Impact Factor, Impact, and Smoke and Mirrors

摘要: I believe that the editorial by Martin Tobin, “Thirty years of impact factor and Journal,” which appeared in August 15, 2004 issue AJRCCM, merits comment (1). was particularly struck figure showing a dramatic upward trajectory recent years. This, to me, bore an ominous resemblance increase NASDAQ stock index late 1990s, before its collapse 2000. We, as physician-scientists, are trained notice unusual investigate further. As author noted, for journal year X is ratio number citations (in any journal) articles interest) X-1 X-2. Thus, can be influenced either decreasing published or increasing citations. There evidence both these case AJRCCM (using Thorax Chest reference points). Any directly influence annually. From 1999 2003 remained stable (196 198) did (701 704). Over same period time, decreased from 600 380. The publications through 2000 when 757 articles, making about 50% decrease. Given explosion research respiratory disease years, it seems ironic have been cut half. Viewed another way, ATS members now paying twice much per article, which, my mind, decreases “impact” Journal. other way articles. Although one would think this primarily result publication additional respected manuscripts reviews, also possible with process known “self-citing,” occurring at higher rate than journals. For example, 3,948 2002 1,122 (30.2%) were “self-cites.” Corresponding values 12.1% 11.3%, respectively. This appears several reasons. First, series “Year Review Blue exclusively cites artificially increases factor. 6 edition cited 380 total 576 times, will (2–7). Second, authors invited resubmit receive following statement: “In revising your manuscript, please update references ensure you discuss work context most area. In particular, need compare against Year accessible File-Cabinet icon on Homepage AJRCCM.” apparently intended citation work. net seen Figure 1; 2003, 44% referenced “self-citations,” compared 19% article references. Other strange phenomena occur. fourth leading citer French Revue des Maladies Respiratoires (with European Respiratory Journal three five). 32% (341/1055), next highest percentages being Journal, 5%. 1. Distribution pattern (A ) (B Thorax. solid part bar represents “self citations,” comprised, 2002, (1,122/2523) (91/491) (9).