Confidence Interval

1Detailed Explanation

Confidence intervals provide information about the precision of an estimate and the uncertainty surrounding it. A 95% CI means that if the study were repeated many times, 95% of the computed intervals would contain the true parameter. Important misconceptions: the CI does not mean there's a 95% probability the true value is in the interval, nor does it mean 95% of data fall within the interval. CI width reflects sample size and variability: larger samples and less variability produce narrower CIs. CI should overlap with clinical significance thresholds, not just statistical significance. Reporting both CI and p-values provides complementary information.

2Examples

• A.Risk ratio = 1.5, 95% CI: 1.2 to 1.9 — the true effect likely ranges from a 20% increase to a 90% increase
• B.Mean difference = 5 mmHg, 95% CI: 2 to 8 mmHg — excludes zero, indicating statistical significance

3Why It Matters in Research

Confidence intervals are preferred over p-values alone because they convey both significance and clinical relevance. Major journals increasingly require CI reporting alongside p-values.

4Related Terms