ABSTRACT
Data often contain outliers, possibly more often than is realized as they are not always detected. Outliers may be due to transmission mistakes such as an incorrect decimal place, or they may be due to faulty measurement without any deeper significance or they may be the most important observations in the data (see [175] on the Antarctic ozone hole1). Undetected outliers can lead to a misleading analysis of the data or to a failure to detect the most important observations.
