I'm having an issue with binary logistic regression for a project I'm working on.
For some of the variables, I'm receiving a significance value of 0.999 and exp(B) of 0. Is this normal?
Asked
Active
Viewed 516 times
0
kjetil b halvorsen
- 77,844
-
2It looks like what has happened is that for some combinations of your variables no events occurred. This has resulted in the model trying it's best to predict a probability of 0, but the standard errors have become extremely large as a consequence. Take a look at the answers to this question to see some more explanation. https://stats.stackexchange.com/questions/254124 – David Thiessen May 29 '22 at 20:28
-
Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. – Community May 29 '22 at 22:50
-
See also for some ideas https://stats.stackexchange.com/questions/260885/binomial-glm-non-significant-difference-between-100-opposite-groups-of-observ/260911#260911 – kjetil b halvorsen May 30 '22 at 17:31