peer review

Peer review is the check a journal runs before it publishes a scientific claim. The Committee on Publication Ethics, which sets standards for scholarly publishing alongside three other bodies, defines it as obtaining advice on manuscripts from reviewers or experts in the manuscript's subject area, and specifies that those individuals should not be part of the journal's editorial team.

Independence is the whole point. Reviewers who did not do the work, and who do not gain if it is celebrated, read the methods, the data and the conclusions and say whether the claim is actually supported. They can recommend rejection, ask for more evidence, or send the authors back to reanalyse what they have. A good deal of work that does not survive that process never becomes a paper at all.

It is a filter rather than a guarantee. Peer review does not normally involve repeating the experiment, reviewers miss things, and published papers are sometimes later shown to be wrong. What it establishes is that people qualified to spot the obvious problems looked and did not find them. That is why a peer-reviewed result and a press release describing the same work are not equivalent evidence, and why we say which of the two we are reading.