Citation Research and Impact Metrics

1. Citation counts are not a measure of quality as articles may be cited for both negative as well as positive reasons. Why something is being cited must always be considered in the assessment.
2. Citation behavior for subject areas varies depending on many disparate factors such as the preferred document type (books vs journal articles vs conference papers vs patents), authors and audience (practitioners vs researchers), and environment (industry vs academic). Consequently, raw citation counts cannot be used to compare authors in different subject areas, even for those subjects that may seem closely related.
3. Authors who have had long careers are most likely to have more publications with more citations than those at the beginning of their careers. Comparing senior faculty to junior faculty using raw counts penalizes junior faculty regardless of whether the counts are for the amount of publications or the amount of citations.
4. Although many indexing or abstracting services provide citation counts, each source only count what's in its own database. The subject focused databases are looking at most, if not all, of the journals within its subject areas while the interdisciplinary databases are trying to cover a greater variety. Both of them cover the same well-know journals, so you can't do a simple addition of scores across the services as there is extensive duplication in the counts. At the same time, each source may be finding citations the other source(s) haven't, so you can't just search one database and be done if you want to be comprehensive.
5. Searching by author name is challenging. Even if the author is consistent with exactly how their name is listed on each document, a database's policy may change the author's name (ex., using initials instead of spelling out the full first and middle names). Use the database's HELP file for author searching to learn the specific policies of the database. Make sure that you have sufficiently covered all possible variants of the authors name in each database you use. For common names, adding in the author's organization/institution to the search helps, however, there may be situations when two or more people with the same name are employed by the same organization; in some cases, manually going through a results list to eliminate unwanted documents may be necessary.
6. Do not compare counts from different sources. For example, do not take a citation count from the Scopus database for "Author A" and compare it to the citation count from PubMed for "Author B."

Best Uses:

• Comparing citation counts for individuals within the same, focused (i.e., narrow) subject area and at approximately the same career point.
• Determining if an individual is producing at an equivalent or better rate than in previous years.