Course contentsShow
← Data Science
Lesson 184 of 2,1455. Continuous Probability DistributionsFree lesson

Beta Distribution: Bounded Between 0 and 1

A distribution for variables constrained to the interval [0,1], controlled by two shape parameters α and β.

Beta Distribution: Bounded Between 0 and 1

What you'll learn: How the Beta distribution models variables that can only take values between 0 and 1, controlled by two shape parameters.

The Core Idea

The Beta distribution is a continuous probability distribution specifically designed for variables confined to the interval [0, 1]. Unlike the Normal or Exponential distributions that can theoretically extend infinitely, the Beta distribution is bounded—perfect for modeling proportions, probabilities, and percentages.

Think of it like a flexible rubber band stretched between 0 and 1. Depending on how you adjust it, the "weight" of the probability can shift left, right, bunch up in the middle, or spread evenly across the interval.

The Two Shape Parameters: α and β

The Beta distribution is controlled by two parameters:

  • α (alpha): Controls the shape on the left side
  • β (beta): Controls the shape on the right side

These parameters determine where the probability density concentrates:

  • When α > β: The distribution leans toward 1 (right side)
  • When β > α: The distribution leans toward 0 (left side)
  • When α = β: The distribution is symmetric around 0.5
  • When α = β = 1: You get a Uniform distribution on [0, 1]

Real-World Examples

  • Click-through rates in digital marketing (proportion of clicks)
  • Baseball batting averages (hits per at-bat)
  • Survey response rates (proportion who respond)
  • Quality control pass rates (proportion of acceptable products)

Key Takeaway: The Beta distribution is the go-to model for any variable naturally constrained between 0 and 1, with α and β shaping how probability mass is distributed across that interval.