Accuracy vs Precision
Accuracy
Accuracy describes how close a measured value is to the true or accepted value of the quantity. A measurement is accurate when it reflects reality — for example, a scale reading 70.0 kg for a 70.0 kg object is accurate, whatever its repeatability.
Example: A thermometer reading 100.1 °C in boiling water at sea level (true value 100 °C) is accurate.
Precision
Precision describes how close repeated measurements of the same quantity are to one another. A measurement set is precise when it is reproducible — the readings cluster tightly — even if the cluster sits far from the true value.
Example: Five weighings giving 72.3, 72.4, 72.3, 72.4, 72.3 kg are precise but inaccurate if the true mass is 70.0 kg.
Key differences
| Aspect | Accuracy | Precision |
|---|---|---|
| Definition | How close a measurement is to the true (accepted) value. | How close repeated measurements are to each other, regardless of the true value. |
| Target analogy | Hitting near the bullseye, wherever the arrows cluster. | Arrows clustered tightly together, even if they all miss the bullseye. |
| Main error source | Caused by systematic errors — faulty calibration, a zero error on a scale, a misread meniscus every time. | Limited by random errors — small unpredictable fluctuations in reading or conditions. |
| How to improve it | Improved by calibrating instruments against standards and removing bias. | Improved by repeating measurements and averaging, which smooths out random fluctuation. |
| Link to significant figures | Significant figures record how precisely a quantity was measured; reporting more digits than the instrument supports falsely implies precision. | Consistent decimal-place agreement across trials signals good precision, which the significant figures should honestly reflect. |
Precision is often mistaken for correctness — a highly precise result can still be wrong if the instrument is miscalibrated, like a scale that always reads 2 kg too heavy. Also, more decimal places do not automatically mean more accuracy; a digital readout can be precisely wrong. Finally, averaging many precise-but-inaccurate measurements does not remove the systematic error; only recalibration does.