Hierarchical linear modeling (HLM, also known as multilevel modeling) is a statistical approach for analyzing hierarchically clustered observations. Observations might be clustered within experimental treatment (e.g., patients within group treatment conditions) or natural groups (e.g., students within classrooms) or within individuals (repeated measures). HLM provides proper parameter estimates and standard errors for clustered data. It also capitalizes on the hierarchical structure of the data, permitting researchers to answer new questions involving the effects of predictors at both group (e.g., class size) and individual (e.g., student ability) levels. Although the focus here is on two-level models with continuous outcome variables, HLM can be extended to other forms of data (e.g., binary variables, counts) with more than two levels of clustering (e.g., student, classroom, and school). The ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles