124
8 Measurement Uncertainty
8.6 Types of Measurement Errors
The gross error is connected with the occurrence in the set of results of one that
deviates significantly from the other values in the series. The sources of the gross
error are most often significant and atypical disturbances of the measuring system
or an error of the person conducting the measurements. If the difference between
the individual result and the rest of the results in a given measurement series is
significant, that result can be arbitrarily rejected or, in the case of doubt, appropriate
statistical tests can be used to reject deviating results. (For the rejection of deviating
results, statistical tests are used.)
The systematic error is connected, for example, with an incorrect setup of the
measuring device, which means that the measured value is systematically underestimated or overestimated with a systematic influence of outside factors on the read (e.g.,
the influence of temperature on the pH measurement). A systematic error remains
constant over a series of measurement and it cannot be reduced by increasing the
number of replicate measurements.
Apart from gross and systematic errors, random errors are also important, meaning those errors that are the most responsible for randomly spread values when
conducting multiple measurements of the same quantity (measuring series). All measurements, even those conducted with the utmost care, are subjected to the influence
of various random factors. In that case, we distinguish two sources of measurement
errors: outside factors (the influence of the surroundings, e.g., changes in temperature during measurements) and internal factors (e.g., the stability of the measuring
device, the quality of the used measuring glassware). The effects connected with the
influence of the outside factors can, to a certain degree, be controlled by the person
conducting the measurements, whereas the effects connected with the influence of
the internal factors are closely connected to the measurement itself and cannot be
removed. Therefore, it is important to correctly calculate or estimate the uncertainty
value, since that is the range in which we expect the true value to be.
At this point, it is worth emphasizing again the relationship between the term ‘measurement error’ and the term ‘measurement uncertainty.’ The error always defines
the difference between the two values, and in the case of the measurement error,
it is a difference between the true value and the value obtained as the result of the
measurement. It follows that for each measurement result, the error will assume a
different value (positive, negative or it can also be zero). Uncertainty, on the other
hand, is a parameter that characterizes the spread of the measurement results—that
is, one that determines the variability limits of the measurement results.
Measurement errors can be systematic and random
Systematic measurement error: a component of the measurement error that in
replicate measurements remains constant or varies in a predictable manner.
Clause 2.17; ISO/IEC Guide 99
8 Measurement Uncertainty
8.6 Types of Measurement Errors
The gross error is connected with the occurrence in the set of results of one that
deviates significantly from the other values in the series. The sources of the gross
error are most often significant and atypical disturbances of the measuring system
or an error of the person conducting the measurements. If the difference between
the individual result and the rest of the results in a given measurement series is
significant, that result can be arbitrarily rejected or, in the case of doubt, appropriate
statistical tests can be used to reject deviating results. (For the rejection of deviating
results, statistical tests are used.)
The systematic error is connected, for example, with an incorrect setup of the
measuring device, which means that the measured value is systematically underestimated or overestimated with a systematic influence of outside factors on the read (e.g.,
the influence of temperature on the pH measurement). A systematic error remains
constant over a series of measurement and it cannot be reduced by increasing the
number of replicate measurements.
Apart from gross and systematic errors, random errors are also important, meaning those errors that are the most responsible for randomly spread values when
conducting multiple measurements of the same quantity (measuring series). All measurements, even those conducted with the utmost care, are subjected to the influence
of various random factors. In that case, we distinguish two sources of measurement
errors: outside factors (the influence of the surroundings, e.g., changes in temperature during measurements) and internal factors (e.g., the stability of the measuring
device, the quality of the used measuring glassware). The effects connected with the
influence of the outside factors can, to a certain degree, be controlled by the person
conducting the measurements, whereas the effects connected with the influence of
the internal factors are closely connected to the measurement itself and cannot be
removed. Therefore, it is important to correctly calculate or estimate the uncertainty
value, since that is the range in which we expect the true value to be.
At this point, it is worth emphasizing again the relationship between the term ‘measurement error’ and the term ‘measurement uncertainty.’ The error always defines
the difference between the two values, and in the case of the measurement error,
it is a difference between the true value and the value obtained as the result of the
measurement. It follows that for each measurement result, the error will assume a
different value (positive, negative or it can also be zero). Uncertainty, on the other
hand, is a parameter that characterizes the spread of the measurement results—that
is, one that determines the variability limits of the measurement results.
Measurement errors can be systematic and random
Systematic measurement error: a component of the measurement error that in
replicate measurements remains constant or varies in a predictable manner.
Clause 2.17; ISO/IEC Guide 99