Outlier Calculator
An outlier is a value that deviates significantly from the rest of your data. This calculator identifies outliers using two complementary methods: the IQR method (distribution-free, robust) and the Z-score method (assumes normality, more sensitive). The IQR method uses the 1.5 * IQR rule, flagging values beyond the quartile fences. The Z-score method standardizes values and identifies those with absolute Z-score greater than 2 or 3, depending on how extreme you want to be. Use both methods together to understand which values are unusual and whether they warrant investigation or removal.
Outlier detection formulas
IQR Method:
Lower Fence = Q1 - 1.5 * IQR
Upper Fence = Q3 + 1.5 * IQR
Outliers: value < Lower or > Upper
Z-Score Method:
Z = (value - mean) / std dev
Outliers: |Z| > threshold (2 or 3)
Comparing the two methods
- IQR method: Distribution-free, robust, does not assume normality. Good for any data shape. Less sensitive to extreme values.
- Z-score method: Requires roughly normal distribution. More sensitive to outliers. |Z| > 2 catches about top 5%, |Z| > 3 catches about top 0.3%.
- When to use each: Use IQR for unknown or skewed distributions. Use Z-score for known normal data.
- Values flagged by both: Very likely to be true outliers.
- Values flagged by one: Worth investigating further, may be distribution-dependent.
Outlier calculator: frequently asked questions
What is an outlier?
An outlier is a data point that is unusual or extreme compared to the rest of the dataset. Outliers may represent genuine variation, measurement errors, or special cases. There is no universal definition, but statistical methods help identify candidates.
How does the IQR method work?
The IQR method uses the 1.5 * IQR rule. Lower fence = Q1 - 1.5 * IQR, upper fence = Q3 + 1.5 * IQR. Any value below the lower fence or above the upper fence is flagged as an outlier.
How does the Z-score method work?
The Z-score method standardizes each value: z = (value - mean) / std dev. A Z-score with absolute value > 2 indicates a moderately unusual value (top 5%), and > 3 indicates an extremely unusual value (top 0.3%).
Which method should I use?
The IQR method is distribution-free and robust to extreme values. The Z-score method assumes roughly normal data. For skewed data or unknown distributions, use IQR. For normally distributed data, Z-score is more sensitive.
Should I always remove outliers?
Not necessarily. Outliers may be genuine, important observations. Investigate why they are unusual before removing them. Some outliers carry valuable information about your process or population.
Official sources
- NIST/SEMATECH e-Handbook of Statistical Methods: NIST Handbook.
- American Statistical Association: ASA.
Reviewed by the CalculatorHub team, edited by James Graham, 14 June 2026. See our methodology.