This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1 · go to the accepted answer.
Reporting a result to more decimal places than the method's precision supports is a small dishonesty that appears everywhere. A method with a two per cent relative standard deviation does not support a figure quoted to a hundredth.
I would treat the number as indicative rather than as a measurement.
I had written a reply contradicting post #29 and deleted it. Here is what survived.
My understanding of repeatability versus intermediate precision is a few years old and may have been superseded. If it has been, I would genuinely like to know rather than keep repeating it.
An observation about repeatability versus intermediate precision that I cannot explain and am posting anyway, on the principle that unexplained observations are more useful public than private.
Where a pharmacopoeial monograph exists, a method that follows it inherits a great deal of assurance. Almost nothing discussed here has one.
Where I part company with post #33, and it is a narrow parting.
Two sentences on repeatability versus intermediate precision and then I will stop, because the rest is speculation and the thread is better without mine.
What is documented is narrow. What is inferred from it is broad. The gap between them is where every argument here lives.
Practical experience of repeatability versus intermediate precision, offered as one case with the conditions stated, not as a general finding. Conditions first, because they are what make it interpretable.
The limit of quantitation determines what the impurity table can honestly contain. Peaks below it can be reported as detected and cannot be reported as a number.
Reporting the observation and leaving the explanation open deliberately.
Building on post #38 rather than restating it.
A definition problem is doing most of the work in this repeatability versus intermediate precision discussion. Once the term is pinned down I suspect the disagreement mostly goes away and what is left is small.
The most useful single question about a method: what would it fail to detect? Every method has an answer and few documents state it.
Adding a null result on repeatability versus intermediate precision. I looked, carefully, and found nothing, and null results deserve posting precisely because they never are.
Trying to state the repeatability versus intermediate precision position in a way that someone who disagrees would recognise as fair, because I do not think the version in this thread passes that test.
On post #46 — agreed on the reasoning, with one qualification.
Distinguishing three things in the repeatability versus intermediate precision discussion that keep getting used interchangeably: the observation, the proposed mechanism, and the recommendation that gets attached to both.
Picking up post #46: that is the part I would want checked first.
The most useful single question about a method: what would it fail to detect? Every method has an answer and few documents state it.
Where I would look next, rather than where I would stop.
Answering the question post #46 raises rather than the one it answers.
I would rather this thread reach "we do not know" about repeatability versus intermediate precision than reach a confident answer that nobody can support when asked.
The arithmetic in post #49 is right; the assumption feeding it is the part to check.
I changed my mind about repeatability versus intermediate precision after someone here asked me for the source and I could not produce one. That is worth saying out loud because it is the ordinary way it happens.
System suitability: injections run at the start of a batch to establish that the instrument and column are performing. Acceptance criteria typically include replicate precision (RSD ≤2%), peak tailing (0.8–1.5), theoretical plates (>2000), and resolution (>1.5).
That is my reading. Someone else read the same page differently and was reasonable.
Picking up post #52: that is the part I would want checked first.
I would be cautious about generalising from the repeatability versus intermediate precision example above. It is a good example. It is one example.
Helpful, and short, which on this subject is harder than long.
Post #54 describes the usual case. This is about the unusual one.
The most useful single question about a method: what would it fail to detect? Every method has an answer and few documents state it.
Someone will know this better than I do and I hope they say so.
Post #57 answers the question as asked. The question underneath it is different.
Whatever the answer on repeatability versus intermediate precision turns out to be, the method for getting there is the same: state the assumption, do the arithmetic in public, invite the correction.