In statistics, it refers to the standard deviation of a distribution divided by the distribution's mean, providing a standardized measure of the variation in a distribution, which does not increase simply because the mean itself increases or because the units of measurement change.