ClarityHow to use this tool
- Enter your measured result and the accepted reference in the same unit.
- Use a nonzero reference; percent error is undefined when the reference is zero.
- Calculate the absolute percentage error and read the signed error separately to see whether the measurement was high or low.
Input reference and sample values
Use the units and choices shown for each field. Sample values are provided for practice, not as recommendations.
| Input | Format and units | Sample |
|---|---|---|
| Measured value | Numeric value. | 98 |
| Reference value | Numeric value. | 100 |
Result for these sample inputs
2 percent
- Signed error
- -2
- Absolute error
- 2
Select Try the example in the form to load these values and calculate them again. This replaces the current form inputs. Compare the result with the method and worked explanation below.
The method, explained
Percent error = |measured − reference| ÷ |reference| × 100. Signed error is also shown separately.
For measured = 98 and reference = 100, the signed error is 98 − 100 = −2. Absolute error is 2, so percent error = 2 / 100 × 100 = 2%. With reference = −100 and measured = −98, the absolute percentage is still 2%.
Understanding your result
Percent error compares a measurement with a designated reference. Percent difference instead compares two values without treating one as accepted, while percentage change compares a later value with an earlier baseline. Select the measure that matches your question.
What to keep in mind
The reference must not be zero. Percent error is not the same as relative change over time or percentage difference between two equally weighted values.
Common questions
Can the result be negative?
Absolute percent error is nonnegative. The signed error shows whether the measurement is above or below the reference.
What if the reference is very close to zero?
The percentage may be extremely large and unstable. Consider reporting absolute error as well.
Methodology maintained by ClarityKit. How these tools are built and checked.