Aging and cognition: methodological differences in outcome.

A series of longitudinal and cross sectional studies, collected as part of the Seattle Study, were reanalyzed. A longitudinal sequential analysis (N = 232), wherein subjects were measured every seven years for four time periods, was completed on the Primary Mental Abilities test. Cohort differences were at least as strong as age differences; cohorts were generally relatively stable over the measured periods, showing little cognitive decline. A cross-sectional sequential analysis (N = 2813) was completed for the same time periods; decline with age was more evident than with the longitudinal sequential analysis. A cross-sectional analysis for the fourth time period (N = 611) showed the most marked decline of all. Interpretation of outcome was highly dependent on the analysis used.