How to read a scientific study

A practical checklist for reading a paper, or a headline about one, like a scientist.

VALREXA Biosciences · 8 min read · November 2026

Key takeaways

Start with who or what was studied: cells, animals or people.
Check size, controls, randomization, blinding and the pre-registered main outcome.
Look at absolute numbers, funding and whether the result has been replicated.

The anatomy of a paper

Section	What to look for
Abstract	A summary. Useful, but it often leaves out limitations.
Introduction	The question and why it matters.
Methods	The most important section: who was studied, how, and what was measured.
Results	The actual numbers, tables and figures.
Discussion	The authors’ interpretation, including limitations.
Funding and conflicts of interest	Who paid for the work and who might benefit.

The ten-question checklist

Who or what was studied? Cells, animals or humans? Healthy volunteers or patients?
How many? Small studies are more likely to produce chance findings.
Was there a control group? Without one, you can’t tell what would have happened anyway.
Was it randomized and blinded?
What was the main outcome, and was it chosen in advance? Check the trial registration. Switching outcomes after the fact is a red flag.
Was the outcome meaningful? A change in a blood marker is not the same as people living longer or feeling better.
How big was the effect? Look for absolute differences, not just percentages. See Statistics that matter.
How long did it last? Short studies can miss late benefits and late harms.
Who funded it? Industry funding doesn’t make a study wrong, but it is worth knowing.
Has it been replicated? One study is a starting point. Check whether independent groups found the same thing.

Preprints, peer review and retractions

A preprint is posted before peer review and may change. Peer review is a quality check, not a guarantee. Papers are sometimes retracted for errors or misconduct; the Retraction Watch Database tracks them.

Reading headlines

Ask: was this in mice? Is it a correlation? What was the actual size of the effect? Many “breakthrough” stories describe early animal or cell results. Our guide to misleading health claims covers the most common tricks.