Direct comparison
Single-Blind vs Double-Blind vs Open Peer Review
Single-blind, double-blind (anonymized), and open peer review compared: what each hides, who sees what, and how fields and journals differ.
Side-by-side comparison
| Dimension | Single-anonymized | Double-anonymized | Open review |
|---|---|---|---|
| Also known as | Single-blind (older term) | Double-blind (older term) | Open review, transparent review |
| Reviewer identity visible to author? | No | No | Usually yes; often published alongside the article |
| Author identity visible to reviewer? | Yes | No, in principle; though often guessable from preprints, self-citation, or writing style | Yes |
| Review reports published? | No, by default | No, by default | Often; decision letters, referee reports, or full correspondence, depending on the journal |
| Main rationale | Reviewers can criticize freely without fear of reprisal from the authors | Judges the work on its merits, reducing bias from an author's name, institution, gender, or nationality | Makes reviewers accountable for their comments and gives reviewers citable credit |
| Main criticism | No reciprocal accountability for the reviewer; author reputation can still bias review | Author identity is often guessable in practice; doesn't fully solve bias | Reviewers, especially early-career ones, may soften criticism or decline to review at all |
| Where it's most common | Broad default across STEM and biomedical publishing | Common in social sciences and humanities; author-selectable option at some STEM journals | Standard at post-publication and preprint-linked publishers; opt-in at others |
| Real-world examples | Default at most Elsevier, Springer Nature, and Wiley biomedical/STEM titles | Author-selectable at Nature and the Nature research journals, since 2015; RSC's Chemical Science; default at many humanities/social-science journals | F1000Research (fully open, post-publication); PLOS Published Peer Review History (opt-in); BMJ prepublication history; Nature Communications opt-in published reports, since 2015; ICLR/NeurIPS/ICML (double-blind during review, made public after decisions via OpenReview) |
| Current preferred term (STM 2020 taxonomy) | Single-anonymized | Double-anonymized | Open peer review; already unambiguous, no revision needed |
Common questions
FAQ
Which peer review type is most rigorous?+
No model has been consistently shown to produce more rigorous reviews; evidence on quality differences is mixed. The real trade-off is which failure mode a journal is more worried about: reviewer timidity (addressed by anonymizing reviewers) or author bias (addressed by anonymizing authors).
Why do some sources say 'double-blind' and others say 'double-anonymized'?+
Both describe the identical practice. A 2020 STM Association peer review taxonomy working group recommended replacing 'blind'/'blinded' with 'anonymized' because the older terms are ableist in origin. Many journals and search results still use 'blind'; expect to see both in the literature.
Can authors choose their peer review model?+
At some journals, yes; Nature and the Nature research journals let authors opt into double-blind review at submission, and several journals let authors choose whether to make their peer review history public. Most journals apply one model uniformly and don't offer a choice.
Does open peer review mean anyone can review the paper?+
No. 'Open' here describes transparency around an already-invited, formal review process (published reports, disclosed identities), not open or crowdsourced participation. Fully open community review is a related but distinct model.
Is there such a thing as triple-blind review?+
Some journals use the term for a model where the handling editor is also kept unaware of author identity, in addition to author-reviewer anonymization. It is uncommon compared to single- and double-anonymized review.
Going deeper
