Course contentsShow
← Mathematics
Lesson 2920 of 3,12157. Common DistributionsFree lesson

Properties and Applications of the Beta Distribution

Derive mean and variance of the beta distribution and explore its use in modeling proportions and Bayesian priors.

Properties and Applications of the Beta Distribution

What you'll learn: How to calculate the mean and variance of the beta distribution and apply it to model proportions and probabilities.

Core Idea

The beta distribution is a continuous distribution defined on the interval [0, 1], making it perfect for modeling proportions, probabilities, and percentages. With two shape parameters α (alpha) and β (beta), it's incredibly flexible—it can be symmetric, skewed left, skewed right, U-shaped, or uniform depending on these parameters.

Mean and Variance

For a beta-distributed random variable X ~ Beta(α, β):

Mean (Expected Value): E[X]=αα+βE[X] = \frac{\alpha}{\alpha + \beta}

This is intuitive: the mean represents the "weight" of α relative to the total α + β.

Variance: Var(X)=αβ(α+β)2(α+β+1)\text{Var}(X) = \frac{\alpha\beta}{(\alpha + \beta)^2(\alpha + \beta + 1)}

Notice that as α and β grow larger (keeping their ratio fixed), the variance decreases—the distribution becomes more concentrated around its mean.

Applications

Modeling Proportions

The beta distribution naturally models any quantity constrained between 0 and 1: completion rates, success probabilities, market shares, or the proportion of defective items in manufacturing.

Bayesian Priors

In Bayesian statistics, the beta distribution serves as the conjugate prior for binomial and Bernoulli likelihoods. If you're estimating an unknown probability p, a Beta(α, β) prior combined with binomial data yields another beta posterior. The parameters α and β can be interpreted as "prior successes" and "prior failures."

For example, Beta(1, 1) represents complete ignorance (uniform distribution), while Beta(10, 10) represents moderate confidence that p ≈ 0.5.

Key Takeaway: The beta distribution's mean α/(α+β) and variance formula make it ideal for modeling proportions and serving as flexible Bayesian priors for unknown probabilities.