The use of a group of statistical techniques to model the way in which people choose between distinct alternatives, such as a university course. The basic concept used is that each alternative has a total utility to the decision-maker, which is the combination of the weighted utilities of all the attributes of the course; for example, the quality of the university teachers, the course content, the entry requirements, distance from home, and local living costs.

 It is then possible to calculate the possibility, P, of choosing one out of j alternatives on the basis of the equation:

where i is the rank, by utility, of the alternative and v_i its utility, but the amount of data and calculation entailed is enormous.