measurement uncertainty
/ MEZH-er-ment un-SER-tin-tee /
If a friend tells you a trip takes 'about 30 minutes', they are admitting it might be 25 or 35. Every measurement carries that same honest 'about'. Measurement uncertainty is the size of the doubt around a measured value — a way of saying not just what you got, but how sure you are of it.
More precisely, measurement uncertainty is a quantified estimate of the range within which the true value of a measured quantity is believed to lie. It is usually reported as a value plus-or-minus a margin (for example, 25.0 plus or minus 0.2 mL). It gathers together random scatter, found by repeating the measurement, and systematic effects, judged from the instrument and method. It is not the same as a mistake or blunder.
Why it matters: without it, a number is hard to use — you cannot tell whether two results really differ, or whether a measurement supports a theory. A caveat: uncertainty is an honest estimate, not a guarantee; quoting too small a margin overstates your confidence, and quoting too large a one throws away real information. Good practice states clearly how the figure was obtained.
Two labs report a reaction's enthalpy as -57.3 plus or minus 0.4 kJ/mol and -57.9 plus or minus 0.3 kJ/mol. Their ranges overlap, so the results agree within uncertainty — there is no real conflict. Had the uncertainties been ten times smaller, the same two numbers would have signalled a genuine disagreement to chase down.
Whether two results agree depends on their uncertainties, not just on the numbers.
Uncertainty and error are often confused. An error is the actual (usually unknown) difference between a measured value and the true value; uncertainty is our estimate of how large that difference could be. A genuine mistake — misreading a scale, mislabelling a flask — is neither, and cannot be fixed by an uncertainty figure.