ABSTRACT
The term ‘generalised linear model’ (GLM) was first introduced in a landmark paper by Nelder and Wedderburn (1972) in which a wide range of seemingly disparate problems of statistical modelling and inference were set in an elegant unifying framework of great power and flexibility. Generalised linear models include all the modelling techniques described in earlier chapters-that is, analysis of variance, analysis of covariance, multiple linear regression, and logistic regression-and open up the possibility of other models (e.g., Poisson regression) that we shall describe in this chapter. A comprehensive account of GLMs is given in McCullagh and Nelder (1989) and a more concise and less technical description in Dobson and Barnett (2008). In the next section, we review the main features of such models.
