Statistical Minimum and Maximum

The statistical minimum or l, also known as the low outlier limit, is the minimum value in a set of values, excluding any outliers. The statistical maximum or h, also known as the high outlier limit, is the maximum value in a set of values, excluding any outliers. Both are computed using either the interquartile rule or a user-defined statistical limit. The interquartile rule is used as the default when the outliers control is left in the automatic position. The user-defined statistical limit is used when you manually change the definition of the statistical minimum.

Outliers

A data value, x, is an outlier from a set of data values {x1, x2,..xn} if xi < l or xi > h.

Interquartile Rule

  and

  where k is usually 1.5.

User-Defined Statistical Limit

l = P(k) where k is a user-defined number between 0 and 24

h = P(k) where k is a user-defined number between 76 and 100

Variable

Description

l

The statistical minimum, also known as the low outlier limit

h

The statistical maximum, also known as the high outlier limit

P

Percentile

k

Either 1.5 (the default for the interquartile rule) or a user-defined percentile

The only chart that uses the statistical minimum and maximum is the box and whisker chart. The statistical minimum is shown as the extent of the whisker below the box. The statistical maximum is shown as the extent of the whisker above the box.