The 1.5(IQR) Statistical Outlier Rule is an easy way to find High Outlier and Low Outlier values in your own datasets based on the following two formulas that can be applied on any column of data:
High Outlier > 75th percentile + 1.5(IQR)
Low Outlier < 25th percentile - 1.5(IQR)
These two formulas apply the statistical theory of the box & whisker plot that calculates the following metrics on a column of data: IQR (Interquartile Range), 75th Percentile, 50th Percentile (Median), 25th Percentile, as well as the UPPER and LOWER Limits (the whiskers of a box & whisker plot).
These formulas use the 50th Percentile to find High Outlier and Low Outlier values. High Outlier values are greater then the UPPER Limit of a box plot, Low Outlier values are less than the box & whisker plot LOWER Limit.
High Outlier and Low Outlier values negatively impact data analysis of all kinds: forecasting and other modeling, scenario analysis, data visualization, and Key Performance Indicator (KPI) reporting.
The 1.5(IQR) Statistical Outlier Rule can be coded using any of the many languages (VBA, R, SAS, Python, SPSS, Tableau, Minitab, SQL, etc.) that data scientists in business and academics currently use to analyze data.
