At irregular intervals British Universities subject themselves to a particularly perverse form of navel gazing called the Research Assessment Exercise (RAE). The exercise is supposed to decide who is doing the best research in their discipline, and to decide how future research money from the government is to be divided up. In the current exercise there are 67 so-called panels that look at specific subject areas. Most academics such as (say) chemists or historians will have a single panel devoted to their speciality. Statisticians are lumped with "operational research", but will be pleased to be distinguished from applied mathematics in this assessment. However, many statisticians work in areas other than statistics and they have a problem with the RAE. Their best applied papers would not be rated highly by the statistics panel, because there are no new statistical techniques on display. Most jobbing statisticians will be too busy analysing data to have the time or the inclination to invent new techniques, especially if an old technique, judiciously applied, will give a "good enough" answer. However, if they did achieve a paper in, say, Biometrika, an applied panel would probably not look upon it as a big advance in their area.
展开▼
