Applicability of selected probability distributions to the number of authors per article in theoretical population genetics

Abstract

Recently scientists have investigated what statistical distributions can be used to describe the distribution of the number of authors per article.Ajiferuke has undertaken the most comprehensive study of this problem. He has found that by and large the Inverse Gaussian-Poisson distribution could describe most properly the observed authorship distributions. However, it is well known that this distribution is rather intricate, soRousseau tried to fit some simple one-parameter distributions to the number of authors of LIS articles. He has found that the geometric and the truncated Poisson distribution adequately describe these authorship data sets. The main purpose of the present paper is to continue these investigations and to analyse and test the viability of simple statistical distributions. As to (sub)fields where the single author dominates the results ofRousseau were corroborated: the truncated Poisson and the geometric distribution give often adequate fits to describe the number of authors. The Lotka distribution should be rejected. The truncated binomial distribution and the truncated negative binomial were investigated as well. However, it is not clear whether they are acceptable candidates.