We show universal behaviour for two indicators of the quality of publications taken from two different data sets, papers from a single institution and those on arXiv. We demonstrate this universality for different years and subjects. This distribution is well fitted by a lognormal with a variance of around 1.3, consistent with Radicchi et al (2008). We will also discuss the evolution over time of our measures describing the data and note that simple models do not have the correct temporal behaviour for our parameters. Based on arXiv:1110.3271 with additional new material.
Poster given at ECCS 2012