Course contentsShow
← Data Science
Lesson 407 of 2,14511. Hypothesis Tests for Proportions and CategoriesFree lesson

Pooled Proportion Estimator

Calculating the combined proportion under the null hypothesis assumption.

Pooled Proportion Estimator

You'll learn how to combine data from two groups into a single proportion estimate when testing whether their true proportions are equal.

What Is It?

When comparing two proportions (like success rates in two groups), the null hypothesis typically states they're equal: H₀: p₁ = p₂. Under this assumption, both groups share the same underlying proportion. The pooled proportion estimator combines all successes and all observations from both samples into one overall proportion estimate.

Think of it like this: if you're testing whether two coins are equally fair, and the null hypothesis says "both have the same probability of heads," then your best guess for that shared probability is to pool all the flips together—count all heads from both coins divided by total flips from both coins.

The Formula

The pooled proportion is calculated as:

p̂_pooled = (x₁ + x₂) / (n₁ + n₂)

Where:

  • x₁ = number of successes in sample 1
  • x₂ = number of successes in sample 2
  • n₁ = size of sample 1
  • n₂ = size of sample 2

Why Pool?

Under the null hypothesis, we assume no difference exists between groups. Pooling gives us a more stable, combined estimate with lower variance than using either group's proportion alone. This pooled value becomes the reference point for calculating the test statistic's standard error.

Example

If Group A has 45 successes out of 100 trials and Group B has 38 successes out of 120 trials:

p̂_pooled = (45 + 38) / (100 + 120) = 83/220 ≈ 0.377

This 37.7% becomes our best estimate of the common proportion under H₀.

Key Takeaway: The pooled proportion estimator combines successes and sample sizes from both groups to estimate the shared proportion assumed under the null hypothesis of equal proportions.