Logistic regression is used in categorical response variable models. The explanatory variables could be categorical or in a continuous scale. Often, parameters are estimated and goodness-of-fit are studied in logistic models, but the ultimate usefulness of a logistic model depends on the prediction rule of the categorical response. Here all found aspects of the logistic regression using a data set were revisited. The four aspects are parameter estimation, goodness-of-fit of the model, prediction rule, and marginal effect of the factors.