The post hoc, passed hack fallacy

The institution in which I profess statistics is UCL. (Initially it was called University College London but now it is called initially. This has something to with branding, apparently.) My department is Statistical Science, the oldest such department in the world, and we are very proud of it. Indeed, it will not have escaped the attention of the attentive reader, that the typical founding father of statistics studied mathematics at Cambridge and then went on to lecture in statistics at UCL. This combination can easily be explained by observing that only Cambridge is good enough for statistics but statistics is not quite good enough for Cambridge.

Returning to UCL, like other departments of statistics throughout Britain and indeed like departments in other disciplines, we have to document our excellence every five years. This is an extremely embarrassing business in which normally shy and retiring dons with no desire whatsoever for self-aggrandisement are forced by cruel and bitter economic and political exigency to list all their publications and public lectures for the ‘Research Assessment Exercise’. Ah, cruel new world. Whatever happened to scholarship?

The statistics panel required us to submit our four best publications to be assessed. It is this that helped the assessors to determine whether we deserved the accolade of a five rating or the ignominy of a one.