Prediction Intervals

Statistics · Axiom Academy

Understanding confidence intervals for mean responses vs. prediction intervals for individual responses 1. Confidence Interval for Mean Response A confidence interval for the mean response estimates the average value of y for all observations at a specific x-value. This interval captures our uncertainty about where the true regression line lies. Formula for Confidence Interval: t α/2 = critical t-value with n-2 degrees of freedom s = standard error of the regression SE = standard error for the mean response 2. Prediction Interval for Individual Response A prediction interval for an individual response estimates the range where a single new observation will fall at a specific x-value. This interval must account for more uncertainty than the confidence interval. Formula for Prediction Interval: Key difference: Notice the "+ 1" term under the square root. This additional term accounts for the variability of individual observations around the mean. 3. Why Prediction Intervals are Wider Prediction intervals are always wider than confidence intervals at the same x-value. This makes intuitive sense when we consider the sources of uncertainty. Confidence Interval (CI) for Mean Response: Only accounts for uncertainty in estimating the regression line Gets narrower as sample size increases Narrowest at x̄ (the mean of x-values) Prediction Interval (PI) for Individual Response: Accounts for uncertainty in estimating the regression line

This is the written version of the interactive lesson above. See the full Statistics course.