Track 15 · Applied Statistics for Professionals · Lesson 6 of 12
Pay percentiles and pay-equity foundations
Build percentile pay bands, compute raw and adjusted pay gaps, and explain what an adjusted gap can and cannot tell you.
A reward team is asked two questions in the same week: “What should the salary band for a Senior Specialist be?” and “Do we have a gender pay gap?” Both are answered with percentiles and group comparisons — and the second can be answered misleadingly in two opposite directions if the analyst does not think about what to adjust for.
Pay data are right-skewed: many people at modest salaries and a long tail of senior pay. Means are pulled up by the tail, so reward teams describe pay with percentiles. The -th percentile is the value below which about of salaries lie. The 10th–90th percentile range (P10–P90) is a common pay band: it covers the middle 80% of incumbents and ignores a few unusual cases at each end.
Included with a Formula Quest membership
It has 8 graded exercises with feedback, hints and worked solutions. After it you will be able to:
- Compute type-7 percentiles and build P10–P90 pay bands, compa-ratios and band positions
- Calculate raw mean and median pay gaps as a share of men's pay
- Explain how differences in representation across levels create a raw gap
- Convert a log-salary regression coefficient into an adjusted percentage gap
- Explain why an adjusted gap can hide inequity when the controls are themselves affected by it