Abstract
Delimitations of hierarchical linear modeling (HLM) were examined in terms of fixed and random effects in multilevel data analyses. The author used examples at the local and national levels to illustrate proper applications of HLM and dummy variable regression. Cautions are raised regarding circumstances under which hierarchical data do not need HLM.

This publication has 1 reference indexed in Scilit: