This is a preview. Log in through your library . Abstract We consider the problem of fitting a generalized linear model to overdispersed data, focussing on a quasilikelihood approach in which the ...