The article presents a method to find and deal with outliers in data used for regression analysis. The method involves two stages: first, identifying and removing outliers, and second, analyzing the remaining data to see how the removed outliers affect the results. By using a robust distance estimator and bootstrapping techniques, the method can accurately classify different types of outliers and leverage points. The researchers applied this method to four examples and found that it effectively identifies outliers, even when they are hard to detect. The interactive nature of the method allows users to gain a better understanding of the data compared to using automated procedures.