Mean deviation for discrete frequency distribution
Let the given data consist of discrete observations x1, x2, x3,……., xn occurring with frequencies f1, f2, f3,……., fn respectively in case
 

Mean deviation about their Median M is given by
 

Mean deviation for continuous frequency distribution

where xi are the mid-points of the classes, x¯ and M are respectively, the mean and median of the distribution.

Variance
Variance is the arithmetic mean of the square of the deviation about mean x¯.
Let x1, x2, ……xn be n observations with x¯ as the mean, then the variance denoted by σ2, is given by
 

Standard deviation
If σ2 is the variance, then σ is called the standard deviation is given by
 

Standard deviation of a discrete frequency distribution is given by
 

Standard deviation of a continuous frequency distribution is given by
Coefficient of Variation

In order to compare two or more frequency distributions, we compare their coefficient of variations. The coefficient of variation is defined as
 

Note: The distribution having a greater coefficient of variation has more variability around the central value, then the distribution having a smaller value of the coefficient 0f variation.