What Does Confidence Mean?

Statistics · Axiom Academy

INTRO What Does Confidence Mean? Discover the surprising truth about confidence intervals and why they're not about probability. Let's start with a scenario. You calculate a 95% confidence interval for the average height of all adults in a city. Click the statement you think is correct: Let's simulate what happens when we create many confidence intervals. The true population mean is marked with a red line. Understanding confidence requires distinguishing between what's random and what's fixed. Why can't we say there's 95% probability after calculating? ❌ Why The Misconceptions Persist Let's address why the wrong interpretations are so tempting. Why it's tempting: This is how we naturally think about uncertainty in everyday life. Why it's wrong: In frequentist statistics, the parameter doesn't have a probability distribution—it's a fixed unknown. "Certainty" about a fixed value isn't probability. (Note: Bayesian statistics does allow this interpretation, but uses different methods.) Why it's tempting: The word "confidence" sounds like it's about the probability of being right. Why it's wrong: Your specific interval [a, b] is fixed—it's just one outcome. The true value either is or isn't inside. There's no probability about it. The 95% describes the procedure's success rate, not any single interval. Why it's tempting: "Confidence" seems like a measure of our belief.

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