What Is Peer Review?

Direct Answer

Peer review is the process by which experts in a field — the author's "peers" — evaluate a research manuscript before it is published, checking whether the methods are sound, the evidence supports the conclusions, and the contribution is novel and worth publishing. In the standard model, a researcher submits a paper to a journal; the editor sends it to two or three anonymous reviewers with expertise in the topic; the reviewers submit detailed critiques and a recommendation — accept, revise, or reject; the editor decides; and the author revises accordingly, sometimes through several rounds. The process is the scientific community's first line of quality control: it catches errors, demands rigor, and prevents the most obviously flawed work from entering the literature. It is not, however, a stamp of truth. Peer review filters; it does not certify.

Historical Context

Peer review grew slowly with scientific publishing. The earliest scientific journals, including the Philosophical Transactions of the Royal Society founded in 1665, relied on the judgment of an editor and the informal scrutiny of the society's members — a system rooted in Bacon's insistence that claims be examined rather than taken on authority. The modern practice of sending manuscripts to external, anonymous referees emerged in the eighteenth and nineteenth centuries, when journals such as the Philosophical Transactions and the Lancet began formalizing referee reports, and it became standard across disciplines in the twentieth century as the volume of research exploded. The rationale reflects the epistemology of Popper: a scientific claim earns its provisional place only by surviving criticism, and peer review institutionalizes that criticism — a community's attempt to falsify a claim before it is published rather than after. It is, in effect, the scientific method applied to the scientific literature itself.

Key Distinctions

Several distinctions clarify how peer review works. Single-blind versus double-blind versus open: in single-blind review, reviewers know the author's identity but the author does not know the reviewers'; in double-blind, neither knows the other; in open review, identities are disclosed and reports may be published. Each form trades anonymity against accountability: double-blind reduces bias by reputation, while open review increases transparency and credit. Pre-publication versus post-publication: traditional review happens before publication; post-publication review — comments, replication attempts, and corrections — continues afterward and is increasingly important. Journal peer review versus grant peer review: the same logic of expert evaluation governs funding decisions. Reviewing versus refereeing versus editing: reviewers advise, editors decide. And the key epistemic distinction: peer review as filter versus peer review as guarantee — a peer-reviewed paper has passed a minimal threshold of plausibility; it has not been proven true, and its claims remain open to the tests of replication and further research.

Contemporary Debates

Peer review is widely criticized and widely defended. Critics note that it is slow, expensive, and imperfect: reviewers are unpaid volunteers, they can miss statistical flaws, and the process is vulnerable to bias — reviewers may favor well-known authors, reject novel work, or be influenced by institutional prestige. The replication crisis proved the point: many peer-reviewed findings failed to reproduce, showing that peer review does not catch p-hacking, small samples, or flexible analysis. Defenders reply that peer review remains the least bad system: it improves the average quality of the literature, catches the most obvious errors, and provides a benchmark for readers. The reform agenda includes registered reports (journals reviewing the design and plan before data collection, committing to publish regardless of outcome), open peer review, and post-publication evaluation. These reforms do not replace peer review; they strengthen the system around it, recognizing that quality control must continue after publication as well as before.

Practical Implications

For readers of science, peer review is a useful but limited signal. "Peer-reviewed" means the work survived expert scrutiny — a reason to take it seriously, but not a reason to treat it as established fact. Combine the signal with the other checks: Was the study replicated? Was it pre-registered? Was the design strong, the sample adequate, the effect size meaningful? Treat one peer-reviewed study as one piece of evidence, and prefer claims supported by convergence across many studies and independent teams. For researchers, peer review is a professional duty: reviewing honestly and carefully is part of the collective enterprise of keeping science self-correcting. For citizens, the lesson is broader: peer review is an example of institutionalized doubt — a system designed to make claims earn their place — and understanding it is part of reading research well.

Related Concepts

• How to Read a Research Study — what to do with a peer-reviewed paper
• What Is the Replication Crisis? — why peer review is not enough
• What Is the Scientific Method? — the larger system peer review serves
• What Is Empirical Evidence? — what reviewers check for
• What Is Falsification? — the critical spirit peer review institutionalizes

Further Learning

• Read the OUP guide to The Peer Review Process
• Explore Reproducibility of Scientific Results at the SEP
• See why publication alone is insufficient in what is the replication crisis
• Visit the Critical Thinking & Logic collection for the full learning path