Elaboration of HLM Growth Modeling Results

Abstract

Standard reporting of the modeling of individual growth or change curves with hierarchical linear models (HLM) typically includes a focus on a certain important results (e.g., the variance of the status of the outcome) at a single time in the growth curve, a time that is determined by the specification of the origin of the time scale. It is argued here that such reporting should be extended to show the variation of these important results over the time span of the study. The required procedure, involving only some simple matrix algebra and a technical graphics program, is illustrated with data for the nonlinear growth of reading ability for young children.