Why it matters when you are writing
The models differ in who knows whom. Single-blind hides reviewer identities from authors; double-blind hides both, which reduces some bias and is imperfect in small fields where a manuscript’s topic identifies its author. Open review publishes the reports, sometimes with names. Registered reports invert the sequence entirely: reviewers assess the design before data collection and accept in principle, which removes the incentive to find a significant result.
The verdict you will see most is major revision, and it is good news dressed as criticism. It means the editor considers the work publishable if specific problems are addressed. The response letter matters as much as the revision: answer every point in order, say exactly what you changed and where, and where you disagree, say so with reasons rather than quietly ignoring the comment.
Be clear-eyed about what review does not do. Reviewers rarely see raw data, almost never rerun analyses, and cannot detect a fabricated dataset that is internally consistent. They also work unpaid, under time pressure, in a system where an unfashionable finding faces more friction than a fashionable one. Peer review is a meaningful filter and a poor guarantee, which is why replication matters more than publication status.
