Procedures for Detecting Outlying Observations in Samples

Abstract

Procedures are given in the report for determining statistically whether the highest observation, or the lowest observation, or the highest and lowest observations, or the two highest observations, or the two lowest observations, or perhaps more of the observations in the sample may be considered to be outlying observations or discrepant values. Statistical tests of significance are useful in this connection either in the absence of assignable physical causes or to support a practical judgement that some of the experimental observations are aberrant. Both the statistical formulae and illustrative applications of the procedures to practical examples are given, thus representing a rather complete treatment of significance tests for outliers in single univariate samples.