accuracy
/ AK-yur-uh-see /
Think of throwing darts at a board. Accuracy asks one simple question: did your darts land on the bullseye? It does not care whether they were tightly grouped or scattered — only whether, on the whole, they hit where the true target is. A measurement is accurate when its result sits close to the real, true value of the thing you are measuring.
Accuracy is the closeness of a measured value to the true (or accepted reference) value. It is the opposite of bias: a measurement is inaccurate when something pushes it consistently above or below the truth, such as a balance that reads heavy or a contaminated reagent. You judge accuracy by measuring something whose true value you already know — a certified reference material or a standard — and seeing how far off you land.
Accuracy matters because a precise but inaccurate result is dangerously convincing: it looks reliable while being consistently wrong. The crucial caveat is that accuracy is not the same as precision. Precision is about repeatability (how tightly your darts cluster); accuracy is about correctness (how close the cluster is to the bullseye). You can have one without the other, and good analytical work demands both.
A lab analyses a certified reference material known to contain 50.0 mg/kg of lead and reports 49.8 mg/kg. Landing within 0.2 of the certified value shows the method is accurate; consistently reporting 60 mg/kg instead would reveal an accuracy problem.
Accuracy is tested against a value you already trust.
Modern metrology often splits 'accuracy' into trueness (closeness of the average to the true value) and precision; accuracy in the strict sense combines both.