Back to Glossary

Data characteristic

Outlier

An observation that differs substantially from other values in a dataset.

Related domain: Data AnalyticsConcept ID: concept-outlier

Definition

An Outlier is an observation that differs substantially from other values in a dataset.

Human Explanation

It is a result that lies noticeably outside the pattern formed by most other observations.

Why it Matters

Outliers may indicate measurement errors, unusual events or important information requiring further investigation. They can also materially affect statistical summaries and models.

Conceptual Boundary

An Outlier is not automatically an error. It may represent a genuine but rare event, and whether a value is unusual depends on context and the method used to identify it.

Practical Perspective

Before removing an Outlier, investigate its origin, document the decision and evaluate how retaining or excluding it changes the analysis.

Related Concepts