Hierarchical linear modeling is an analysis that takes the hierarchical structure of the data into account. Hierarchically structured data is nested data where groups of units are clustered together in an organized fashion, for example, students within classrooms within schools. 

HLM incorporates one or more continuous variables (such as intake scores) and class scores, such as diagnosis, to predict a continuous dependent variable, such as the change score. 

HLM can also be called multi-level modeling. It can be used for the purpose of prediction. It can also be used for the purpose of data reduction and drawing out causal inference.