Chapter 8
Measurement Uncertainty
Certain with uncertainty
In a chemical laboratory, tests are often conducted so that the final result is a mean of
a number of individual measurements, and the experimentally determined standard
deviation is used as a measure for the dispersion of the experimentally obtained data.
Thus, the standard deviation determines the precision of the measurements since it
shows how close the result was repeated in a given measurement series. It should be
noted, however, that the precision of measurements does not indicate the accuracy of
the result. It could happen that the mean value from even very precise measurements
may differ significantly from the true value or the value recognized as the reference
value.
In a chemical laboratory, tests are most often conducted in such a way that the
given result is a mean value or a median of a set of values obtained within multiple
measurements. The set of values from which the mean (or the median) is calculated might come from a series of repetitions for the single test sample or from a
series of repetitions for a few test samples taken from the primal sample. As a measurement range, in which the values can occur, a standard deviation—determined
experimentally—is used (variance or coefficient of variance are also used). Those
values determine the precision of the measurements, as they show how well the result
was repeated in a given measurement series. In Chap. 7, the terms ‘accuracy,’ ‘precision’ and ‘trueness’ have been discussed in detail. It is worth remembering that a
high precision of measurements does not guarantee their accuracy.
The term of ‘measurement error’ is closely linked to performing measurements.
According to the definition, ‘measurement error’ is the measured quantity value
minus a reference quantity value (p. 2.16, VIM 3).
According to the basic axiom of metrology, there are no measurements completely
free of errors. Therefore, when conducting measurements, one should be aware that
the results are burdened with an error. Those errors contribute to the uncertainty
connected to the course of the measurement process and its results.
© Springer Nature Switzerland AG 2018
E. Bulska, Metrology in Chemistry, Lecture Notes in Chemistry 101,
https://doi.org/10.1007/978-3-319-99206-8_8
115
Measurement Uncertainty
Certain with uncertainty
In a chemical laboratory, tests are often conducted so that the final result is a mean of
a number of individual measurements, and the experimentally determined standard
deviation is used as a measure for the dispersion of the experimentally obtained data.
Thus, the standard deviation determines the precision of the measurements since it
shows how close the result was repeated in a given measurement series. It should be
noted, however, that the precision of measurements does not indicate the accuracy of
the result. It could happen that the mean value from even very precise measurements
may differ significantly from the true value or the value recognized as the reference
value.
In a chemical laboratory, tests are most often conducted in such a way that the
given result is a mean value or a median of a set of values obtained within multiple
measurements. The set of values from which the mean (or the median) is calculated might come from a series of repetitions for the single test sample or from a
series of repetitions for a few test samples taken from the primal sample. As a measurement range, in which the values can occur, a standard deviation—determined
experimentally—is used (variance or coefficient of variance are also used). Those
values determine the precision of the measurements, as they show how well the result
was repeated in a given measurement series. In Chap. 7, the terms ‘accuracy,’ ‘precision’ and ‘trueness’ have been discussed in detail. It is worth remembering that a
high precision of measurements does not guarantee their accuracy.
The term of ‘measurement error’ is closely linked to performing measurements.
According to the definition, ‘measurement error’ is the measured quantity value
minus a reference quantity value (p. 2.16, VIM 3).
According to the basic axiom of metrology, there are no measurements completely
free of errors. Therefore, when conducting measurements, one should be aware that
the results are burdened with an error. Those errors contribute to the uncertainty
connected to the course of the measurement process and its results.
© Springer Nature Switzerland AG 2018
E. Bulska, Metrology in Chemistry, Lecture Notes in Chemistry 101,
https://doi.org/10.1007/978-3-319-99206-8_8
115