Andrea Scolari

The likelihood methods are used to find efficient estimators, and to construct tests with good power properties. One such tool is the likelihood ratio (LR) statistic which is used to assess the goodness-of-fit of two statistical
models. A problem with parametric likelihood inferences is that we might not know which parametric family to use and misspecification can cause likelihood-based estimates to be inefficient. We will discuss the empirical
likelihood (EL) method which is a nonparametric counterpart of the likelihhood method. We will examine the necessary background pieces of EL, compare with the parametric likelihood and provide an example of EL
method for univariate mean.