Check your outliers! An introduction to identifying statistical outliers in R with easystats

Abstract

Beyond the challenge of keeping up-to-date with current best practices regarding the diagnosis and treatment of outliers, an additional difficulty arises concerning the mathematical implementation of the recommended methods. Here, we provide an overview of current recommendations and best practices and demonstrate how they can easily and conveniently be implemented in the R statistical computing software, using the *{performance}* package of the *easystats* ecosystem. We cover univariate, multivariate, and model-based statistical outlier detection methods, their recommended threshold, standard output, and plotting methods. We conclude by reviewing the different theoretical types of outliers, whether to exclude or winsorize them, and the importance of transparency.

On the rapid use of verb-control information in sentence processing

“…Once the maximally converging model has been identified, buildmer calculates value of p for all fixed effects based on Satterthwaite denominator degrees of freedom using the lmerTest package ( Kuznetsova et al, 2020 ). We followed current best practices for automatic and reproducible statistical outlier detection ( Thériault et al, 2023 ). After identifying the maximally converging model, outliers were identified using the check outliers function from the performance package ( Lüdecke et al, 2023 ).…”

Unmet expectations: social inclusion and the interaction between social anxiety and ambiguous or positive feedback

“…We used Welch t-tests per recommendations (Delacre et al, 2017). The transformed data satisfactorily met all the univariate and model-based assumptions, and there were no outliers based on three median absolute deviations (Leys et al, 2013;Thériault et al, 2023a). We report raw descriptive statistics (before transformations) of all relevant variables in Table 1.…”