Hi,

Here is a quick query....

If for a Logistic Regression Test, the Whole Model Test is significant (p<0.05) and the Lack of Fit test is also significant (p<0.05), with just one parameter, what should be our interpretation and next steps. I was using the test to validate one of the Xs.

Nirankar

Hello,

Could someone help me solve a probability equation in binary logistic regression when one of the Xs is a nominal data i.e. a day.

While finding out the probability of the event we solve below equation:

log(p/1-p) = a + bx1 + cx2+ ......

Here if X1 is a nominal variable ie. above equation cannot be solved as "nominal" variable has no value that we can enter in the equation

Please help.. if I am missing onto some important info about of treatment of "nominal variables" in logistic regression equation

Thanks

Deepak Ghalsasi