Regression Analysis Used in Compensation Administration

Example of Coefficient of Determination

Consider the following 2 cases:

  • Executive compensation vs. company revenue
  • Factory worker compensation vs. company revenue

The correlation coefficient (r) of:

  • executive compensation vs. revenue is 0.95
  • factory worker compensation vs. company revenue is 0.2

To find R2, square the correlation coefficients.

(0.95)2 = 0.9025
(0.2)2 = 0.04

This results in an R2 of 0.9025 for executive compensation vs. company revenue, compared to an R2 of 0.04 for factory worker compensation vs. company revenue.

Executive compensation vs. company revenue
Factory worker compensation
vs. company revenue

Interpretation: 90% of the variation in executive compensation can be attributed to differences in company revenues, while only 4% of the variation in factory worker compensation can be attributed to differences in company revenues. This means company revenue is NOT a good predictor of factory workers' compensation.

Memory Jogger

An executive compensation vs. tenure (years with the company) correlation graph shows a correlation coefficient of 0.32. What may you conclude?

Continue