Hierarchical Linear Modeling

Hierarchical linear modeling (HLM, also known as multilevel modeling) is a statistical approach for analyzing hierarchically clustered observations. Observations may 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.

This entry illustrates key basic concepts using a two-level model with a continuous outcome variable. The example is based on a subset of the 1982 High School and Beyond Survey, which includes data on 7,185 students nested within 160 ...

  • 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