How a referee writes a report is interesting for two main reasons. The general one is that it is an example of how humans perform tasks, and in this it is similar to maintaining correspondence, whether done by letters, email, or messaging systems. Here recently debates have raged about whether such activity leads - perhaps intrinsically - to power-law statistics and how that would follow from priority queue -type models. The second, not necessarily less important one, is that it gives us insight into how science, journals, and peer review works. Here we discuss empirical data from two physics journals, and what can be distilled from the data. We present a simple "many tasks" model, which fits the data fairly well. Conclusions are presented about both the issues, and in particular about refereeing as seen in the light of both the data and models.
Go to day