LAUSR

The purpose of this study was to assess the gamma distribution as a model of the distribution of ages of cited references in corpora of scientific literature, and to derive inferences from the parameters of the distributions. The ages of cited references in 2867 articles published in 40 distinguished journals in the fields of accounting, economics, finance, management, marketing, operations and information systems, organisation behaviour and human resources were analysed. The distributions of ages of cited references in each subject area were fitted to gamma distributions with the parameters estimated using minimum distance estimation. In contrast to extant literature, it is shown for all subject areas in the study that the goodness-of-fit statistics for gamma distributions were superior to those for lognormal distributions. The rate of growth of knowledge and the temporal profile of the growth of knowledge are derived from the parameters of the gamma distributions and differentiate the subject areas. Longitudinal analysis demonstrates that the gamma distribution models are stable but illustrate the evolution of specific corpora of literature over time. The gamma distribution parameters and derived metrics can be applied diagnostically and descriptively to characterise corpora of literature, or prospectively to set norms, expectations and criteria for new research. The results have implications for future bibliometric studies, authors, editors, reviewers, and knowledge researchers. Opportunities for further research and verification of prior research are created from this novel bibliometric approach.