ABSTRACT
Social systems are mostly organized in hierarchical structures. Therefore, conducting analyses without considering the effect of clusters that individuals or groups belong to can lead researchers to engage in type I error. Multilevel analysis is an analytical technique that enables researchers to interpret hierarchically organized data. It is also known as multilevel models, hierarchical linear models or mixed models. Using this technique, researchers can account for higher level units’ (e.g., leadership) effect on lower level units (e.g., followers) in their analyses. In this chapter, we discuss different types of multilevel structures such as hierarchical designs and non-hierarchical designs. We also introduce different forms of multilevel models such as random intercept models, random slope models, full multilevel models and cross-level models. Finally, we present the statistical applications of multilevel models such as moderation, mediation and structural equation modeling.
