The Logic Behind Confidence Intervals
You'll learn why confidence intervals work by understanding how repeated sampling and the Central Limit Theorem guarantee our estimates capture the truth most of the time.
The Core Logic
A confidence interval doesn't mean "the true parameter has a 95% chance of being in this range." Instead, it means: if we repeated our sampling procedure many times, about 95% of the intervals we construct would contain the true parameter.
Here's the key insight: When you take a sample mean, it's a random variable that varies from sample to sample (sampling variability). But thanks to the Central Limit Theorem, we know the sample mean follows approximately a normal distribution centered at the true population mean μ, with standard error σ/√n.
How It Works
Think of it like casting a net: You don't move the fish (the true parameter) into your net—you throw a net that's wide enough to capture the fish most of the time.
- Before we sample, the sample mean is random and will fall within a predictable range around μ about 95% of the time (say, μ ± 1.96·SE)
- After we sample, we flip the perspective: our observed mean is fixed, so we build an interval around it using that same margin (X̄ ± 1.96·SE)
- The guarantee: Because 95% of sample means fall within 1.96 standard errors of μ, 95% of intervals built this way will trap μ inside them
The sampling variability is what requires an interval rather than just a point. The CLT is what makes the interval calculable with known probabilities.
Key Takeaway: Confidence intervals leverage the predictable sampling distribution of estimators (via the CLT) to create ranges that will capture the true parameter in a specified percentage of repeated samples—that's the "confidence" in confidence intervals.