Outliers May Not Be Automatically Removed

Researchers often remove outliers when comparing groups. It is well documented that the common practice of removing outliers within groups leads to inflated Type I error rates. However, it was recently argued by André (2022) that if outliers are instead removed across groups, Type I error rates are...

Full description

Saved in:
Bibliographic Details
Published in:Journal of experimental psychology. General 2023-06, Vol.152 (6), p.1735-1753
Main Author: Karch, Julian D.
Format: Article
Language:English
Subjects:
Concepts
Confidence intervals
Error Analysis
Group Differences
Human
Hypotheses
Hypothesis Testing
Type I Errors
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Researchers often remove outliers when comparing groups. It is well documented that the common practice of removing outliers within groups leads to inflated Type I error rates. However, it was recently argued by André (2022) that if outliers are instead removed across groups, Type I error rates are not inflated. The same study discusses that removing outliers across groups is a specific case of the more general concept of hypothesis-blind removal of outliers, which is consequently recommended. In this paper, I demonstrate that, contrary to this advice, hypothesis-blind outlier removal is problematic. Specifically, it almost always invalidates confidence intervals and biases estimates if there are group differences. It moreover inflates Type I error rates in certain situations, for example, when variances are unequal and data nonnormal. Consequently, a data point may not be removed solely because it is deemed an outlier, whether the procedure used is hypothesis-blind or hypothesis-aware. I conclude by recommending valid alternatives. Public Significance Statement Removing outliers is widespread in psychological research. This study demonstrates that removing a data point solely because it is an outlier is problematic. As a remedy, robust techniques are recommended.
ISSN:0096-3445
1939-2222
DOI:10.1037/xge0001357