Multilevel Regression Models for Mean and (Co)variance: with Applications in Nursing Research

Abstract

In this chapter, a concise overview is provided for the statistical techniques that are applied in this thesis. This includes two classes of statistical modeling approaches which have been commonly applied in plenty of research areas for many decades. Namely, we will describe the fundamental ideas about mixed effects models and factor analytic (FA) models. To be specific, this chapter covers several types of these two classes of modeling approaches. For the mixed effects models, we briefly describe the linear, generalized and multivariate mixed effects models, while for the FA models, exploratory FA (EFA), confirmatory FA (CFA) models and multilevel FA (MFA) models are covered. As an extension of FA models, structural equation modeling (SEM) and multilevel SEM are also briefly described. we also discuss the two classical estimating methods, i.e. the frequentist and the Bayesian approach, with the latter chosen as the analytic algorithm for our proposed models.