Publication metrics are a measure of academic productivity. They are a count of the number of scholarly outputs by academic authors, and other metrics derived from this count.
Citation metrics quantify the use of scholarly publications by citing academics as a count of the number of citations and other metrics derived from this count.
Both publications and citation metrics provide an indication of academic influence because they only count scholarly outputs and the citations from scholarly outputs.
Publication and citation metrics are best used in combination, e.g. Citation Count and Citations per Output are complementary to Field-Weighted Citation Impact:
| Metric | Size-normalised | Field-normalised | Publication-type normalised | Resistant to database coverage | Difficult to manipulate | Time-independent |
|---|---|---|---|---|---|---|
| Outputs in Top Percentiles | Percentage | |||||
| Field-Weighted Citation Impact | Yes | Yes | Yes | Yes | Yes | |
| Citations per Output | Yes | |||||
| Publications in Top Journal Percentiles | Percentage | Yes | Yes | Yes |
When presenting publication and citation metrics, it is important to state the time frame that metrics refer to. For example, when benchmarking individual researchers, the year of first publication must be stated. It is also important to state the database source.
Counts of Scholarly Outputs will vary between sources, and there is no comprehensive database covering all outputs. Comparisons of the number of Scholarly Outputs should, therefore, only be made from the same source.
The number of citations recorded for a publication depends on the source database. This is because only the citations from articles indexed in the same database will be recorded. Because the publications (e.g. journals, books, book chapters, conference proceedings) indexed in a database will vary, the number of citations recorded for each output will also vary between databases. Therefore, a single source should be used when benchmarking research entities with citation metrics.
Citation metrics for the same publication(s) will vary between database sources. This is because only the outputs indexed in a database will be included in a citation calculation from that data source, and citation data sources will vary in their coverage of outputs. No data source will have comprehensive coverage of all outputs. Comparisons should only be made from the same source.