A range likely to contain the true effect if the study were repeated many times.
The 95% CI gives the range of values within which the true population parameter (e.g., a treatment effect or mean difference) would be expected to fall in 95% of repeated samples drawn the same way. A narrow CI suggests a more precise estimate, while a wide CI reflects greater uncertainty, often due to small sample size. It is commonly misread as 'a 95% probability the true value lies in this range' — technically incorrect under the frequentist framework, though widely used this way in practice.
This guide was auto-drafted and is pending editorial review.