noun Statistics. a process by which a method that works for one sample of a population is checked for validity by applying the method to another sample from the same population.