Contact Us!

Bibliometrics Email

library.bibliometrics@noaa.gov

Sarah Davis - Senior Bibliometrics Librarian

sarah.davis@noaa.gov

Jan Thomas

jan.thomas@noaa.gov

Hope Shinn

hope.shinn@noaa.gov

Keith Locantore

keith.locantore@noaa.gov

What are Bibliometrics?

Bibliometrics are the quantitative analysis of academic publications. Using academic publications as a data source, bibliometric analysis attempts to provide a better understanding of how research is produced, organized, and interrelated. It also attempts to evaluate academic publications and sets of publications based on the number of citations these publications have received. Bibliometrics and citation analysis is one way we are able to illustrate NOAA's status as producer of world-class research.

Why are Bibliometrics Important?

• Bibliometrics can help illustrate the impact of a scholarly publication or group of publications in the greater research community and can support application for grants and research funding.
• When used with other methods such as peer review, bibliometrics is a useful tool in evaluating the research output of programs and researchers.
• Bibliometrics can be used to inform decisions about future research or identify research strengths and gaps.

Limitations

• Bibliometric indicators are imperfect and can only be used for their specific purpose. For example: Journal Impact numbers can only be used to evaluate journals and not the papers in those journals.
• Citation counts and behaviors vary among fields and over time and comparisons between subject areas should be avoided. Additionally, articles need at least two years to accumulate enough citations to provide meaningful analysis.
• Most bibliometric datasets have skewed distributions and as such an average is not necessarily representative of the whole. Providing median or other metrics in addition to mean helps resolve this problem.
• Many metrics are easy to manipulate such as increasing H-Index through self-citation and these metrics should be used with care and a full awareness of the potential for manipulation.
• Bad data makes for bad bibliometrics; to produce a quality analysis it is important to start with a clean dataset of high quality and adequate size. Using data from a reliable source such as Web of Science makes this much easier.
• There’s no way to definitively know what indicators actually measure. For instance, high citation rates don’t necessarily correlate to credibility and an article may be highly cited because the research it presents is being questioned.