In examinign LR results it is very important to look at the 95% CI no matter how large the OR is or if the significance value is less than 0.05. Convention is that if the 95% CI includes the value of 1.0 then the variable is not considered a useful predictor of the dichotomous outocme variable you are trying to predict.
Martin Sherman wrote:
> Dear list,
> I got the following output from a multiple logistic regression (SPSS11.0)and I am not sure how to explain the results. Two variables were entered (stepwise) into the equation, yet one of the variables has a significance level of .711 with an Exp(B) of 9564 and a 95% CI of .00 to 1.02E+25. In addition, there is a variable which is not included but which has a significance of .004 (it is included in the list of variables not in the equation). Could multicollinearity explain why this is happening. TIA
> martin sherman
"Make it idiot-proof and someone will make a better idiot".
Michael Kruger "A True Prince"
Wayne State Univ.School of Medicine
Dept. of OB/Gyn