A Web of Science Citation Report is a summary view that aggregates citation data across an entire search result set. Where a single record on Web of Science shows only its own “times cited” count, the Citation Report tool sits above a set of results — a saved search, an author’s publication list, or any other result set — and rolls those individual counts up into a small number of set-level metrics: total publications, total times cited, citations per year, average citations per item, and the h-index for that specific set of records.
This guide covers what the Citation Report actually computes, how to generate one, what each metric on it means, and where it fits relative to the two other Web of Science workflows CASRAI already covers in depth — building the search that produces the result set and disambiguating an individual author’s records before reporting on them.
What the Citation Report tool actually is
Citation Report is a feature within Clarivate’s Web of Science platform, available from the results page after running any search. It is not a separate database or a different metric source — it recomputes standard citation statistics over whatever result set is currently active, using the same underlying Web of Science Core Collection citation data that powers individual record pages. Two constraints matter operationally:
- Result-set size limit. Web of Science generates a Citation Report only for result sets up to a fixed maximum number of records (historically 10,000). A search that returns more results than that ceiling needs to be narrowed — by date range, document type, or additional field tags — before a report can be generated.
- It reflects only what’s in the active result set. If the underlying search under- or over-includes records (a name-disambiguation miss, a missing variant spelling, an unintended document-type filter), the Citation Report metrics inherit that error silently. The report itself has no way to flag that its input set is wrong — the accuracy work happens upstream, at the search and disambiguation stage.
How to generate one
- Run a search on Web of Science that produces the result set you want metrics for — most commonly an author’s publications (via Author Search, a name+affiliation query, or a saved Author Record) or a topic/institution search.
- From the results page, select “Create Citation Report” (the exact link/button label and placement is set by Clarivate and can shift between interface versions — look for it near the results-list sort/export controls).
- Optionally exclude specific records before generating the report — self-citations aren’t automatically separated out, and if particular items (editorials, corrections, retracted records) shouldn’t count toward the aggregate, deselect them first.
- Web of Science computes the report from the resulting set. For an author’s full record set, this is often generated directly from a confirmed Author Record rather than a manual query — see CASRAI’s guide to Web of Science Author Search and disambiguation for how to get a clean, merged Author Record before pulling a report from it.
What each metric on the report means
Sum of the times cited
The total number of citing instances across every record in the set, counted only from other items indexed in Web of Science Core Collection. This is a raw sum, not deduplicated for co-authorship — if two papers in the set cite each other, that citation still counts toward the total (self-citation is not excluded by default).
Average citations per item
The sum of times cited divided by the number of records in the set. Because citation counts are typically right-skewed (a small number of highly-cited items pull the mean up), this average can be misleading for a set with even one outlier record — a sole highly-cited paper in an otherwise modestly-cited body of work will inflate the per-item average well above what most records in the set actually received.
Citations per year (the trend chart)
A bar chart of how many citing events, across the whole set, occurred in each calendar year. This is the metric best suited to showing a trajectory — whether attention to a body of work is growing, flat, or declining — rather than a single cumulative figure. It is commonly attached to promotion, tenure, and grant renewal narratives as evidence of sustained (not just historical) impact, though it reflects the citing behavior of Web of Science-indexed literature specifically, not scholarly attention as a whole.
h-index (for the set)
The report computes an h-index for the specific result set it was generated from — h publications in the set have each been cited at least h times. This is mechanically the same h-index calculation used elsewhere, but the number it returns is scoped to whatever set produced the report. An h-index computed from an incompletely disambiguated author search will differ from one computed after a proper Author Record merge, and both will generally differ from the h-index the same author’s work would produce on Scopus or Google Scholar, because each source indexes a different set of citing and cited literature.
Why the same person’s h-index differs by source
A recurring point of confusion in tenure and grant documentation is that a researcher’s h-index “changes” depending on where it was pulled. This isn’t a computational inconsistency — the formula is the same everywhere — it’s a coverage difference. Web of Science Core Collection, Scopus, and Google Scholar each index a different (overlapping but non-identical) set of source publications and citing literature, so the same underlying body of work produces a different citation count, and therefore a different h-index, on each platform. When citing a Citation Report h-index in a CV, dossier, or grant narrative, name the source and the date the report was pulled — an unlabeled h-index is not independently checkable.
