Statistical measures, such as journal impact factors, are frequently controversial. This series has explained what impact factors are, how they are calculated, and how they are used. This final article will address some of the criticisms and limitations of impact factors.

A major criticism of impact factors is that, over time, they have been used to determine "author impact". The impact factor was created to compare journals, not authors, but some tenure committees evaluate the quality an individual's research and publications by looking at the impact factors of the journals in which they have been published.

English language journals dominate the international research and clinical literature and editors of foreign language journals are critical of impact evaluations. Local clinical journals are less relevant for most researchers and are cited less frequently.

Impact factors are also skewed toward journals that mainly publish review articles, since those tend to be cited more frequently.

In addition, a typical paper in a journal with a high impact factor may not, in fact, be cited more frequently than the average paper in a lower-ranking journal.

The time required to review manuscripts can have some effect on the accuracy of impact factor data. If the publication and review of articles is delayed, and references are no longer current, they will not be included in the JCR impact calculation.

Many citation analysts believe that one of the major problems with impact factor data is the frequency of errors made when citing articles. Errors in cited references translate into misplaced credit. Citation errors increase as they are copied from paper to paper.

ISI, the company that creates and publishes impact factors, has cautioned against the misinterpretation and misuse of this data. According to David Carter, Vice-President for Corporate Communications, "These scores were never designed by ISI to be proxies for the influence of papers, or when aggregated, the work of individuals."

Impact Factors are not perfect. Ideally, they should be used in conjunction with other measures to evaluate the quality and influence of a journal.

• Previous Article in the Series
• Citation Analysis: The Counting House. Nature. February 14, 2002