Analyze values, quartiles, fences, and suspicious observations instantly. Separate regular, mild, and extreme cases clearly. Save reports, review trends, and support cleaner analytic decisions.
This sample shows one mild outlier and one extreme outlier.
| Rank | Value | Expected Class |
|---|---|---|
| 1 | 12 | Regular |
| 2 | 13 | Regular |
| 3 | 13 | Regular |
| 4 | 14 | Regular |
| 5 | 15 | Regular |
| 6 | 15 | Regular |
| 7 | 16 | Regular |
| 8 | 16 | Regular |
| 9 | 17 | Regular |
| 10 | 18 | Regular |
| 11 | 25 | Mild Outlier |
| 12 | 40 | Extreme Outlier |
Mild outlier analysis uses quartiles and the interquartile range.
Q1 is the lower quartile. Q3 is the upper quartile. IQR = Q3 − Q1.
Lower Inner Fence = Q1 − 1.5 × IQR
Upper Inner Fence = Q3 + 1.5 × IQR
Lower Outer Fence = Q1 − 3 × IQR
Upper Outer Fence = Q3 + 3 × IQR
Values outside the inner fences are outliers. Values between inner and outer fences are mild outliers. Values beyond outer fences are extreme outliers.
Mild outliers are unusual values. They are not always errors. They often show shift, segmentation, behavior change, or process drift. Data teams review them before removing anything. That protects useful evidence and improves decision quality.
The calculator sorts the dataset first. It then estimates Q1, the median, and Q3. These values describe the middle structure of the data. The interquartile range shows spread inside the central half. Inner and outer fences create practical review limits. Those limits work well when a quick nonparametric screen is needed.
A mild outlier should trigger a question. Was there data entry noise? Was a customer segment different? Did a batch change? Did a campaign create a spike? A mild outlier can be valid and important. The classification should start analysis, not end it.
This page combines descriptive statistics, ranked classification, export tools, and a Plotly graph. That makes review faster. Analysts can inspect distributions, preserve a table for reporting, and compare suspicious values against computed fences. The example table also helps teams verify interpretation before using live data.
Always pair outlier detection with domain knowledge. A value can be statistically unusual and still operationally correct. Use this calculator as a screening step inside a broader data quality workflow.
A mild outlier falls outside the inner fences built from the interquartile range. It is unusual, but not as far from the middle spread as an extreme outlier.
Mild outliers are outside the 1.5 × IQR fences. Extreme outliers are outside the 3 × IQR fences. Extreme values usually need stronger investigation.
Yes. The calculator sorts valid numeric values before computing quartiles, fences, ranks, and classifications. You can paste the data in any order.
Yes. The parser accepts integers, decimals, and negative values. It ignores invalid tokens and reports how many tokens were skipped.
Not automatically. Some mild outliers are meaningful observations. Review source quality, business context, and downstream impact before deleting or winsorizing values.
Quartile-based analysis needs enough data to split the sorted sample into meaningful sections. Very small datasets can produce unstable fences and weak interpretation.
If Q1 and Q3 are equal, the inner and outer fences collapse. In that situation, repeated identical values may appear regular while distant values can become extreme.
The CSV export saves summary metrics and the ranked classification table. The PDF export creates a compact report with the same computed tables.
Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.