Age-Period-Cohort Analysis

9

Mixed Effects Models: Hierarchical APC-Growth Curve Analysis of Prospective Cohort Data

9.1 Introduction

The repeated cross-sectional data designs for which generalized linear mixed models (GLMMs) in the form of hierarchical age-period-cohort (HAPC) models were specified and estimated in Chapters 7 and 8 rely on information from synthetic cohorts that contain different cohort members at each point in time. That is, the cohorts defined in these designs are synthetic in the sense that they contain different individuals as opposed to the same group of individuals at each calendar period. Inferences drawn from such designs about cohort and age effects therefore rest on the assumptions that synthetic cohorts mimic true cohorts and changes ...