Preprints rarely change after peer review and are almost never retracted

An analysis of 70,000 studies published in Nature found that preprints — versions of papers released before peer review — rarely change significantly after the review process and are almost never retracted. The findings challenge the widespread assumption that preprints are unreliable.