In logistic regression, if the model is not a good fit, which options below might improve the goodness of fit? Select ALL correct answers. Consider some transformation on the predictors Discard some predictors with small estimated coefficients Consider other link functions. O Use methods that correct for overdispersion