SUCCESSFUL 2008 PREDICTIONS
The Power of Citations:
Strong Signals of Excellence in Science
The Nobel Prize in Chemistry for 2008 honors three researchers who developed fluorescent protein probes for studying cellular function: Osamu Shimomura, Martin Chalfie, and Roger Y. Tsien. And the 2008 Prize for Economics recognizes the work of Paul Krugman for his theories of international trade and analysis of geography in economic transactions.
The Scientific business of Thomson Reuters predicted in September that Tsien was a contender for this ultimate accolade in science. In 2006, Krugman was picked as one of our Citation Laureates who were in the running for the Nobel Prize. Since 1989, predictions of who may win the Nobel Prize, based on citations in the scientific literature and other factors (such as the receipt of top awards), have time and again turned up Nobelists to be. In fact, since 1989, Thomson Reuters has correctly predicted at least one Nobel Laureate each year, with the exception of the years 1993 and 1996. In some years, namely, 2002, 2003, 2005, and 2008, these Citation Laureates won in the same year they were named by Thomson Reuters. In 2007, two Nobel Prizes were accurately predicted, in Physiology or Medicine and in Physics, based on nominees posted in 2006.
What accounts for the predictive power of citations? It is not really that mysterious. Citations in the literature represent repayments of intellectual debts that one researcher pays to another. The social norms and formal publication procedures in science demand that relevant work is acknowledged — and it is considered misconduct not to recognize previous work which a researcher has used in his or her own work. Moreover, these citation links represent a coherent pattern of intellectual connections. In other words, there is much meaning in all those millions of footnotes appended to research reports. The counting of citations, over specific periods and in different fields, can reveal the most highly cited people and papers, and these ranked lists have been shown to contain a large number of scientists who have already won the Nobel Prize (and other prestigious awards) and the names of those who later go on to receive the Nobel Prize.
The intellectual coherence of citations permits their use in searching the literature. The first citation index for science, created in 1963 by information pioneer Eugene Garfield, was designed for information retrieval and represented a powerful new way for searching the literature. The principle behind a citation index: If one knew of a work relevant to one's research, a search of the index could reveal what other publications had cited that work, and these citing publications likely dealt with the same subject. Thus, one could search forward as well as backward in time. Researchers worldwide today depend on ISI Web of Knowledge from Thomson Reuters, as it is the only citation index which provides such in-depth current and retrospective coverage in the sciences, social sciences, arts, and humanities.
Early on it was recognized that this vast, rich resource offered other possibilities, namely, insight on the social structure and history of science. It has also been exploited for quantitative studies of achievement in scientific research (see Henk F. Moed, Citation Analysis in Research Evaluation, Springer, 2005). The efforts by Thomson Reuters to predict Nobel Prize winners to be is merely one extension of this use.
--David Pendlebury, Citation Analyst, Scientific business of Thomson Reuters