Case
Abstract
This is a case study of spatially mediated inequalities in diabetes prevalence in Oslo, Norway. I present a multilevel framework in which individuals are nested in neighborhoods and outline the key tenets by which such a framework facilitates the analysis of hierarchical data structures. The concepts of variance partition, random intercept, and random slope are defined and illustrated with reference to variation in the source of disease risk across analytic parameters. Guidance is provided for how to conceptualize and construct measures of higher level units and how to operationalize a multilevel model in empirical research practice.