FISCHER FISCHERSCOPE X-RAY XDLM 231, Operator'S Manual

Page: 116 / 236

Measurement device monitoring for the Fischerscope X-RAY Trueness and precision
114 FISCHERSCOPE
®
X-RAY
explained by the lacking precision of the instruments involved and
does not need to cause a dispute. In such cases, one can no longer
speak of different measurement results. However, if one wishes to
prove that the difference is systematic, then the precision of the
involved instruments must be increased (e.g., by increasing the
measurement time or by decreasing the measurement distance or
by increasing the number of measurements) and in this manner the
confidence interval made narrower. If this establishes a systematic
deviation, it must be researched which of the two results is correct
and which one is not or even if both are false. How to do this will be
shown in the next section.
15.6 Trueness and precision
Trueness Trueness is a property of measurement readings, namely to
possess larger or smaller systematic deviations from the correct
values.
As a rule, a systematic deviation is discovered when an instrument
has changed. It is a result of inadequate measurement device
monitoring if false readings are measured over longer periods.
However, the calibration of the instrument may also have been
wrong from the outset (e.g., through erroneous entries of nominal
values during the calibration or when using a wrongly labeled
calibration standard). Or the calibration standard in its composition
does not fit the properties of the instrument setup. A wrong
calibration can be corrected by re-measuring with one or more
standards and possibly through normalization or re-calibration.
Errors caused by faulty calibration standards or software influences
would become apparent only through comparative measurements
between different calibration standard sets. Both types of
deviations are systematic. Systematic measurement deviations
lead to situations that different users (e.g., supplier and customer)
measure different results at the same specimen.
Precision In contrast to trueness, precision is an instrument property. It can
be easily characterized, e.g., by repeated measurements on a
sample (whose true value must not even be known) at one and the
same location under repeatability conditions. The mean value
produces the requested measurement reading and the standard
deviation of the test series is a measure for the instrument’s
precision. If the measurement application has been calibrated
correctly, then the mean value of the test series is also correct and
corresponds to the true value (which is the mean value that results
from infinitely repeated measurements) within the confidence
interval.
However, if one knows the correct measurement result of a sample
at a marked location and measures it with a different instrument and
by a different user, the measurement result is typically different.