Report uncertainty with the result
Distinguish variability, measurement uncertainty and missing knowledge.
Research note · November 14, 2026Research question
What uncertainty accompanies a timing, accuracy or physical measurement?
Method
- Identify the quantity being estimated and the sources of variation. Separate repeated-run variation from assumptions about instruments, models or unobserved conditions.
- Choose an uncertainty procedure appropriate to the sampling design. Preserve repetitions and grouping so dependent measurements are not treated as independent evidence.
- Report the estimate, uncertainty expression, coverage interpretation and assumptions together. Do not label an unexplained plus-or-minus value as a confidence interval.
Records to retain
- Measured quantity and units
- Repeated observations
- Uncertainty components
- Calculation and coverage assumptions
Source basis
NIST TN 1297 distinguishes Type A and Type B evaluations of uncertainty and discusses combining and reporting uncertainty. The terminology matters because a standard deviation, standard uncertainty and expanded uncertainty are not interchangeable labels.
Interpretation & limits
Uncertainty estimates depend on their model and data. They cannot compensate for an unrepresentative population or an invalid measurement procedure. No uncertainty interval is calculated in this note.
Next experiments
- Document the dominant uncertainty sources for one bounded measurement.
- Retain raw observations alongside the published summary.
References
Technical reference
Implementation details supporting this note, drawn directly from the maintained documentation. These are technical references, not additional papers or new experimental results. Expand a topic to inspect its method, fields and limits.