Questions: Prediction Intervals in Regression

5 questions to test your understanding

Score: 0 / 5
Question 1 Multiple Choice

A researcher fits a regression model with n = 10,000 observations and reports a very tight 95% confidence interval for the mean response at x = 5. A colleague says this means they can predict any individual patient's outcome with high precision. What is wrong with this claim?

ANothing — a tight confidence interval implies a tight prediction interval for the same data
BThe confidence interval estimates the population mean response, not individual outcomes; a prediction interval would be much wider due to irreducible person-to-person variation
CThe colleague should use a 99% confidence level instead of 95% for medical applications
DThe model must be misspecified if individual predictions are not as precise as the confidence interval
Question 2 Multiple Choice

As sample size n approaches infinity, what happens to a 95% prediction interval for a new observation?

AIt collapses to zero width, as all intervals do with sufficient data
BIt narrows to zero only if the true error variance σ² equals zero
CIt approaches a fixed non-zero width determined by the irreducible observation variance σ²
DIt becomes equivalent to the confidence interval for the mean response
Question 3 True / False

A confidence interval for the mean response and a prediction interval for a new observation answer the same underlying statistical question.

TTrue
FFalse
Question 4 True / False

Prediction intervals are always wider than confidence intervals at the same x value and confidence level.

TTrue
FFalse
Question 5 Short Answer

Explain why a prediction interval cannot shrink to zero width even with an arbitrarily large sample, while a confidence interval for the mean response can.

Think about your answer, then reveal below.