The power of outliers (and why researchers should ALWAYS check for them)

484 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

A Survey of Evaluation Metrics Used for NLG Systems (2020, background): “ation is lower. Pearson correlation can only reflect whether the relation between the two variables is linear (as opposed to, say, quadratic). Pearson correlation is also highly sensitive to outliers [116]. As illustrated in figure3g, the Pearson correlation changes drastically in the presence of ou”
The Tripartite Associations between Bacteriophage, Wolbachia, and Arthropods (2006, unknown): “Outliers can artificially increase or decrease the value of a correlation coefficient and their removal can have recommended benefits that are not typically considered in practice [63,64].”
Tangled up in BLEU: Reevaluating the Evaluation of Automatic Machine Translation Evaluation Metrics (2020, background): “Pearson’s r is highly sensitive to outliers (Osborne and Overbay, 2004): even a single outlier can have a drastic impact on the value of the correlation coefficient; and in the extreme case, outliers can give the illusion of a strong correlation when there is none, or mask the presence of a true rel”
Metacognitive Monitoring Accuracy and Student Performance in the Postsecondary Classroom. (2005, background): “74-76) and Osborne and Overbay (2004). No univariate outliers were found.”