
The Bayesian interval containing a stated fraction of the posterior — the answer to 'where is the parameter', unlike a confidence interval.A credible interval is the Bayesian answer to the question 'where is the parameter?' It is a range of values that contains a specified fraction of the posterior probability distribution. For example, a 95% credible interval contains 95% of the posterior mass, meaning that, given the data and the model, there is a 95% probability the parameter lies within that interval. This direct probabilistic interpretation is a key difference from the frequentist confidence interval, which does not make a probability statement about a fixed parameter.
The concept was introduced by Thomas Bayes in the 18th century, though the formal term 'credible interval' came into common use in the 20th century with the development of Bayesian statistics.
To construct a credible interval, you first compute the posterior distribution, which combines prior beliefs with the likelihood of the observed data via Bayes' theorem. Then you find the interval that contains the desired probability mass. The most common method is the 'highest posterior density' interval, which is the narrowest interval containing the specified mass, but equal-tailed intervals (cutting off equal tails) are also used. The choice of interval type depends on the shape of the posterior and the question at hand.
Credible intervals are widely used in medical trials, economics, and machine learning to report uncertainty in parameter estimates, such as the effect size of a drug or the prediction error of a model.
A 95% credible interval can be interpreted as 'there is a 95% chance the parameter lies in this interval' — a statement that is mathematically forbidden for confidence intervals.

