How publishing works in this field
Bibliometric papers are judged on reproducibility of the query. The database, the version or snapshot date, the exact search string, the document types included, and the deduplication and disambiguation rules all have to be reported, because the same question run on Scopus, Web of Science and OpenAlex returns different answers.
The field has a strong open-access tradition and a correspondingly low tolerance for closed practice. Depositing data and code is close to expected, and so is acknowledging the limits of whichever database was used rather than treating it as the record of scholarship.
What slows papers down here
Author disambiguation quietly breaks results. Common names, inconsistent affiliations and missing ORCID coverage distort author-level analyses, and reviewers ask what was done about it. "We used the database's author IDs" is an answer only if their error behaviour is discussed.
Coverage bias is the other standing objection. Each database under-represents non-English publication, books, and some regions and disciplines, and a paper that reports a global finding from one database without that caveat will be challenged.
Worth reading next
Questions we are asked in this field
The database and its snapshot date, the complete search string, document types and language filters, the deduplication approach, the disambiguation method, and the export date. Someone else should be able to re-run it and get the same set.
Whichever fits the question, with its limits stated. Scopus and Web of Science differ in coverage, OpenAlex is broader and noisier, and all three under-cover books and non-English work. Using more than one, and reporting where they disagree, strengthens a paper considerably.
State the method — ORCID where available, affiliation and subject heuristics otherwise — and report how it was validated, for instance by manual checking of a sample. Reviewers in this field know how badly naive matching performs.