Repeatability and Reproducibility of Measurements

What happens in real analytical laboratories There is very rarely sufficient time or resources to perform experiments enough times on each test material to establish a reasonable mean and standard deviation (and therefore confidence interval). Duplicate measurements are the norm rather than the exception. 

Many working laboratories will have performed a similar analysis many times over, whether it be the analysis of active ingredients for a pharmaceutical production line, or the analysis of an element in an ore for a mining company. When the analytical method was first established, method validation will have determined repeatability and reproducibility standard deviations, and these will have been verified for use in the particular laboratory. 

The repeatability standard deviation is defined as The precision of a method expressed as the standard deviation of independent determinations performed by a single analyst using the same apparatus and techniques. Hence the repeatability is what governs any replicate measurements made in your laboratory by you on the test material. The reference to independent measurements means that a number of separate test portions should be weighed, dissolved, and measured, not that the same solution should be presented to the 

Reproducibility standard deviation is The precision of a method expressed as the standard deviation of determinations performed in different laboratories. Remembering the discussion of chapter 1 we would expect the reproducibility to be greater than the repeatability, as each laboratory will have its own repeatability (which might be expected to be about the same) but the differences between the laboratories reflecting different biases will now add to this to give the reproducibility. Experience has shown that the interlaboratory reproducibility is about two to three times the repeatability. 

For completeness it should be noted that intralaboratory reproducibility is sometimes used to refer to the standard deviation of measurement results obtained within the same laboratory, but perhaps by different analysts and/or different equipment and/or different days. 

---