Outlier

An outlier is an observation that differs markedly from the other data in a sample. A common rule flags values more than 1.5 times the interquartile range beyond the quartiles, or several standard deviations from the mean, though any threshold is a convention rather than a law.

Why it matters

Outliers can distort averages, standard deviations and fitted models, but they can also be the most important points: a fraud transaction, a failing sensor or a latency spike that signals an outage. The right response depends on the cause. Fix or remove clear entry errors, keep genuine extremes, and use robust statistics such as the median when extremes should not dominate.

Example

Daily visitors: 1020, 980, 1010, 995, 15000
IQR rule flags 15000.
Cause A: bot traffic -> filter it out
Cause B: viral article -> keep and analyse it

Common mistakes

• Deleting outliers automatically without investigating.
• Keeping obvious errors, such as a negative age, in the analysis.
• Using methods that assume normal data, such as the mean and SD, on data with heavy tails.

Related

• Mean vs median
• Standard deviation
• Data cleaning
• Clustering

This page is general information written by our editorial team and updated when documentation changes. Servers and providers differ, so test changes safely first. See our editorial policy and disclaimer.