Outlier Detector
Paste your data to flag unusual values. IQR is the most robust choice, Z-score suits roughly normal data, and MAD resists distortion by the outliers themselves.
Method:
Threshold:
Output:
Separator:
Statistics Result
Download CSV
| No. | Original | Outlier? | Deviation Score |
|---|
Introduction to the tool and how to use it
An outlier is an individual number that sits clearly apart from the rest of the data — it may be a data-entry error, or it may have a genuine cause, and it is often exactly the handful that deserve attention. This tool picks them out using three methods commonly used in practice.
Three detection methods:
· Interquartile range (IQR) — the most robust and widely used. It computes the 25th and 75th percentiles; their difference is the IQR. Values above upper quartile + threshold×IQR or below lower quartile − threshold×IQR are flagged (threshold defaults to 1.5; set 3 to flag only extremes);
· Z-score — measures how many standard deviations a value sits from the mean; suited to data near a normal distribution (threshold often 2 or 3);
· Median absolute deviation (MAD) — centers on the median and gauges spread by the median of deviations; it resists being skewed by the extreme values themselves better than Z-score (threshold often 2.5 or 3.5).
Note: Deviation means different things under each method — under IQR it is the excess multiple in IQR units (1.5 sits exactly on the threshold); under Z-score it is |z|; under MAD it is the deviation multiple in units of 1.4826×MAD. Larger means more abnormal, and you can sort directly by this column for a second-pass screen. Every value is output in input order so you can compare row by row against the original data; when spread is 0 (for example all values identical), there are no outliers.
Three detection methods:
· Interquartile range (IQR) — the most robust and widely used. It computes the 25th and 75th percentiles; their difference is the IQR. Values above upper quartile + threshold×IQR or below lower quartile − threshold×IQR are flagged (threshold defaults to 1.5; set 3 to flag only extremes);
· Z-score — measures how many standard deviations a value sits from the mean; suited to data near a normal distribution (threshold often 2 or 3);
· Median absolute deviation (MAD) — centers on the median and gauges spread by the median of deviations; it resists being skewed by the extreme values themselves better than Z-score (threshold often 2.5 or 3.5).
Note: Deviation means different things under each method — under IQR it is the excess multiple in IQR units (1.5 sits exactly on the threshold); under Z-score it is |z|; under MAD it is the deviation multiple in units of 1.4826×MAD. Larger means more abnormal, and you can sort directly by this column for a second-pass screen. Every value is output in input order so you can compare row by row against the original data; when spread is 0 (for example all values identical), there are no outliers.
How to use
- Paste the data into the input box; commas, spaces, line breaks, or any symbol are fine.
- Choose the detection method.
- Set the detection threshold — smaller values are more sensitive and flag more outliers.
- Choose the output: list all values (marking each row as abnormal or not) / list only the outliers.
- Set the separator to control the layout of the text-box output.
- Click start; results appear in the text box and the table below (index / value / abnormal or not / deviation).
Message board
All messages →-
No one has spoken up yet — want to go first?