The degree to which test scores for a group of test takers are consistent for repeated evaluations of a measurement procedure. A measure is said to have a high reliability if it produces consistent results under consistent conditions, and for multiple evaluators.