Sebastian Raschka • 7/19/2014

Dixon's Q test for outlier identification

This article provides a detailed explanation of Dixon's Q test, a statistical method for outlier detection in very small sample sizes (typically 3-10 observations). It covers the test's methodology, its common application in fields like chemistry, and important criticisms regarding its assumptions and proper use. The content includes a step-by-step implementation guide and the author's perspective on cautiously using the test for outlier identification rather than automatic removal.

0 comments

#data analysis #statistics #Outlier Detection