How to find outliers step by step
- Find Q1 and Q3 (see the IQR calculator for the method).
- Compute IQR = Q3 − Q1.
- Multiply the IQR by 1.5.
- Subtract that from Q1 for the lower fence; add it to Q3 for the upper fence.
- Any value outside the fences is an outlier. Using 3 × IQR instead identifies “extreme” outliers.
What to do with an outlier
An outlier is a flag, not a verdict. First check for a typo or measurement error; fix or remove those, and say so. If the value is genuine, keep it and consider reporting resistant statistics (median and IQR) alongside the mean and SD. The z-score method (|z| > 3) is an alternative for roughly normal data, but outliers themselves inflate the SD, which can hide them.
Frequently asked questions
Why 1.5 × IQR?
John Tukey proposed it in Exploratory Data Analysis (1977) as a practical cut-off. For normally distributed data it flags about 0.7% of values, so real outliers stand out without flagging too many ordinary points.
Is a value exactly on the fence an outlier?
No. Only values strictly beyond the fences are outliers.