Confidence Interval for the Difference of Two Proportions
What You'll Learn
You'll learn how to estimate the difference between two population proportions and interpret the resulting confidence interval to compare two groups.
The Core Idea
Imagine you want to compare the success rate of two different treatments, or the proportion of voters favoring a policy in two different states. Instead of just comparing sample proportions, you can construct a confidence interval for the difference between the two population proportions.
If you have two independent samples:
- Sample 1: size n₁, with x₁ successes, giving proportion p̂₁ = x₁/n₁
- Sample 2: size n₂, with x₂ successes, giving proportion p̂₂ = x₂/n₂
The point estimate for the difference is simply: p̂₁ - p̂₂
But we want an interval estimate. Using the Central Limit Theorem (which you've studied), when sample sizes are large enough, the sampling distribution of p̂₁ - p̂₂ is approximately normal.
The Confidence Interval Formula
The confidence interval for (p₁ - p₂) is:
(p̂₁ - p̂₂) ± z · √[(p̂₁(1-p̂₁)/n₁) + (p̂₂(1-p̂₂)/n₂)]*
where z* is the critical value from the standard normal distribution for your desired confidence level (e.g., 1.96 for 95% confidence).
Interpretation
If your interval is entirely above zero, you can be confident p₁ > p₂. If it's entirely below zero, then p₁ < p₂. If the interval includes zero, you cannot confidently say one proportion differs from the other.
Key Takeaway: A confidence interval for the difference of two proportions quantifies the plausible range for how much two population proportions differ, allowing you to assess whether that difference is meaningful or could plausibly be zero.