How to calculate
Multiply each input’s standard uncertainty by its partial derivative. Sum the squared contributions and add twice their product times the correlation. The square root is the combined standard uncertainty.
Worked example
Use Operation A plus B · Estimate A 20 · Standard uncertainty of A 3 · Estimate B 10 · Standard uncertainty of B 4. The result is ≈ 5.
Input and result notes
This is a first-order approximation about the input estimates. Use standard uncertainties and correlation from −1 to 1. Choose units consistently.
NIST on the propagation of measurement uncertainty