ABSTRACT
This chapter discusses the feasibility of investigating cause–effect relations, the traditional basis of inferential statistics. We also specifically advocate the generalized (mixed effects) regression framework, including linear mixed effects regression models (mixed models), multinomial logistic regression and generalized additive (mixed) models (GAM[M]s), in which the time series data, treated as single-cluster data, become the target of inferential statistics. All of these represent important statistical developments of the last 45 years and provide an extremely useful set of tools for analyzing data within the field of applied linguistics. Our goal is to focus on what these models are, how they work and why/when applied linguists should use them. We close with a critical discussion of limitations of inferential statistics and desiderata for future research, referring to Chapter 8, “Advancing quantitative research methods”.
