Multilevel Modelling


Health Economics

  • The analysis of data that have a hierarchical structure and that yield information about the impact each level of the hierarchy has on the dependent variable of interest. This is done by decomposing the error term in regression equations into component parts that are attributed to the various levels. For example, if hospital care is organized hierarchically by region, area and district (descending order) the outcome in any particular district is a linear function, having an intercept, a slope and three random error terms relating to region, area and district respectively.