7021.237.4163.676[家庭状况=4]0b..0....[政治面貌=1]19.649.904472.2401.0003.415E858036750.6192.009E9[政治面貌=2]16.914.576861.7771.00022167105.4727165900.72768572058.661[政治面貌=3]17.277.000.1.31871631.32231871631.32231871631.322[政治面貌=4]0b..0....a.参考类别是:1。

千问 · 2011-12-9 07:30:57

B列是回归系数，Exp(B)是OR值，显著水平是P值，logistic回归中真正有现实意义的是OR值而非回归系数。校园满意度汇总分类过多，从你的数据看，校园满意度应该按连续变量处理。即使校园满意度是分类变量，但它们按分值明显存在对应关系，所以也可以按连续变量处理。

千问 · 2011-12-9 07:30:57

列B就是对应的回归系数赞同