Variance of the Geometric Distribution
What you'll learn: You'll derive the formula for the variance of a geometric random variable and understand what it tells you about the spread of waiting times.
The Core Idea
Recall that a geometric random variable (X) counts the number of trials until the first success in repeated Bernoulli trials with success probability (p). You already know (E(X) = \frac{1}{p}). Now we want to quantify how much variability or spread exists around this mean.
The variance measures this spread, and for the geometric distribution:
Deriving the Formula
Start with the shortcut formula for variance:
We know (E(X) = \frac{1}{p}), so ([E(X)]^2 = \frac{1}{p^2}).
The tricky part is finding (E(X^2)). Using the definition with the geometric PMF (P(X=k) = (1-p)^{k-1}p):
Through calculus techniques (differentiating the generating function twice) or algebraic manipulation, this sum evaluates to:
Therefore:
Interpretation
- When (p) is small (rare success), both the mean and variance are large—you wait longer and with more unpredictability.
- When (p) approaches 1 (nearly certain success), variance shrinks toward zero—you almost always succeed on the first trial.
- The standard deviation is (\sigma = \sqrt{\frac{1-p}{p^2}} = \frac{\sqrt{1-p}}{p}).
Key Takeaway: The variance of a geometric random variable is (\text{Var}(X) = \frac{1-p}{p^2}), capturing how waiting times spread around the mean, with more variability when success is rare.