How to use this outlier calculator
Paste your data above and press Calculate. The tool builds the standard Tukey fences (Q1 − 1.5×IQR and Q3 + 1.5×IQR) and lists every value that falls outside them.
What counts as an outlier?
An outlier is a value that stands unusually far from the rest of the data. "Unusually far" needs a rule, and the standard one is John Tukey's: compute the interquartile range (IQR), stretch it by 1.5 in both directions from the quartiles, and flag anything beyond those fences. The rule is deliberately simple and blind to the actual extremes. That is what makes it fair.
Delivery times (minutes) for a courier over nine drops: 32, 35, 30, 33, 31, 34, 33, 32, 95. Here Q1 = 32, Q3 = 34, so IQR = 2 and the fences are 29 and 37. Eight deliveries sit comfortably inside; 95 is flagged, three times the typical drop. Flagged does not mean deleted: that 95 was a real delivery (a flooded road, most likely), and the honest summary mentions it rather than hiding it.
Flag, investigate, then decide
- Data error? A mistyped 950 for 95:0, a sensor glitch. Fix or remove, and say so.
- Genuine but exceptional? Keep it, and summarise with the median and IQR, which it cannot distort.
- A different population? A professional athlete in a class fitness dataset is not an error; perhaps it belongs in a separate analysis.
Frequently asked questions
What is the 1.5 IQR rule for outliers?
Values below Q1 − 1.5×IQR or above Q3 + 1.5×IQR are flagged as outliers. It is Tukey's rule, used by every box plot.
Should outliers be removed?
Only if they are demonstrated errors, and the removal should be reported. Genuine extreme values are information, not dirt.
Can a small dataset have outliers?
The rule needs enough values to form quartiles; with fewer than about 8–10 points it becomes unreliable, and judgement should lead.