Patterns Behind The Rankings: Mining Academic League Tables To Reveal Variability And Covariation

University League Tables are widely used to compare the standing of individual departments or universities, both nationally and internationally. In response to criticisms, the League Tables have become more sophisticated and use more resources to ensure better measurement, although difficulties still abound. Rather than being limited to uni-dimensional measures of standing, it is argued in this paper that the data from League Tables can also be used to investigate the range of characteristics of universities and the covariation of these characteristics. Presenting factor analyses based on many of the available League Tables shows that quite complex patterns are often yielded although some traditional distinctions such as research versus teaching dimensions sometimes occur.