MAS473   Extended Linear Models   (10 credits)

Description

This course considers two recent developments in statistical modelling which together dramatically extend the scope of analysis, enabling it to cope effectively with problems from agricultural experimentation, industrial planning, environmental and medical research. The first extension is to models for situations in which variation arises from several sources ¿ from different life-style choices, for example, in relation to patients¿ responses to medical treatment, or from variations in field fertility as well as local micro-climate in the growth of crops. The second extension is to problems in which data are naturally modelled by distributions other than the Normal: count data or non-negative data for example. The unit will show how standard Normal theory regression methods can be profoundly generalised to deal with the relationship of response variables to explanatory factors. Applications include drug cure-rates as functions of dose, the analysis of contingency tables and the modelling of sports data.

Reading List