An outlier is a data point that stands apart from the rest of a data set — sometimes a genuine extreme observation, sometimes a data-entry mistake. This calculator applies two of the most common statistical methods, the IQR (Tukey) fence rule and the z-score threshold, so you can flag suspicious values, compare methods, and export the cleaned data.

How the Outlier Calculator works

The IQR method sorts your data, finds Q1 and Q3 (the 25th and 75th percentiles), and computes the interquartile range as Q3 minus Q1. Any value beyond 1.5 times the IQR below Q1 or above Q3 is flagged. This method makes no assumption about the shape of your distribution, which makes it robust to skewed data.

The z-score method instead computes how many standard deviations each value sits from the sample mean. A value with an absolute z-score above the chosen threshold — 3 by default — is flagged. This method assumes your data is roughly normally distributed, and its own mean and standard deviation are pulled by the outliers it's trying to detect, which is worth keeping in mind on small data sets.

Selecting "Both" combines the two outlier lists so you can see where the methods agree and where they diverge.

Inputs and what they mean

Data Values — paste any list of numbers separated by commas, spaces, semicolons, or new lines. At least 4 values are required so quartiles and standard deviation are meaningful.

Detection Method — choose IQR, Z-Score, or Both. IQR tends to be the more conservative, distribution-agnostic default; Z-Score is more familiar if your data is approximately normal.

Z-Score Threshold — the number of standard deviations from the mean that counts as an outlier. 3 is the common default; lowering it to 2 flags more values as "mild" outliers, while raising it flags only the most extreme.

Limits and edge cases

With very small data sets (fewer than about 10 values), both methods can be unstable — a single extreme value pulls the mean, standard deviation, and even the quartiles, which can mask or exaggerate outliers. Highly skewed data can also produce IQR fences that flag legitimate values on the long tail as outliers.

Flagging a value as an outlier is a statistical signal, not an automatic verdict — always look at the source of the data before deciding whether to remove a flagged point. For deeper distribution analysis, pair this tool with the Standard Deviation or Descriptive Stats calculators.