What is the Law of Large Numbers, and how is it different from the Central Limit Theorem?
StatisticsLevel: FoundationStatistics and probability
What is the Law of Large Numbers, and how is it different from the Central Limit Theorem?
Hints
- What is random in the experiment and what is being averaged?
- Check independence, finite-moment assumptions and suitability for dependent data.
Answer
The law of large numbers concerns convergence: for independent identically distributed observations with a finite absolute first moment, the sample mean approaches the population mean. It does not by itself specify a normal finite-sample distribution.
The usual iid central limit theorem additionally assumes finite, nonzero variance σ². It says √n(average − μ)/σ converges in distribution to a standard normal variable. Thus the mean has an approximate standard error σ/√n under those conditions.
The LLN answers where the average converges; the CLT describes its normalized fluctuations. Neither theorem supplies a universal rule that n = 30 is enough. Heavy tails, dependence and nonstationarity require checking assumptions or using other limit results.
Source and licence: Ishan Dutta — Awesome Data Science Interview Q&A · MIT · answer: Quantoros editorial