How to compile a publication report for your lab or department

1. Decide what the report covers — and write it down

Most disagreements about a publication report are really disagreements about scope. Settle these first:

• Who. Faculty only, or also trainees and staff scientists? What about people who joined or left during the period — do their papers count for the whole period, or only while they were members?
• When. Calendar, academic or fiscal year? And by which date: a paper can appear online months before its issue. Pick one — publication date or online-first date — and apply it to every paper.
• What. Research articles and reviews, certainly. Editorials, letters, errata, preprints, book chapters? Decide whether they are in, out, or reported separately.

Put these rules in a short methods note at the end of the report. Whoever compiles next year’s report will need them, and readers comparing years will too.

2. Build the roster

For each person, record everything a search needs:

• Name as it appears on their papers, plus variants (initials, a previous surname, a hyphenated form).
• ORCID iD, if they have one.
• Their affiliations over the reporting period, not only the current one.
• The division, program or center they report under, and their start and end dates.

3. Find every paper

Search PubMed for each person, using the method in how to find all of a researcher’s publications: author and full-name searches, ORCID iD, then affiliation and date to narrow common names. PubMed covers the biomedical literature well, so for most health-science departments it is the backbone. Ask faculty for anything outside it — conference proceedings, books, work in non-indexed journals.

Emailing faculty for their citations is still the most common method, and the least reliable: replies arrive late, in a dozen formats, and people forget papers. Use it to catch what the search missed, not as the source.

4. Remove duplicates, but keep the attribution

A paper co-authored by three members appears in three searches. In the report’s totals it counts once; in each member’s or each division’s section it can appear under every author. Match duplicates on PMID or DOI rather than title, which varies between sources.

5. Have faculty confirm their lists

Send each person their list with a clear deadline and one question: is anything here not yours, and is anything missing? This catches the same-name mistakes a search can’t, and it means the finished report is one the department has agreed to — which matters when it goes to a dean or an external reviewer.

6. Organize it the way the report is read

Group papers by the units the report is about — division, program, center or research theme — and within each, sort by date or by author. If the report highlights trainee authorship or collaboration between divisions, decide how you will mark it before you start formatting.

7. Add grants, if the report asks for them

For NIH funding, NIH RePORTER lists every award with its project number, dates and amounts; search by investigator name. Other funders usually have to come from your grants office or from faculty.

8. Use metrics carefully

• Counts are the least contested number: papers per year, per division, per member.
• Citation counts depend on the database — Web of Science, Scopus, Google Scholar and PubMed all give different numbers. Name the source, and don’t mix sources within one report.
• Journal impact factors describe journals, not the papers in them. The San Francisco Declaration on Research Assessment (DORA), signed by many institutions and funders, recommends against using them to judge individual papers or researchers. If your report shows them, present them as context rather than a score.

9. Format consistently and keep a snapshot

Use one citation style throughout — see NIH citation format if the report feeds an NIH application — and export from one place rather than pasting from several. Finally, keep a copy of the final list with the date you compiled it: next year starts from it, and questions about this year’s numbers will come back to it.