Statistics Glossary → Coefficient of variation (CV)
Coefficient of variation (CV)concept
The ratio of the standard deviation to the mean; a unit-free measure that expresses spread as a percentage of the mean.
It is computed as CV = s / x̄ (often multiplied by 100%). Because it has no unit, it compares the relative spread of variables measured on different scales or in different units: for example, the relative variability in the weights of elephants and mice, or the repeatability of a laboratory assay.
It is meaningful only for ratio-scale variables that have a true zero and take positive values; it misleads on interval scales such as temperature in Celsius or for data whose mean is close to zero, because the CV grows without bound as the mean shrinks. In small samples the sample CV slightly underestimates the true value.
An everyday example
A shop sells on average 200 loaves of bread a day with a standard deviation of 20 (CV = 10%), and 10 cakes a day with a standard deviation of 4 (CV = 40%). Although the standard deviation is larger for bread, cake sales are relatively far more volatile.
In MerQur
In Descriptive Statistics, the 'CV, geometric/harmonic mean, sum' option reports the coefficient of variation; it is also among the statistics for which the Bootstrap CI analysis can build an interval.
Analyses where this concept appears
Descriptive StatisticsBootstrap CI