Concept
Separate repeatable measurements from correct measurements
A clock can be reliably five minutes wrong.
When it fits
- A device or process gives very consistent readings and you are tempted to call it accurate.
When to avoid it
- The reference itself has uncertainty, and a check at one value does not validate every operating condition.
The idea
Precision describes agreement among repeated measurements; accuracy concerns closeness to the relevant reference. Repeating the same biased measurement reduces neither the bias nor the need for a suitable check. Ask which property the evidence actually demonstrates.
An example
A scale that consistently reads a known reference mass too high is precise in those readings but not accurate for that reference.
Check your result
Repeatability and reference agreement are reported separately.
Keep this limit in mind
- The reference itself has uncertainty, and a check at one value does not validate every operating condition.
Evidence and sources
Supports
Precision concerns agreement among repeated measurements; accuracy concerns agreement with the relevant reference value.
Real measurement assessment also requires suitable references and uncertainty, not merely two labels.
Chemistry 2e, 1.5: Measurement Uncertainty, Accuracy, and Precision · Accuracy and Precision