Simpson's Paradox: A Data Set and Discrimination Case Study Exercise

Abstract

In this article, we present a data set and case study exercise that can be used by educators to teach a range of statistical concepts including Simpson's paradox. The data set and case study are based on a real-life scenario where there was a claim of discrimination based on ethnicity. The exercise highlights the importance of performing rigorous statistical analysis and how data interpretations can accurately inform or misguide decision makers.

• Univariate analysis
• Bivariate analysis
• Specific variation
• Outliers
• Weighted average
• Simpson's paradox