Hierarchical Linear Modeling

Abstract Hierarchical linear modeling (HLM) is an extension of linear regression analysis that allows analyzing variance in a dependent variable at multiple hierarchical levels. It is known by various names in the literature, such as multilevel analysis (or multilevel modeling), random coefficient model, mixed‐effects model and covariance components model. It is a statistical regression approach that is used when data have a hierarchical or so‐called nested or clustered structure which means that observations at one level of analysis are nested within observations at another level. In cross‐cultural research, the hierarchical structure consists usually of individuals (referred to as level 1 or individual level) who are nested within a specific country or culture (level 2 or country level), which requires a two‐level analysis. The outcome variable is then measured at the individual level and predictor variables can be used at both the individual and country level. A particular characteristic of multilevel modelling is that it takes into account the error associated with sampling at each level (e.g., individuals and countries) of analysis.