Method of Moments for Multiple Parameters
What you'll learn: How to estimate several unknown parameters at once by matching multiple sample moments to their theoretical counterparts.
The Core Idea
When a distribution has multiple parameters—like the normal distribution with both mean μ and variance σ²—you need multiple equations to solve for all unknowns. The method of moments extends naturally: set the first k sample moments equal to the first k theoretical moments, then solve the resulting system of equations simultaneously.
Think of it like having two locks that need two keys: one moment equation isn't enough to pin down both parameters, but two equations give you exactly what you need.
How It Works
For a distribution with parameters θ₁, θ₂, ..., θₖ:
- Write theoretical moments as functions of the parameters: E[X], E[X²], E[X³], etc.
- Compute sample moments from your data: m₁ = x̄, m₂ = (1/n)Σxᵢ², etc.
- Set them equal: E[X] = m₁, E[X²] = m₂, and so on for k equations
- Solve the system for θ₁, θ₂, ..., θₖ
Example: Normal Distribution
For N(μ, σ²), we have two parameters:
- First moment: E[X] = μ
- Second moment: E[X²] = σ² + μ²
Setting equal to sample moments:
- μ̂ = x̄ (sample mean)
- σ̂² + (x̄)² = (1/n)Σxᵢ²
Solving gives σ̂² = (1/n)Σxᵢ² - (x̄)² = (1/n)Σ(xᵢ - x̄)²
Both parameters estimated simultaneously from the same data!
Key Takeaway: The method of moments for multiple parameters requires matching as many sample and theoretical moments as you have unknown parameters, then solving the resulting system of equations together.