Tests to determine the goodness of fit of a model



Hi I am working on logistic regression . I have 4 independent variables. I modelled using all of them first , second iteration with 3 and 2 and last iteration with 1. But there wasn’t any change AIC ,SC values in all the cases. However there was impact on concordance. Can some one please tell what are other goodness of fit test we can run to be more confident of findings.

Also can someone please suggest a book for predictive analysis .



You can look at the multiple metrics to evaluate the performance of model. Below are the list of some metrics:

• Chi-square
• Classification tables
• ROC curves
• Logistic regression R-Square or Adjusted R-Square