An outlier can be defined in terms of bquartiles/b, as any value which is:

either (> Q₃ + 1.5×IQR )

or (< Q₁ - 1.5×IQR )

Alternatively, an outlier can be defined in terms of the bmean and standard deviation/b, as any value which is:

either (> \bar{x} + 2\sigma )

or (< \bar{x} - 2\sigma )

Questions may specify different parameters, such as (\pm 3\sigma).

bCleaning data/b is the process of removing outliers from a data set.