Sir

As rightly pointed out by David Colquhoun in Correspondence —“Challenging the tyranny of impact factors” (Nature 423, 479; 200310.1038/423479a) — the citation rate of the individual paper is essentially uncorrelated to the impact factor of the journal in which it was published.

Adding insult to injury, the impact factors of many journals also change over time. To quantify this further, I selected journals whose impact factor is available from 1992 to 2001 from Roman Woelfel's website (http://staff-www.uni-marburg.de/~woelfel/impact.html). When I compared data for more than 3,000 journals, I found that 26.8% of them had at least doubled their impact factors over this time period, whereas 1.8% had decreased by up to half.

A few journals (1.9%) had increased their impact factor more than tenfold by 2001, although most of these had very low impact factors to start with. Among the most notable increases from 1992 to 2001: Behavioral and Brain Sciences from 0.30 to 17.31; CA: A Cancer Journal for Clinicians from 5.02 to 35.93; Current Opinion in Immunology from 2.16 to 13.72; The Journal of the American Medical Association from 5.56 to 17.57.

Notable decreases included the Federation of American Societies for Experimental Biology's FASEB Journal from 18.21 to 8.82 and Quarterly Reviews of Biophysics from 13.0 to 3.94.

When scientists are being evaluated on the basis of the impact factors of the journals in which they publish, such distortions should be kept in mind.