Publishing an inconvenient result: a case study — the long version Writing it up because I had to work it out twice and would rather nobody else did.
A narrow question about Publishing an inconvenient result, deliberately narrow, because the broad version has been asked here four times and produced four long threads and no answer.
One question, stated units, stated method, and what I have already ruled out.
This follows the opening post rather than contradicting it.
Blind sampling: submitting without the laboratory or seller knowing the source removes the two easiest ways for a result to be shaped: selective submission and selective reporting. The extra effort is worth it.
What would change my mind on Publishing an inconvenient result is a second dataset collected by someone with no stake in the first. Until then I hold it loosely and I would rather say so than pretend to more.
I changed my mind about Publishing an inconvenient result 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.
The opening post and I disagree about the size of the effect, not about the direction.
Verification records are dated because they expire in usefulness. A result from two years ago describes a lot from two years ago.
Taking post #5 at face value and following it one step further.
A note on scope: what I am saying about Publishing an inconvenient result applies to the case in the first post and I would not extend it further without checking.
Post #5 put the caveat in the right place and I want to underline it.
Practical note on Publishing an inconvenient result: write down what you expect before you look. The number of times I have found what I went looking for is higher than chance would allow.
I disagree with the framing of Publishing an inconvenient result above, and I think it is a substantive disagreement rather than a terminological one. Setting out why, so it can be checked.
The reasoning depends on an assumption that is doing a lot of work and is never stated. If the assumption holds, the conclusion follows. I do not think it holds generally.
Post #10 is the version of this I will quote in future. One addition.
The record shows what was measured. Everything beyond that is inference and this site tries to keep the line visible.
Marking that as an opinion rather than a finding.
Where I part company with post #8, and it is a narrow parting.
Where two services agree with each other and with the declared figure, that is a genuinely strong result and it applies to that lot.
That is what the documentation says. What happens in practice is usually close.
I have been on both sides of the Publishing an inconvenient result argument in this category within eighteen months, which should tell you how strong the evidence for either side is.
Repeat determination on the same product line months apart is the measurement that speaks to consistency, and consistency is what most buyers actually want.
I checked the source rather than the summary, and they differ.
Post #12 put the caveat in the right place and I want to underline it.
Reading this Publishing an inconvenient result thread as someone who came in with a fixed view: the third and seventh replies moved me and the confident ones did not.
Asking the supplier before publishing a discrepancy is the convention here. Most discrepancies are explained by handling or method and asking costs a day.
Repeat testing: if you want to verify the same lot a second time, testing with a different service strengthens the conclusion. Testing with the same service verifies reproducibility but uses the same method.
What I would check first on Publishing an inconvenient result is whether the thing being measured moved or whether the way of measuring it moved. Those look identical in a graph.
No disagreement from me. Posting only so the question does not look ignored.
The record shows what was measured. Everything beyond that is inference and this site tries to keep the line visible.
Worth separating two things that post #22 runs together.
Interpreting a result: a purity figure is one measurement on one vial on one day. Agreement across two services is much stronger. A single high number does not establish a pattern.
If this contradicts something upthread, the upthread version may well be the better one.
This follows post #22 rather than contradicting it.
Combining a chromatographic result with a documentation review gives a more complete picture than two chromatographic results, because the failure modes are different.
The part I am sure of is shorter than the part I have written.
Where two services agree with each other and with the declared figure, that is a genuinely strong result and it applies to that lot.
Narrowing post #25, because the general version has more than one answer.
An observation about Publishing an inconvenient result that I cannot explain and am posting anyway, on the principle that unexplained observations are more useful public than private.
Post #25 and I disagree about the size of the effect, not about the direction.
Two sentences on Publishing an inconvenient result 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.
A record with four results over sixteen months is a thin record rather than a bad one, and the honest description says which.
Filing this under things that are true until someone shows me otherwise.
Answering the question post #25 raises rather than the one it answers.
Whatever the answer on Publishing an inconvenient result turns out to be, the method for getting there is the same: state the assumption, do the arithmetic in public, invite the correction.
Blind sampling: submitting without the laboratory or seller knowing the source removes the two easiest ways for a result to be shaped: selective submission and selective reporting. The extra effort is worth it.
Noting that the question and the thing people usually mean by it are different.