Convergence in Distribution

Graduate Depth 85 in the knowledge graph I know this Set as goal
Unlocks 54 downstream topics
convergence distribution limit-theorems

Core Idea

Xₙ converges to X in distribution if lim_{n→∞} Fₙ(x) = F(x) at continuity points of F, or equivalently lim_{n→∞} φₙ(t) = φ(t) for all t. This is the weakest form of convergence—Xₙ and X need not be defined on the same probability space. Characteristic function convergence provides the most convenient criterion.

Explainer

From your study of distribution functions and densities, you know that a distribution function F(x) = P(X ≤ x) completely characterizes a random variable's probabilistic behavior. Convergence in distribution (also written Xₙ →_d X or Xₙ ⟹ X) says that the sequence of distributions, as described by their CDFs Fₙ, approaches the distribution F — not that the random variables themselves get close. This distinction is crucial and is what makes convergence in distribution the weakest of the three main modes.

To understand why "weakest" is meaningful, consider what stronger convergence demands. Almost sure convergence says that the actual sample paths Xₙ(ω) converge to X(ω) for nearly every outcome ω — the random variables must be defined on the same probability space and their values must track each other. Convergence in probability says that P(|Xₙ − X| > ε) → 0, again requiring both to live on the same space and their values to be close with high probability. Convergence in distribution only requires that probabilities of events like {X ≤ x} converge — it says nothing about whether Xₙ and X are close as numbers in any specific sample. In fact, Xₙ and X don't even need to be defined on the same probability space, since only their marginal distributions matter.

The CDF definition has a subtle caveat: convergence is required only at continuity points of F. This is necessary because CDFs can have jump discontinuities, and at a jump, a sequence of CDFs could converge to either boundary value. At continuity points, this ambiguity disappears. The characteristic function criterion — that convergence in distribution is equivalent to pointwise convergence of characteristic functions φₙ(t) = E[e^{itXₙ}] — is often more tractable in proofs. Characteristic functions are always continuous and bounded, so no caveat about continuity points is needed, and Fourier-analytic tools become available.

The Central Limit Theorem, which you'll encounter next, is the canonical example of convergence in distribution: for i.i.d. random variables with finite mean and variance, the standardized sums √n(X̄ₙ − μ)/σ converge in distribution to a standard normal N(0,1). Notice that this doesn't say the sample means literally converge to a normal — they converge in probability to μ by the law of large numbers. Rather, the *shape of their fluctuations*, properly scaled, approaches the normal distribution. Convergence in distribution is precisely the right tool for describing this kind of limiting shape, which is why it sits at the heart of classical probability theory.

Practice Questions 5 questions

Prerequisite Chain

Counting to 10Counting to 20Understanding ZeroThe Number ZeroCounting to FiveOne-to-One CorrespondenceCombining Small Groups Within 5Addition Within 10Addition Within 20Two-Digit Addition Without RegroupingTwo-Digit Addition with RegroupingAddition Within 100Repeated Addition as MultiplicationMultiplication Facts Within 100Division as Equal SharingDivision as Grouping (Measurement Division)Division: Grouping (Repeated Subtraction) ModelDivision: Fair Sharing ModelDivision as Equal SharingDivision as GroupingBasic Division FactsDivision Facts Within 100Two-Digit by One-Digit DivisionDivision with RemaindersRemainders and Quotients in DivisionDivision Word ProblemsIntroduction to Long DivisionFactors and MultiplesPrime and Composite NumbersEquivalent FractionsRelating Fractions and DecimalsDecimal Place ValueReading and Writing DecimalsComparing and Ordering DecimalsAdding and Subtracting DecimalsMultiplying DecimalsDividing DecimalsDividing FractionsMixed Number ArithmeticOrder of OperationsInteger Order of OperationsVariable ExpressionsCombining Like TermsOne-Step EquationsTwo-Step EquationsSolving Multi-Step EquationsEquations with Variables on Both SidesAngle Pairs: Complementary, Supplementary, and VerticalParallel Lines and TransversalsCorresponding AnglesAlternate Interior AnglesTriangle Angle Sum TheoremExterior Angle TheoremTriangle Inequality TheoremSimilar Triangles: AA SimilaritySimilar Triangles: SSS and SAS SimilarityProportions in Similar TrianglesRight Triangle Trigonometry IntroductionTrigonometric Ratios ReviewRadian MeasureConverting Between Degrees and RadiansThe Unit CircleGraphing Sine and CosineGraphing Tangent and Reciprocal Trigonometric FunctionsDerivatives of Trigonometric FunctionsAntiderivativesIndefinite IntegralsBasic Integration RulesRiemann SumsDefinite Integral DefinitionFundamental Theorem of Calculus Part 1Fundamental Theorem of Calculus Part 2U-SubstitutionPartial Fraction Decomposition for IntegrationImproper Integrals - ConvergenceIntegral TestP-SeriesComparison TestLimit Comparison TestAbsolute vs. Conditional ConvergencePower SeriesTaylor PolynomialsTaylor SeriesMoment Generating FunctionsCharacteristic FunctionsConvergence in Distribution

Longest path: 86 steps · 458 total prerequisite topics

Prerequisites (2)

Leads To (7)