How to tell if a research source is trustworthy

Not everything that looks like research is reliable. Being able to size up a source quickly — before you cite it, share it, or make a decision based on it — is one of the most valuable skills there is. Here's how to do it without a PhD.

Peer review: what it does and doesn't guarantee

Peer review means independent experts checked a paper before publication and judged it sound enough to publish. It's a real quality filter — but it's not a guarantee of truth. Reviewers can miss errors, good studies get rejected, and flawed ones slip through. Treat peer review as a positive signal, not a stamp of certainty.

A quick credibility checklist

• Who wrote it, and what's their affiliation? Real researchers at real institutions, with other work in the field, are a good sign.
• Where was it published? A recognized journal indexed in trusted databases beats one you've never heard of.
• Is the method described clearly enough that someone could repeat it?
• Is the data or code shared, so the findings can be checked?
• Do the conclusions match the evidence, or reach far beyond it?
• How old is it, and has newer work confirmed or overturned it?

Journal metrics, and why to distrust them a little

You'll see numbers like the Impact Factor (how often a journal's papers are cited) and an author's h-index (a mix of how much they publish and how often they're cited). These can hint at influence, but they're easily gamed and vary wildly between fields — a brilliant paper in a niche area will always look “small” next to a mediocre one in a huge field. Use metrics as one weak signal among many, never as the deciding factor.

Red flags

• Predatory journals that publish anything for a fee, with little or no real review.
• Claims with no data, no method, or no way to check them.
• Sensational conclusions that go far beyond what a single study could show.
• No named authors, no institution, or an “about” page that avoids specifics.
• A journal not listed in DOAJ (for open access) or major indexes.

Check whether it's been retracted

Occasionally a paper is withdrawn after publication because of errors or misconduct — but copies keep circulating as if nothing happened. Before you rely on a surprising or widely shared result, it's worth checking whether it has been retracted (databases like Retraction Watch track this). A retracted paper should never be cited as if it still stands.