1.17 Identifying Outliers

In everyday language, the word “outlier” is often used imprecisely, simply to describe an unusual outcome or event.  Even in the field of statistics, there is no single, universal definition for the term.  Some statisticians might refer to a variable’s maximum and minimum values as outliers, others might refer to any value more than three standard deviations above or below the mean as an outlier, and others still might use it in a more subjective fashion, just labeling anything atypical in this way.

Some statistical software will identify points as outliers when they are greater than the 75th percentile plus 1.5x the IQR, or less than the 25th percentile minus 1.5x the IQR.   In Chapter 2 (Data Visualization) we will look at boxplots, and we will see outliers identified in this way.