This document discusses outlier detection and treatment using the interquartile range (IQR) method. It defines outliers as values that behave differently than other observations. As an example, it shows athlete performance increases where one athlete, Sam, had a decrease of -0.56m, making them an outlier. It then explains how IQR divides the data distribution into quartiles to identify outliers, with the lower bound set at Q1-1.5*IQR and upper bound at Q3+1.5*IQR. Outliers are values outside this range. Python code is provided to demonstrate outlier treatment using IQR.