How to tell if a scientific study is trustworthy

You do not need a science degree to catch the worst mistakes. Here is a concrete check you can run on any study in about two minutes. None of these steps alone is a verdict, but together they filter out most of what fools people.

The two-minute check

1. Does the DOI resolve. Real papers carry a DOI (like 10.1126/science.000000). Paste it at doi.org. If it does not resolve, be suspicious of the source you got it from.
2. Has it been retracted. This is the highest-value single check. A retracted paper has been formally withdrawn, and its claim should not be relied on. See what a retraction is, or check any DOI yourself.
3. Preprint or journal. A preprint is shared before formal review. That is not a verdict, it is a flag: weigh it, do not treat it as settled. A journal paper has been reviewed, which reduces but does not remove the risk of error.
4. Find the caveat. Good papers state their own limits: small sample, one population, correlation not cause. If the write-up you are reading has no caveat, that is usually the write-up's fault, not the paper's. Go find the limitation.

What none of this tells you

These checks confirm a paper is real, live, and not withdrawn, and they set your expectations. They do not tell you the conclusions are correct: that takes replication and time. Anyone who promises you a single trust score for science is selling something. The honest position is to weigh the evidence and keep the caveat in view.

What Lucent does for you

Lucent runs the retraction and live-DOI check on every paper before it reaches you, rewrites the finding in plain language, and always states the caveat. We check integrity signals, not whether the conclusions are correct, and we tell you exactly what we checked.

What is a paper retraction  ·  Latest issue  ·  All issues  ·  All papers  ·  All guides  ·  Privacy