Citation Report vs. Journal Citation Reports (JCR)
These two Clarivate products share almost the same name and are frequently confused. They measure entirely different things:
- Citation Report (this page) is generated on demand from any result set — an author’s papers, an institution’s output, a topic search — and reports metrics about that specific set of records.
- Journal Citation Reports (JCR) is a separate, annually-published Clarivate database reporting journal-level indicators, including the Journal Impact Factor, for indexed journals as a whole — not for any individual author or paper.
A Citation Report never produces a Journal Impact Factor, and JCR never produces an author- or paper-level h-index. If a task calls for “the impact factor of the journal,” that’s JCR; if it calls for “how many times this author’s work has been cited,” that’s Citation Report. For the JCR side specifically, see CASRAI’s guide to looking up a journal’s impact factor.
Common uses in research administration
- Tenure and promotion dossiers — a Citation Report exported for a candidate’s full publication record documents cumulative citation counts, the citations-per-year trend, and a Web of Science-sourced h-index as supporting evidence, typically alongside (not instead of) narrative peer assessment.
- Grant renewal and progress reports — some funders and internal review committees ask for citation trend evidence on prior outputs; the per-year chart is the direct source for that.
- Departmental or institutional research-output review — aggregating a Citation Report at the institution or department level (rather than per-author) gives a rollup of output volume and citation impact for a reporting period, though this is more often done through a CRIS or bibliometrics unit than a manual per-search export — see CASRAI’s persistent identifiers and research information systems hub for how that data typically flows into institutional systems.
Because these figures often carry real consequences, treat a Citation Report the way any standards-conscious research office should treat a metric: document the source (Web of Science Core Collection specifically, not “Web of Science” as a vague label), the exact search or Author Record it was generated from, and the pull date — the same discipline CASRAI’s Author Search guide recommends for disambiguated records generally.
Frequently asked questions
Does a Citation Report exclude self-citations automatically?
No. Self-citations are included in the default sum of times cited, average citations per item, and h-index unless specific records are manually excluded before generating the report.
Is there a limit to how many records a Citation Report can cover?
Yes — Web of Science applies a maximum result-set size for generating a Citation Report. A search that exceeds it needs to be narrowed with additional filters (date range, document type, refined field tags) before a report can be produced.
Can I generate a Citation Report for a whole institution, not just one author?
Yes, in principle — any result set can produce a report, including an institution-level search (using the OG= or affiliation field tags). In practice, institution-scale bibliometric reporting is more commonly handled through a CRIS or a dedicated bibliometrics team, since institution-level searches carry a higher risk of affiliation-matching noise than a well-disambiguated individual Author Record.
Why does my h-index look different on Web of Science than on Google Scholar?
Because each platform indexes a different set of source and citing publications, not because the h-index formula differs. See “Why the same person’s h-index differs by source” above.
Related CASRAI resources
- How to Search Web of Science Effectively — building the underlying search and using field tags, Boolean/proximity operators, and Cited Reference Search.
- Web of Science Author Search: Finding and Disambiguating an Author’s Records — producing a clean, merged Author Record before pulling a Citation Report from it.
- h-index — the underlying metric definition.
- Journal Citation Reports (JCR) and How to Look Up a Journal’s Impact Factor — the separate, journal-level Clarivate product this is often confused with.
- Scopus — the comparable citation database from Elsevier, for cross-checking a set-level metric against a different index.
- Persistent Identifiers & Research Information Systems — the CASRAI cluster hub for this topic.
Sources
- Clarivate, Web of Science platform documentation and Core Collection product pages (clarivate.com/academia-government/scientific-and-academic-research/) — Citation Report feature, result-set size limits, and self-citation handling.
- Clarivate, Journal Citation Reports product page (clarivate.com/academia-government/scientific-and-academic-research/research-funding-analytics/journal-citation-reports/) — for the Citation Report vs. JCR distinction.
