A simple method for the detection of outliers

Abstract

This article presents a simple and efficient method to detect multiple outliers using a modification of the Akaike's Information Criterion. The proposed method is exemplified using various examples and implemented in the statistical package R. The core features of the method are that it depends on simple computations, it does not depend on traditional p values, and it signals the absence of outliers. The proposed procedure is applicable to analyse data in all sciences.