Quick answer: A 95% confidence interval is the range of values that the data support for a population parameter — if the study were repeated many times, 95% of such intervals would capture the true value. It answers the question a p-value cannot: not just “is there an effect?” but “how big could it plausibly be?” The mean difference 8 mmHg with 95% CI [3, 13] tells you the effect could plausibly be trivial (3) or clinically huge (13) — that honesty is why modern journals want CIs first. Companion reading: the p-value explainer.
Table of Contents
What the confidence level actually means
The 95% describes the procedure, not your one interval: across countless repeated experiments, intervals built this way capture the true value 95% of the time. Your single interval either contains the truth or it doesn’t — there’s no probability left about it. What you can read off it: the width is the precision of your study (driven by sample size and variance), and any value inside the range is compatible with your data. A CI that includes zero corresponds to p > .05; one that excludes it corresponds to p < .05 — they are two views of the same test at the same α.
Reading intervals like a reviewer
| Interval (for a mean difference vs 0) | Honest reading |
|---|---|
| [3.0, 13.0] | Effect clearly present; even the worst case is non-trivial |
| [0.1, 13.0] | Significant but uninformative — the true effect could be almost anything |
| [−2.0, 13.0] | Not significant; the data cannot rule out a small negative or a large positive |
| [−1.0, 1.0] | Precisely near-zero — strong evidence of no meaningful effect (precise null) |
The fourth row is the one most writers get wrong: “not significant” is not “no effect” unless the interval is also precise. A wide interval that straddles zero means the study was inconclusive; a narrow one that hugs zero is genuine evidence of absence. This distinction decides how non-significant results should be discussed — and the width responds directly to n, as covered in the power analysis guide.
Reporting and the software reality
Every mainstream tool prints CIs: jamovi renders them in the t-test and regression coefficient tables (one tick-box — see the t-test walkthrough), R’s t.test() and R Commander output them by default, and the APA format is mean difference = 8.0, 95% CI [3.0, 13.0]. The habit that upgrades any results section: report the estimate with its interval before the p-value, and interpret the interval’s endpoints in the units of the problem — that is the sentence that shows you understood your own data. Effect-size measures and their CIs are catalogued in the effect size guide.
For one-to-one coaching on estimation-driven reporting, Ampersand Academy teaches statistics one-to-one.
Official documentation
For the full reference behind this guide, read the NIST/SEMATECH e-Handbook page on confidence limits for the mean. It is the source the details above are drawn from, and the first place to check when an option or attribute name changes.
Frequently asked questions
What does a 95% confidence interval mean?
Across repeated experiments, 95 percent of intervals constructed this way would contain the true population value. It is a property of the method, quantifying how precisely your data pin down the parameter.
Is a 99% confidence interval better than 95%?
It is more conservative – wider and more certain to capture the truth – but less informative per width. Most fields standardize on 95; use 99 when false claims are especially costly.
Does a confidence interval crossing zero mean no effect?
It means the effect is not demonstrated at that alpha level. If the interval is wide, the study may simply be underpowered; only a narrow interval hugging zero is real evidence of no meaningful effect.
What makes a confidence interval narrower?
Larger samples, less outcome variance, and lower confidence levels. Paired designs narrow intervals too, because they remove between-subject variance from the comparison.
How do I report a confidence interval in APA style?
Give the estimate with the interval in brackets: mean difference = 8.0, 95% CI [3.0, 13.0]. Interpret the endpoints in the units of the problem rather than restating that zero lies outside.
Last updated on · Written by Dinesh Kumar R
