This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1.
Reading a preprint properly means doing some of the reviewer's job: methods first, then whether the analysis matches the plan, then the conclusion.
One more caveat and then I will stop qualifying: the sample selected itself.
Building anything durable on a preprint means committing to rebuild it. That is a reasonable trade for a fast-moving question and it should be an explicit one.
I would want the raw data before agreeing with my own summary of it.
The arithmetic in post #62 is right; the assumption feeding it is the part to check.
Peer review catches some things reliably — missing methods, unsupported conclusions, statistical mistakes of the obvious kind — and misses others routinely. Knowing which is which makes preprint reading easier.
The most useful convention this subcategory has: state the status, state the version, state the date you checked. Three short clauses and the citation stays honest.
That is the shape of it. The detail is where I would expect to be corrected.
Post #64 and I disagree about the size of the effect, not about the direction.
Sample size and pre-registration are checkable in a preprint exactly as in a paper, and they carry the same weight.
Unrefereed work is not worthless. A great deal of it survives review substantially unchanged. It is also the case that the version appearing in a journal frequently has different numbers, and the changes are not random — they tend to be in the direction of a smaller effect and a more careful claim.
I keep a log of this specifically because memory is unreliable about it.
Screening on a preprint server is minimal and is not review. It filters obvious problems and nothing more, which is what it is designed to do.
Author track record is a weak signal and it is a signal. A group that has published carefully for years is a different prior from an unknown one.
I have seen it go both ways, which is why I hedge.
Bookmarking this. I will come back when I have something worth adding.
Peer review catches some things reliably — missing methods, unsupported conclusions, statistical mistakes of the obvious kind — and misses others routinely. Knowing which is which makes preprint reading easier.
The reasoning is more useful than the number, which is why I have shown it.
Adding the measurement that post #71 says would settle it.
Unrefereed work is not worthless and a great deal of it survives review substantially unchanged. That is the accurate framing and it is not a licence to treat it as final.
Where I part company with post #71, and it is a narrow parting.
Unrefereed work is not worthless. A great deal of it survives review substantially unchanged. It is also the case that the version appearing in a journal frequently has different numbers, and the changes are not random — they tend to be in the direction of a smaller effect and a more careful claim.
Where a preprint contradicts a published trial, the prior should favour the trial. Where it extends one, the preprint can be genuinely useful.
Worth reading the earlier posts in this thread before acting on mine.
How to cite a preprint: author last name, first initial, et al. Title. Server name [preprint]. Posted date. Doi. Include the server name and note that it is a preprint.
A single observation, in a thread that deserves better than single observations.
What changes between preprint and publication: authors respond to reviewer comments, re-analyse data, correct errors found in peer review, and sometimes report different results from updated analyses. The published version is not the same as the preprint.
I would be glad to be shown a cleaner way of putting this.
Reading rather than contributing, but this is the most useful thread I have found on it.
Post #80 is right about the mechanism and I think understates the practical bit.
Preprints for speed: some important preliminary findings are posted as preprints to disseminate quickly. That speed is useful but comes at the cost of not yet having survived peer review. The trade-off is real.
Good question, well framed, and I would like to see it answered properly.
Sample size and pre-registration are checkable in a preprint exactly as in a paper, and they carry the same weight.
I had written a reply contradicting post #83 and deleted it. Here is what survived.
Author track record is a weak signal and it is a signal. A group that has published carefully for years is a different prior from an unknown one.
Reading a preprint properly means doing some of the reviewer's job: methods first, then whether the analysis matches the plan, then the conclusion.
The interesting part of this is the exception, and I do not understand the exception.
Screening on a preprint server is minimal and is not review. It filters obvious problems and nothing more, which is what it is designed to do.
Written from notes rather than memory, which is why the numbers are specific.
That matches what I have seen, for whatever a single anecdote is worth.
How to cite a preprint: author last name, first initial, et al. Title. Server name [preprint]. Posted date. Doi. Include the server name and note that it is a preprint.
I have kept the units in throughout, for the obvious reason.