Accuracy and precision

Accuracy is the closeness of a measured or derived data value to its true value, while precision is the closeness of repeated measurements to each other (Fig. 8.1). A balance with a fault in it (i.e. a bias, see below) could give precise (i.e. very repeatable) but inaccurate (i.e. untrue) results. Unless there is bias in a measuring system, precision will lead to accuracy and it is precision that is generally the most important practical consideration, if there is no reason to suspect bias. You can investigate the precision of any measuring system by repeated measurements of the same sample: the nearer the replicate values are to each other, the more precise the measurement.

Absolute accuracy and precision are impossible to achieve, owing to the limitations of measuring systems. It is particularly important to avoid spurious accuracy in the presentation of results; include only those digits which the accuracy of the measuring system implies. This type of error is common when changing units (e.g. inches to metres) and in derived data, especially when calculators give results to a large number of decimal places.

Fig. 8.1 'Target' diagrams illustrating precision and accuracy.