|
Article Excerpt
ABSTRACT
THE MATTHEW EFFECT FOR COUNTRIES (MEC) consists of the systematic deviation in the number of actual (observed) citations from the number of expected citations: A few countries, expecting a high impact (i.e., a high number of cites per paper) receive a surplus of citations, while the majority of countries, expecting a lower impact, lose citations.
The MEC is characterized by numerous facets, but two are the most impressive. The first is the possibility of ranking the science nations by their overall efficiency of scientific performance, thus making the MEC attractive for science policy. The second is the concentration of the MEC in a small number of scientific journals which happen to be the most competitive markets for scientific papers and, therefore, are of interest to librarians as well as scientists.
First, by using an appropriate measure for the above-mentioned deviation of the observed from the expected citation rate one can bring the countries under investigation into a rank order, which is almost stable over time and independent of the main scientific fields and the size (i.e., publication output) of the participating countries. Metaphorically speaking, this country rank distribution shows the extent to which a country is using its scientific talents. This is the first facet of the MEC.
The second facet appears when one studies the mechanism (i.e., microstructure) of the MEC. Every journal contributes to the MEC. The "atoms" of the MEC are redistributed citations, whose number turns out to be a new and sensitive indicator for any scientific journal. Bringing the journals into a rank order according to this indicator, one finds that only 144 journals out of 2,712 contain half of all redistributed citations, and thus account for half of the MEC. We give a list of these "Matthew core journals" (MCJ) together with a new typology relating the new indicator to the well-known ones, such as publication or citation numbers. It is our hypothesis that the MCJ are forums of the fiercest competition in science--the "Olympic games in science" proceed in this highest class of scientific journals.
INTRODUCTION
The Discovery of the Matthew Effect for Countries
It is often regretted that research papers, especially in the natural sciences, follow the stereotyped approach "introduction-method-results-conclusions," while the circumstances under which the authors achieved their results remain hidden. In contrast, this paper starts with a historical survey of the research lines we have followed since 1990. Impatient readers may skip this introductory section.
The effect was detected in 1994. A eureka moment of the kind known from discoveries in the natural sciences encouraged us to call what we could see on the computer screen "Matthew effect"--later, more precisely, "Matthew effect for countries" (MEC) (Bonitz, Bruckner, & Scharnhorst, 1995a). This event was no accident, it was preceded by long years of investigations into the structure of national science systems (Bonitz, Bruckner, & Scharnhorst, 1991, 1992, 1993). For instance, a hypothesis of the existence of two worlds in science--a "Right World" and a "Left World"--was a forerunner of the MEC (Bonitz, Bruckner, & Scharnhorst, 1995a, 1995b, 1996a, 1996b). Furthermore, if we hadn't had in the backs of our minds that there was a "Matthew effect in science," as introduced by the eminent scholar R. K. Merton into the sociology of science (Merton, 1968), we never would have dared name our phenomenon "Matthew effect."
In the first phase of our investigations, we studied the effect's time-stability, field-dependency, and its order of magnitude. The effect turned out to be stable over time, independent of scientific fields, and to have a small order of magnitude (Bonitz, Bruckner, & Scharnhorst, 1997). It is not an artifact. At this time, any speculations concerning the practical impact of our findings were beyond the scope of our considerations. Then, a measure for the effect was developed--"Matthew-Index"--the value of which must be computed for each country (Bonitz, Bruckner, & Scharnhorst, 1999a). Countries can be ranked according to this measure, and one can easily see how a certain country is affected by the MEC. We found our country ranking method more expressive than a relational charts representation developed by other researchers (Braun, Glanzel, & Schubert, 1989).
At the beginning, we were taking for granted the public understanding of the Gospel parable described in St. Matthew 25:14-30, and of Merton's Matthew effect in science. We declared, that the Right World (citation rich) countries were "taking away" citations from the Left World (citation poor) countries (Bonitz, Bruckner, & Scharnhorst, 1995a). It was some time before we grasped the very essence of the famous Biblical parable and then found it surprisingly compatible with our findings (Bonitz, 1997). This fact helped us in treating the "meaning" of the country rank distribution. It enabled us to understand whether it reflects a "discrimination against certain countries," the "quality of national science systems," the "usage of scientific talents," or the "efficiency of competition in science."
Next, we looked for the mechanisms underlying the country rank distribution for the "microstructure of the MEC." It became clear that we had to shift from the "country side" of the MEC to its "[journal side." All of the nearly 3,000journals in the database were expected to (more or less) contribute to the MEC. It seemed evident that the journals are like molecules--molecules that combine to produce the entity we call the MEC. We could not, however, presuppose that these molecules themselves are composed of different sorts of "atoms" (i.e., citations given to the journals), and that only one of these sorts is responsible for producing the MEC. This special sort, which had never before been described in journal investigations, we called "redistributed citations" or "Matthew citations." Our investigations into the behaviour and the properties of the Matthew citations yielded surprising results. The Matthew citations have a very skewed distribution over all journals: Most of the Matthew citations are concentrated in few journals, with 144journals containing half of the Matthew citations.
These lucky journals also play a highly distinguished role in scientific communication. We proved this hypothesis by trying to falsify it. However, journal ranking by number of publications, by number of citations, by number of participating countries, and by impact factor, failed to yield journal rank distributions highly correlated with the distribution by Matthew citations. Thus, a special role of the "Matthew core journals" (MCJ), as we called them, seemed to be established.
How should this new type of scientific journal be characterized? The journals with the highest reputation? The most important journals in science? The journals with the highest quality articles? If we would choose one of these features we could easily run into boundary problems by having to distinguish between journals of high and low reputation, between those of high and low importance, between high and low quality papers. The appearance of the Matthew citations offered a new possibility: Our proposal is that the number of Matthew citations in a journal reflects the degree of scientific competition going on within its pages. Those competing are scientists, scientific institutions, and countries in science. MCJ are thus the most competitive markets in science. They mirror the experience of other competitive areas, such as the economy or sports, which supports our metaphor that "the Olympic games in science" proceed in the highest class of scientific journals--in the MCJ. Another metaphorical filiation appears, reaching from the MEC to the treatment of the Parable of Talents in St. Matthew: Not "the rich are becoming richer and the poor poorer," but those who are most effectively competing, irrespective of the amount...
|
