<p>请问各位在用SPSS软件进行累积LOGISTIC回归后结果显示 The log-likelihood value is pratically zero. There maybe a complete seperation in the data. THe maximum likelihood estimate don't exist.是代表什么？这样构造的模型是否不可使用？</p><p>结果栏中还有显示There are 92(66.7%)cells (dependent variable levels by combination of predictor variable values)with zero frequency.这个对模型结果有影响吗？</p><p>感激不尽</p>