Outliers
Values that are considerably larger or smaller than the bulk of the data are called outliers.
An outlier may have been incorrectly recorded, or there may have been other anomalous circumstances associated with it. Outliers must be carefully checked if possible. If anything atypical can be found, outliers should be deleted from the data set and their deletion noted in any reports about the data.
Outliers and skew distributions
Deciding whether a value is an outlier or not is affected by the shape of the distribution of values for the rest of the data.