Google

Extreme outliers may sometimes represent honest and accurate estimates rather than errors. To manage these valid but extreme scores, researchers can either employ resistant statistics like the median, or they can compare analyses conducted both with and without the outliers. If the results differ substantially between the two analyses, best practice dictates reporting both sets of results and thoroughly discussing the discrepancies.

Handling Valid Extreme Outliers

Extreme outliers do not necessarily indicate measurement errors or misunderstandings; they can sometimes represent honest and accurate estimates of extreme behavior. For example, in a large study of university students (Brown & Sinclair, 1999), the vast majority of participants reported having had fewer than $$15$$ lifetime sexual partners, but a few participants reported extreme scores of $$60$$ or $$70$$. While these could be intentional exaggerations or errors, it is also plausible that they reflect accurate estimates of highly atypical behavior, requiring researchers to carefully consider how to handle them analytically rather than automatically discarding them.

Example of a Valid Extreme Outlier

When a researcher finds that an extreme score in a dataset is valid and accurate (not an error), best practice is to run the analysis both with and without that score and, if the results differ substantially, report both sets of results.

Learn Before

Related