Overdispersion

Viac o knihe

In the frequentist framework, a quasi-likelihood approach has long been used as a means of allowing for overdispersion, a common occurrence in many areas of scientific research. An alternative approach is to improve models by including random effects that account for the extra variability. This book describes both approaches in detail and compares them in terms of the numerical differences for each data set, using several examples involving real data. Additionally, their performance is assessed by simulations based on examples.