I have seen people use ANOVA to compare models to see if adding additional variables is significant, such as this post:
Significance of categorical predictor in logistic regression
Why ANOVA can be used to compare models? What assumptions is it based on?