Page 157 - Vogel's TEXTBOOK OF QUANTITATIVE CHEMICAL ANALYSIS

PREClSlON 4.4

or they rnay be prepared by the analyst and subjected to rigorous purification
by recrystallisation, etc. The substances must be of known purity. The test of
the accuracy of the method under consideration is carried out by taking varying
amounts of the constituent and proceeding according to specified instructions.
The amount of the constituent must be varied, because the determinate errors
in the procedure rnay be a function of the amount used. The difference between
the mean of an adeauate number of results and the amount of the constituent
actually present, usually expressed as parts per thousand, is a measure of the
accuracy of the method in the absence of foreign substances.
The constituent in question will usually have to be determined in the presence
of other substances, and it will therefore be necessary to know the effect of these
upon the determination. This will require testing the influence of a large number
of elements, each in varying amounts - a major undertaking. The scope of
such tests rnay be limited by considering the determination of the component
in a specified range of concentration in a material whose composition is more
or less fixed both with respect to the elements which rnay be present and their
relative amounts. It is desirable, however, to study the effect of as many foreign
elements as feasible. In practice, it is frequently found that separations will be
required before a determination can be made in the presence of varying elements;
the accuracy of the method is likely to be largely controlled by the separations
involved.

Comparative method. Sometimes, as in the analysis of a mineral, it rnay be
impossible to prepare solid synthetic samples of the desired composition. It is
then necessary to resort to standard samples of the material in question (mineral,
ore, alloy, etc.) in which the content of the constituent sought has been
determined by one or more supposedly 'accurate' methods of analysis. This
comparative method, involving secondary standards, is obviously not altogether
satisfactory from the theoretical standpoint, but is nevertheless very useful in
applied analysis. Standard samples can be obtained from various sources (see
Section 4.5).
If several fundamentally different methods of analysis for a given constituent
are available, e.g. gravimetric, titrimetric, spectrophotometric, or spectrographic,
the agreement between at least two methods of essentially different character
can usually be accepted as indicating the absence of an appreciable systematic
error in either (a systematic error is one which can be evaluated experimentally
or theoretically).

4.4 PREClSlON
Precision rnay be defined as the concordance of a series of measurements of the
same quantity. Accuracy expresses the correctness of a measurement, and
precision the 'reproducibility' of a measurement (the latter definition will be
modified later). Precision always accompanies accuracy, but a high degree of
precision does not imply accuracy. This rnay be illustrated by the following
example.
A substance was known to contain 49.10 + 0.02 per cent of a constituent A.
The results obtained by two analysts using the same substance and the same
analytical method were as follows.