Statistische Modellierung

Abstract

This paper provides an overview of central aspects of statistical modeling of linguistic data. Starting from a general definition of model, the paper discusses the goals of modeling as well as a variety of issues bearing upon the formulation/definition of statistical models. It then surveys model selection, the choice of 'the best model' and three fundamental notions affecting the interpretation of models. Finally, the paper turns to validation and replicability and addresses a variety of challenges researchers face during modeling processes.