The geometric mean is the nth root of the product of n values. It’s particularly useful for data representing rates, growth rates, and proportional changes.
When to Use Geometric Mean
- Growth rates: Average annual growth rates
- Ratios: Investment returns over multiple periods
- Proportional data: Data where values are multiplied, not added
- Percentage changes: Average rate of increase/decrease
Advantage over arithmetic mean: Less affected by extreme values in certain contexts.
Geometric Mean for Ungrouped Data
Formula:
GM = ⁿ√(x₁ × x₂ × ... × xₙ)
GM = (x₁ × x₂ × ... × xₙ)^(1/n)
Or using logarithms (practical):
GM = 10^(Σ log(xᵢ) / n) or e^(Σ ln(xᵢ) / n)
Example: Investment returns over 3 years: 10%, 15%, 8%
GM = ³√(1.10 × 1.15 × 1.08) GM = ³√(1.3662) GM ≈ 1.1096 or 10.96% average annual return
Geometric Mean for Grouped Data
Formula:
GM = ⁿ√(x₁^f₁ × x₂^f₂ × ... × xₖ^fₖ)
where:
xᵢ = class midpoint
fᵢ = frequency
n = Σfᵢ (total frequency)
Using logarithms (practical):
GM = e^(Σ(fᵢ × ln(xᵢ)) / Σfᵢ)
Key Characteristics
- Always less than or equal to arithmetic mean (AM)
- Useful for normalized/indexed data
- Cannot be used with zero or negative values (for logarithm method)
- Represents the “middle” rate of change
Related: Relationship Between Means
For positive values:
Harmonic Mean ≤ Geometric Mean ≤ Arithmetic Mean
HM ≤ GM ≤ AM