Confidence Intervals for the Difference Between Two Means
What you'll learn: How to estimate and interpret the range of plausible values for the true difference between two independent population means.
The Core Idea
Sometimes you don't just want to compare one group to a fixed value—you want to compare two groups to each other. For example: Do customers in Region A spend more than those in Region B? Is a new drug more effective than the standard treatment?
A confidence interval for the difference between two means (μ₁ - μ₂) gives you a range that likely contains the true difference. If the entire interval is positive, you have evidence that group 1 has a higher mean. If it includes zero, you can't rule out that the groups are equal.
How It Works
You collect two independent samples (one from each population), calculate their means and standard errors, then combine them:
Formula structure:
(x̄₁ - x̄₂) ± t* × SE(x̄₁ - x̄₂)
Where:
x̄₁ - x̄₂is the observed difference in sample meanst*is the critical value from the t-distribution (based on your confidence level and degrees of freedom)SE(x̄₁ - x̄₂)is the standard error of the difference
The standard error combines variability from both samples. Larger sample sizes or smaller variances make the interval narrower (more precise).
Interpretation
A 95% CI of [2.1, 5.8] for (μ₁ - μ₂) means: "We're 95% confident the true mean difference is between 2.1 and 5.8 units." Since zero isn't in this range, group 1 likely has a higher mean than group 2.
Key Takeaway: Confidence intervals for the difference between two means let you quantify how much two groups differ, accounting for sampling uncertainty—essential for A/B tests, medical trials, and comparative studies.