Finding Moments from MGFs
What you'll learn: How to extract mean, variance, and higher moments from a moment generating function by taking derivatives.
The Core Idea
The moment generating function M(t) = E[e^(tX)] is called a "moment generating function" for good reason: its derivatives at t = 0 directly give you the moments of your random variable X.
Here's the magic formula:
The nth moment E[X^n] equals the nth derivative of M(t) evaluated at t = 0:
E[X^n] = M^(n)(0)
This works because when you differentiate M(t) = E[e^(tX)] with respect to t and then set t = 0, the exponential simplifies beautifully.
How It Works
First moment (mean):
- Differentiate once: M'(t) = E[X·e^(tX)]
- Evaluate at t = 0: M'(0) = E[X·e^0] = E[X] = μ
Second moment:
- Differentiate twice: M''(t) = E[X²·e^(tX)]
- Evaluate at t = 0: M''(0) = E[X²]
Variance: Once you have the first two moments, use the familiar formula: Var(X) = E[X²] - [E[X]]² = M''(0) - [M'(0)]²
Example Analogy
Think of M(t) as a "moment storage vault." Each time you differentiate and evaluate at t = 0, you're turning a combination lock that releases the next moment. The first turn gives you the mean, the second turn gives you E[X²], and so on.
Why t = 0?
Setting t = 0 eliminates the exponential weighting (e^0 = 1), leaving just the pure moment E[X^n].
Key Takeaway: To find the nth moment of a random variable, take the nth derivative of its MGF and evaluate at t = 0; variance comes from Var(X) = M''(0) - [M'(0)]².