ABSTRACT
The likelihood principle plays a prominent role in statistical methods research for developing powerful statistical inference tools. The likelihood method is arguably the most important concept for inference in parametric modeling when the underlying data are subject to various assumed stochastic mechanisms and restrictions related to biomedical and epidemiological studies. This chapter provides strong arguments to position the likelihood function as an essential and optimal tool in biostatistical inference. Chapter 3 briefly introduces principles and aspects of the maximum likelihood estimation and its applications, likelihood ratio based procedures and their optimality, maximum likelihood ratio tests and their properties as well as examples related to correct model based likelihood formations.
