Measurement Errors and Standards Questions and Answers

Answer: A

Correct answer: False, since resolution and accuracy are different properties

Extra digits let the instrument discriminate finer changes, which is resolution. Whether those digits are correct is accuracy, and that depends on calibration. A five-digit meter that is 1 per cent out displays its error to five figures - the additional digits make the wrong answer more precise, not more right.

Answer: B

Correct answer: a best estimate with a stated uncertainty

A result without an uncertainty cannot be used: the reader has no way to judge whether it agrees with another figure, meets a specification, or supports a conclusion. Reporting the best estimate together with its uncertainty, and the basis on which that uncertainty was arrived at, is what makes the measurement usable by someone else.

Answer: C

Correct answer: throws away information the measurement actually contains

Over-rounding is as much a fault as quoting spurious figures, though a less conspicuous one. If the uncertainty is 0.02 then quoting 12.3 discards a digit the measurement genuinely determined. The convention is to round the uncertainty to one or two significant figures and match the result to the same decimal place.

Answer: D

A pointer resting off zero shifts every reading by the same amount, which is the definition of a systematic error, and the adjuster exists to remove it. Checking it before and after a set of readings is standard practice.

Answer: A

Correct answer: a drift during the readings would otherwise go unnoticed

Checking the zero at both ends brackets the readings: if it has moved, the readings taken between are suspect and can be corrected or repeated. Checking only beforehand assumes stability over the whole session, which for a sensitive instrument warming up or in a changing ambient is not a safe assumption.

Answer: B

The correction from the certificate is itself uncertain, and that uncertainty is inherited by every measurement made with the instrument. It is one link in the traceability chain and has to be combined with the others.

Answer: C

Correct answer: Thermal noise in a measuring circuit

Thermal noise is a fluctuating quantity of unpredictable sign, so it scatters readings about a mean and is reduced by averaging. The other three all shift the reading consistently in one direction - an uneven scale, a persistent temperature offset, a pointer off zero - and are systematic errors, correctable but not averageable.

Answer: D

The error of a particular measurement is a definite but unknowable number, and the uncertainty is the interval within which it is believed to lie. Modern practice reports the second because the first cannot be known, and confusing them leads to statements that cannot be checked.

Answer: A

Temperature range, humidity, warm-up time, orientation and supply stability are all part of the specification, and using the instrument outside them invalidates the figure. Reading the small print before the measurement is what distinguishes a result from a number.

Answer: B

Repeating a measurement immediately with the same setup gives the narrower figure, and repeating it in another laboratory with other apparatus gives the wider one. Which is quoted matters greatly when a result is being compared with someone else's.