Comparative precision
The comparative precision (also comparability or traceability ) is defined in measurement technology , especially in analytical chemistry , as the precision under comparative conditions.
The comparability is therefore the degree of agreement between results of measurements of the same analyte in different sub-samples of a substance quantity. The measurements are carried out under changing conditions (“comparison conditions”), but always using the same procedure. Can differ z. B .:
- the measuring instruments used (of a similar type),
- the places or times of measurement,
- the person taking the measurements,
- the general experimental conditions as long as the procedure remains the same.
Application in analytical chemistry
Comparative conditions are usually present when different laboratories carry out analyzes using the same procedure. The comparative precision is therefore an important parameter in round robin tests in which several laboratories participate. A sample with known content (e.g. a reference material ) is sent to the laboratories. Using a specific, defined procedure, a defined measured variable is analyzed in each laboratory. The laboratories deliver individual results with details of the standard deviation . Based on the known values of the measured variable in the sample material, it can be determined which results are accepted and which are rejected as incorrect. The laboratories whose results are accepted have passed the round robin test.
The comparability (comparative precision) "x" of the accepted results (at a 95% confidence level ) is given by multiplying the standard deviation of the individual results from a laboratory by the factor 2.8.
If two laboratories examine partial samples of the same sample and the results differ by a greater amount than the comparative precision x, then the quality of at least one result must be questioned. A comparison is then not permitted (there is no comparability). Either the method has to be improved or another method with better comparative precision has to be used.
See also : repeatability , laboratory precision , measurement system analysis