 |
 |
Solutions
|
|
Accuracy and Precision Accuracy and precision are often confused. Accuracy gives an indication how close a test result is to the "true" value, or, more practically, how close it might be to the average of the results from a number of independent laboratories. Laboratories monitor their accuracy either by repeatedly measuring samples or things with known values, or by comparing themselves to other laboratories in proficiency trials. Both methods are routinely used in wool testing. Precision is a measure of how repeatable a result might be. Precision is often split into two components – within laboratory and between-laboratory. The overall precision includes both components. It is often expressed as the 95% Confidence Limit or 95%CL, which is the range within which a result is expected to fall 19 times out of 20.
In practice, clients are often interested in how close two results should be from two independent laboratories. This can be expressed as the Maximum Probable Difference, or MPD, and gives the maximum expected difference which will occur 19 times out 20 between two independent results obtained on the same lot of wool. Statistically, once in 20 times, differences could fall outside the MPD.
Most of the test methods under which we operate have well-documented precision data. IWTO has a method for determining precision data and for comparing test methods and instruments (IWTO-0). Further information can be obtained from the following Info-bulletins: |
