gen-define-regression-to-the-mean-in-performance-and-give-one-sh-061a

Define regression to the mean in performance and give one short worked example.

# Regression to the Mean in Performance **Definition:** Regression to the mean is the statistical phenomenon where extreme performances tend to be followed by more moderate performances closer to the average. This occurs because any single performance includes both true skill/ability and random variation (luck, conditions, measurement error). Extreme results typically involve unusually favorable random factors that are unlikely to repeat. ## Worked Example **Scenario:** A sales manager notices that the top 10 performers in Q1 averaged 50 sales each, well above the company average of 35 sales. **Q1 Results (Top Performers):** - True skill component: ~42 sales - Random positive factors (luck): +8 sales - **Observed: 50 sales** **Q2 Results (Same People):** - True skill component: ~42 sales (unchanged) - Random factors: ±0 sales (averages out) - **Observed: ~42 sales** **Outcome:** The Q2 average for these "top performers" drops to 42 sales—still above the company average (35), but notably lower than their Q1 performance (50). **Key Insight:** The manager might mistakenly think these employees became complacent or that a Q1 bonus program "stopped working," when actually this decline is statistically expected. The extreme Q1 results were partly due to luck, which didn't persist. This is why comparing extreme performers to themselves (before/after) can be misleading without a control group.