The work proposed in this competing continuation application will combine the methods and data developed in our NGCS studies to evaluate and improve the large-scale cognitive aging data based on the HRS. Our previous research on the Wechsler Adult Intelligence Scale (WAIS;NIA grant AG02695) and the Woodcock-Johnson tests (WJ-R;AG04704, and AG07407) has produced: (1) a large-scale data base of WAIS/WJ-R information useful foraging research, (2) novel structural equation models (SEM) appropriate for aging research, (3) practical results about the complex measurement functions of the WAIS/WJ-R, and (4) dynamic representations of the kinematic and kinetic functions of intellectual ability. During the past five years of this project (AG-07407) we have analyzed (5) ability measurement data on large samples (N>12,000) with multiple batteries, (6) three-wave longitudinal retest data on the National Growth and Change Study (NGCS) sample, (7) a seventh-wave of longitudinal data on the Bradway children (now 68-73), and (8) a sixth-wave of longitudinal data on the Berkeley Growth Study children (now 67-71). We have also developed (9) dynamic models for unraveling patterns of leading and lagging indicators of aging processes, and (10) IRT-based item-level analyses for converging information from different tests at different ages. Our most recent work uses cognitive data from the Health and Retirement Study (MRS), including analyses of latent age- curves of cognition, and the factorial structure of current MRS measures. Our working hypothesis is that the current MRS cognitive testing is efficient, and the resulting measures are useful, but HRS does not measure some key constructs in cognitive aging (e.g., reasoning, executive functioning, speed). To assess and improve the HRS procedures we will: (1) Develop new cognitive tests based on the WJ-R/ WAIS measures for telephone administration in the HRS;(2) Collect HRS telephone and NGCS in-person cognitive measures on a new large and nationally representative sample of persons (A/=1,118) born in 1947 or before, who will be measured longitudinally at two separate occasions at varying intervals (Dt=1-23 months);(3) Identify the age-related cognitive ability factors in the HRS across new telephone measures and traditional face-to-face measures by convergent measurement analyses based on contemporary SEM/IRT techniques;(4) Analyze the dynamic age changes across cognitive measures, and in relations among spouses in cognition and related to demographics, personality, lifestyle, health, and economics;(5) Improve the cognitive measurements in current HRS analyses and future HRS data collections, and create a public-use data set. The resulting NGCS+HRS data set will provide a large-scale nationally representative public-use longitudinal dataset with persons from many households measured on indicators of many key constructs related to normal cognitive abilities and functioning. These new results will help synthesize past research on current HRS cognitive measures, explicitly define the internal validity of the HRS using standard cognitive batteries, add to the external validity of adult intellectual abilities with a variety of demographic, health, and economic variables, and provide a rich multivariate-longitudinal data source for other researchers.
Showing the most recent 10 out of 33 publications
