The goal of this project is to rigorously evaluate the utility of intra-individual variability (IIV) on aging-related changes in cognitive, health and emotional outcomes using a recently developed statistical model. The project covers three years and makes use of existing data sets of adult populations and it includes data-based simulations to examine minimal and optimal data needs for inclusion of intensive measurement designs within long-term longitudinal studies. The project addresses a number of critical issues that surround current and past research results. Up to date, the vast majority of studies that investigate the relation of short-term variability, as expressed in IIV, are based on multistep approaches which separate intra-individual means and trends from the computation of IIV. These approaches assume that means and standard deviations are completely independent and that they can be reliability estimated. These assumptions typically go untested but if they are not met, the results of analyses based on IIV can be severely biased in the sense that IIV explains undue amounts of variance. Mixed-effects location scale models (LSM) overcome these issues as they incorporate all relevant effects simultaneously. A main objective is to replicate existing results on the basis of these novel models and to evaluate new questions about the relation of and its independent predictive value for mortality and change-based outcomes in cognitive, health, and affect. The second goal focuses on optimization of intensive measurement burst designs (IMD), increasingly important for aging-related research, in terms of study design features that optimize statistical power, accuracy, and precision to estimate IIV. Specifically, the project follows three specific aims: 1) To perform multi-study replication of the predictive utility of IIV on subsequent change in cognition, health, and emotions and to evaluate the sensitivity of results to statistical models. 2) Evaluate novel multivariate hypotheses about the dynamics of IIV to level of functioning and to other aging-related outcomes. 3) Investigate optimal design features for maximizing power and reliability of embedded intensive measurements within longitudinal studies of aging. The long-term objective is to contribute to the early and precise detection of cognitive, health and emotional change. This project rigorously tests the predictive power of IIV in different domains and clarifies its utility in agin research.
This project evaluates the predictive value of intra-individual variability on long-term changes in cognition, health and affect. A number of existing longitudinal studies are investigated using novel modeling techniques that allow rigorous testing of the relation among intra-individual variability, individual levels, and long term outcomes. The project will clarify the importance of intra-individual variability as early marker of later change in affet, health, and cognition in older adults.
| Karr, Justin E; Areshenkoff, Corson N; Rast, Philippe et al. (2018) The unity and diversity of executive functions: A systematic review and re-analysis of latent variable studies. Psychol Bull 144:1147-1185 |
| Rush, Jonathan; Hofer, Scott M (2017) V. DESIGN-BASED APPROACHES FOR IMPROVING MEASUREMENT IN DEVELOPMENTAL SCIENCE. Monogr Soc Res Child Dev 82:67-83 |
