Confidence Interval

What a confidence interval actually is

A confidence interval is a range of plausible values around a study’s result. When researchers measure something, like how much a treatment lowers anxiety scores, they get a single number from their sample. But that number is just an estimate. The confidence interval shows the range where the true value probably falls.

You’ll usually see it written like “a 7-point drop, 95% CI 4 to 10.” That means the best estimate is 7, and the data are consistent with the real effect being anywhere from about 4 to 10.

How it works

The width of the interval tells you how precise the estimate is. A narrow interval means the study pinned the result down tightly. A wide one means there’s a lot of uncertainty, often because the study was small or the data were noisy.

The “95%” refers to the method’s reliability over many studies, not the odds for this one result. If researchers repeated the study many times, about 95% of the intervals they calculated would capture the true value. In practice, readers treat it as a sensible range for where the truth likely sits.

What it isn’t

A confidence interval isn’t a guarantee that the true value is inside it. It’s a range produced by a method that’s right most of the time, not a certainty about this single estimate. It also isn’t the same as the spread of individual people in the study. It describes uncertainty about the estimate, not how much people differed from each other.

And it isn’t only about being statistically significant. Even when a result clears that bar, a very wide interval can mean the real effect could be tiny or large.

Related terms you’ll see next

• Statistical significance
• P-value
• Effect size
• Randomized controlled trial

Why it matters when you read about mental health

Confidence intervals tell you how much to trust a headline result. A study reporting that a new therapy “reduced symptoms” looks very different if the interval is narrow and clearly helpful versus wide and barely budging off zero. Checking the interval, not just whether something “worked,” helps you judge how solid the evidence really is.

<blockquote>
  <p><strong>Confidence Interval:</strong> A confidence interval is a range of values that likely contains the true result. It shows how precise a study's estimate is.</p>
  <p>&mdash; <a href="https://shrinktionary.com/terms/confidence-interval/">Shrinktionary</a>, a plain-English mental health dictionary, medically reviewed by a board certified psychiatrist.</p>
</blockquote>