Slide 1
Tightly grouped is not necessarily correct
Precision describes agreement among repeated measurements. A balance can give nearly identical readings while consistently reporting the wrong mass.
More to Read
Imagine a balance that was not zeroed before use. Repetition may produce a narrow cluster, but the whole cluster is displaced. More repeats alone do not remove that shared offset.
Slide 2
Separate scatter from bias
Random variation makes repeated results spread out. A systematic effect can shift many results together in the same direction.
More to Read
Temperature can change an instrument’s response over a day. Calling every discrepancy random would hide a pattern that could be measured and corrected, provided the relevant conditions were recorded.
Slide 3
Compare with a reference
Calibration compares an instrument’s indications with a reference. The comparison reveals how readings relate to values assigned by an established procedure.
More to Read
Calibration does not necessarily include adjustment. A laboratory can report a correction for the user to apply, together with uncertainty in that correction; changing the instrument is a separate operation.
Slide 4
Keep the uncertainty
Correcting a known offset improves a result, but the correction itself is uncertain. The reference and the measuring procedure both contribute.
More to Read
A reference mass is supplied with a value and associated uncertainty, not access to perfect truth. Its uncertainty joins effects such as resolution and repeatability in the final measurement budget.
Slide 5
Ask how the result was obtained
A trustworthy result needs more than many decimal places. Its method, reference chain and uncertainty make the number interpretable.
More to Read
Metrological traceability belongs to a measurement result through a documented chain of calibrations. A sticker on an instrument cannot guarantee that every later use, environment or calculation produces a traceable result.
Sources
- NIST TN 1297: Terminology
National Institute of Standards and Technology
- Metrological Traceability
National Institute of Standards and Technology