An outlier is a data point that is very different from most of the other data points. For example, if the mean math test score for a class is 45%, and one student gets 98%, that student could be considered an outlier. It is up to the analyst to decide exactly how far out a data point needs to be before it is called an outlier.

Outlier data points should be checked carefully to make sure there was not an error in data entry.

See also: data cleaning

Return to the Evaluation Dictionary