Convergence in Probability
What you'll learn: You'll understand what it means for a sequence of random variables to converge in probability—a formal way to say that as we take more observations, a random quantity becomes "almost certainly" close to a fixed value.
The Core Idea
Imagine you're estimating the average height of all adults by measuring random samples. As your sample size grows, your sample average should get closer and closer to the true population average. But since each sample is random, there's always some chance of getting an unusual result.
Convergence in probability formalizes this intuition: a sequence of random variables X₁, X₂, X₃, ... converges in probability to a constant (or another random variable) if the probability that Xₙ is "far away" from that limit becomes vanishingly small as n increases.
Formal Definition
We say that a sequence {Xₙ} converges in probability to a value c (written Xₙ →ᵖ c) if:
For any tiny distance ε > 0 (no matter how small),
P(|Xₙ - c| > ε) → 0 as n → ∞
In plain English: the chance that Xₙ differs from c by more than ε shrinks to zero as n grows.
Why It Matters
This concept is the mathematical foundation for the Weak Law of Large Numbers you just saw. It tells us that sample averages don't just "tend toward" the mean—they converge in probability, meaning extreme deviations become infinitely unlikely with enough data.
Unlike almost-sure convergence (a stronger notion), convergence in probability allows for rare "blips" at arbitrarily large n—what matters is that these become vanishingly rare.
Key Takeaway: Convergence in probability means the probability of being far from the limit shrinks to zero as n increases—formalizing the idea that randomness "washes out" with enough observations.