Instant Notes: Analytical Chemistry

Section B – Assessment of data

B2 ASSESSMENT OF ACCURACY

AND PRECISION

These two characteristics of numerical data are the most important and the most

frequently confused. It is vital to understand the difference between them, and

this is best illustrated diagrammatically as in Figure 1. Four analysts have

each performed a set of five titrations for which the correct titer is known to be

20.00 cm^3. The titers have been plotted on a linear scale, and inspection reveals

the following:

● the average titers for analysts B and D are very close to 20.00 cm^3 - these two

sets are therefore said to have good accuracy;

● the average titers for analysts A and C are well above and below 20.00 cm^3

respectively -these are therefore said to have poor accuracy;

● the five titers for analyst A and the five for analyst D are very close to one

another within each set – these two sets therefore both show good precision;

● the five titers for analyst B and the five for analyst C are spread widely

within each set -these two sets therefore both show poor precision.

Accuracy and
precision

Key Notes

Accuracy is the closeness of an experimental measurement or result to

the true or accepted value. Precision is the closeness of agreement

between replicated measurements or results obtained under the same

prescribed conditions.

The standard deviation of a set of values is a statistic based on the normal

error (Gaussian) curve and used as a measure of precision.

Relative standard deviation (coefficient of variation) is the standard

deviation expressed as a percentage of the measured value.

A standard deviation can be calculated for two or more sets of data by

pooling the values to give a more reliable measure of precision.

This is the square of the standard deviation, which is used in some

statistical tests.

An estimate of the overall precision of an analytical procedure can be

made by combining the precisions of individual measurements.

This is the range of values around an experimental result within which

the true or accepted value is expected to lie with a defined level of

probability.

Related topic Errors in analytical measurements (B1)

Accuracy and

precision

Standard deviation

Variance

Relative standard

deviation

Pooled standard

deviation

Overall precision

Confidence interval