We are engaged in methodological research focused on the measurement and modeling of between-person differences in within-person change and variation. The study of development and aging-related processes presents many challenges and has contributed a rich history of substantive and methodological debates and advances. The resulting innovative developmental designs, improvements in measurement for detecting within-person change, and statistical advances in dynamic modeling and population inference conditional on mortality and attrition are meeting many of these concerns. Central to progress in research on development and aging processes is a focus on within-person change and variation from a variety of repeated measurement designs. Inferences regarding within-person change, however, are conditional on the research design and often complex given the joint influences of multiple processes, including age and aging, health, learning, and historical differences in context. A major emphasis is on the replication and generalizability of results across different samples, measurements and designs in research involving measurement harmonization and integrative data analysis across studies differing in country, culture, and cohort.
Hofer and Sliwinski (2001; Hofer, Sliwinski, & Flaherty, 2002; Hofer, Flaherty, & Hoffman, 2006) demonstrate, both analytically and in a simulation, how variance decomposition analysis of age-heterogeneous cross-sectional studies may be misleading because high levels of association between time-dependent processes result simply from average population change. Using an alternative narrow age-cohort design, Hofer, Berg, & Era (2003) did not find evidence congruent with findings from age-heterogeneous studies of high proportions of shared age-related variance among measures of perceptual acuity, balance, muscle strength and cognitive capabilities. Cross-domain associations were relatively weak and inconsistent in combined 75-year-old cohort samples from Denmark, Finland, and Sweden (NORA Study). These findings indicate that the effects of aging on sensory acuity, balance and cognitive functioning are likely to be largely independent, multidimensional, and complex at the level of the individual.
Integrative data analysis methods include meta-analysis, pooled-data approaches, and coordination of measurement and analysis protocols. Integrative analysis methods have enormous potential for making efficient use of the substantial resources provided by the many excellent studies on aging. Synthesizing results across studies has important benefits in terms of direct evaluation of the generalizability of results and increasing statistical power.