“…Following the more general theoretical literature on testing and evaluation, such as Gill and Sgroi (, ), a Bayesian model in this context is simply one that produces a posterior probability, p i , that a given submitted paper indexed by i is of type a rather than type b . An external rule could then be used to further divide papers into categories: for example if we wanted four categories as indicated by the REF on page 10 of the consultation document we could make use of four ordered threshold probabilities { p r }, where r ∈ {1, 2, 3, 4}.…”