general research answers carried a fabricated or mismatched citation in testing.
A citation that says 42% when the paper says 24% passes peer review more often than anyone would like. The reference looks right; the number is not.
A reference is not evidence until it matches.
Merlise resolves each citation to its DOI and reads what the paper actually reports. When the figure in the manuscript does not match the source, the claim is flagged rather than trusted.
Peer reviewInside the editorial workflow you run.
Authors and editors upload the manuscript. References are checked against Crossref and the literature, with mismatches surfaced before review begins.
Resolve the references
Every citation is matched to its DOI through Crossref.
Read the source
The cited paper's reported figures are compared with the claim.
Flag the gaps
Mismatched statistics and unsupported claims are surfaced for the editor.
Read it claim by claim.
Every statement gets a calibrated reading, a verdict, and the record behind it. The ones that need a person rise to the top.
“We screen submissions before an editor spends an hour on them. The mismatched citations surface in minutes.”
Managing Editor · a Cambridge methods journal
See it on one of your own documents.
Bring a real scientific document. We will walk you through the ledger, claim by claim.