ABSTRACT
This chapter develops a generalized linear mixed model (GLMM) approach to the analysis of prospective panel data using accelerated longitudinal cohort designs. It includes empirical applications that continue to reveal patterns and mechanisms underlying social stratification of aging and health. The chapter shows how to use GLMMs to disentangle the effects of aging and birth cohort in longitudinal panel study designs. It examines different questions essential to the understanding of independent age and cohort effects. The life course changes in health disparities that were found in different studies were actually due largely to cohort differences. Taking cohort effects into account substantially modifies the existing understanding of the relationships between social inequalities and aging. The waves of the Health and Retirement Survey (HRS), as compared to the Americans’ Changing Lives, increased the power of the hierarchical age-period-cohort-growth curve models in tests of aging-related effects within each cohort, but the fewer HRS cohorts may decrease the power of tests for cohort differences.
