Measurement Errors and Standards Questions and Answers

Answer: C

Circulating one artefact and comparing the reported values against the claimed uncertainties reveals any laboratory whose confidence exceeds its performance. It is how accreditation bodies keep the whole traceability chain honest.

Answer: D

Listing each source, its magnitude, its distribution and its sensitivity coefficient makes the combination auditable and shows at once where an improvement would pay. It is the working document behind any quoted uncertainty.

Answer: A

Multiplying each input's uncertainty by its coefficient converts them all into the units of the result so that they can be combined. An input the result barely depends on can be measured crudely, which is often where the effort is best saved.

Answer: B

Repeating the measurement and computing the standard deviation of the mean is the Type A route, and everything else, from certificates to handbook data to reasoned judgement, is Type B. The distinction concerns how the figure was obtained rather than its quality.

Answer: C

Multiplying the combined standard uncertainty by two gives roughly ninety-five per cent confidence for a normal distribution, and the factor used has to be stated with the result. Omitting it makes two quoted uncertainties incomparable.

Answer: D

Dividing the expanded figure by the coverage factor recovers the standard uncertainty, and the factor of two corresponds to about ninety-five per cent for a normal distribution. Reading the interval as a guarantee rather than a confidence statement is the usual misunderstanding.

Answer: A

One or two figures in the uncertainty and a result rounded to the same decimal place is the convention, which conveys the resolution honestly. Quoting six figures against a one per cent uncertainty misrepresents the measurement as badly as rounding away real information.

Answer: B

When a component fails in service or a dispute arises, the record is what establishes whether the original acceptance was sound and what instrument and standard stood behind it. Nothing about the record changes the measurement, but a great deal depends on being able to reconstruct it.

Answer: C

Traceability is what connects a reading to the definition of the unit, and without it no amount of resolution or repetition establishes correctness. Repetition reduces the random component only, and extra digits may show nothing but noise.