فرآیند رقابت دو مرحله ای منجر به توزیع درآمد قدرت شبه قانون : کاربرد انتشار علمی و توزیع استناد

We apply a two-step competition process as a model to explain the distribution of citations (‘income’) over publications (‘work’). The first step is the competition amongst scientists to get their work published in better journals, and the second to get this work cited in these journals. Generally, citation distributions are supposed to follow a power law, like most other ‘income’ distributions. So far, no satisfactory theoretical model of citation distribution has been developed. On the basis of two Boltzmann type distribution functions of source publications, we derive a distribution function of citing publications over source publications. This distribution function corresponds very well to the empirical data. It is not a power law, but a modified Bessel-function. In our view, the model presented in this article has a more generic value, particularly in economics to explain observed income distributions. We developed a new model to explain the distribution of citations over publications. Bibliometric measurements of the distribution of citations over publications suggest a power-law function (see, for instance, [1] and [2]). But so far, no satisfactory theoretical model of citation distribution has been developed. Our model consists of two steps. First, the competition amongst scientists for ‘publication status’. This status is determined by the way the journal is cited by other journals. We argue that the underlying distribution originates from the journal in which a publication appears and it is operationalized in the form of an equilibrium distribution of publications according to their ‘status’. Second, within their status level, scientists again have to compete with their publications (i.e., with their ‘work’), in terms of getting cited (‘income’). On the basis of these two distribution laws, a third one results, the distribution of citations (i.e., citing publications) over source publications. A more detailed discussion and comparison of the model with further empirical findings based on bibliometric measurements is given elsewhere [3]. The basic concept of our model is the idea that scientific communication is characterized by a large number of publications that has to be divided according to attributed status. This concept is based on the following assumptions: