Longitudinal selectivity in aging populations: Separating mortality-associated versus experimental components in the Berlin Aging Study (BASE)

The authors examined 3.7-year selectivity in the Berlin Aging Study by comparing the T1 parent sample (N = 516) with the T3 sample (N = 206). Selectivity was partitioned into a mortality-associated component, reflecting the degree to which individuals still alive at T3 (T3 survivors, N = 313) differ from the T1 parent sample (N = 516) from which they originated, and an experimental component, reflecting the degree to which the T3 sample (N = 206) differed from T3 survivors (N = 313). Across 48 variables representing medical, sensorimotor, cognitive, personality-related, and socioeconomic domains, the mortality-associated component accounted for 64% of total selectivity, and the experimental component for 36% (0.18 vs 0.10 SD units; t = 7.20,p < .01). Except for age and intelligence, experimental selectivity effects regarding means and prevalence rates were generally small. Partitioning selectivity into mortality-associated and experimental components is a useful tool in the longitudinal study of aging populations.