Data Linkage Methods
My recent methodological work focuses on novel approaches to longitudinal data linkage and methodologies that offer advantages over conventional studies of singular datasets and types of temporal measures for life course and aging research.
Longitudinal Integrative Data Analysis (IDA)
This approach is one of recent developments on complex coordinated data analysis (CDA) of multiple study samples. It is a methodological framework to combine and synthesize separate longitudinal datasets for enhanced inferences for testing life course hypotheses. My team has endeavored to make our work a resource that enables new analytic and modeling approaches for longitudinal studies across a variety of outcomes, supports data sharing, and helps build a more cumulative population science.
Research Projects:
- “Life Course Process of Alzheimer’s Disease: Sex Differences and Biosocial Mechanisms (National Institute on Aging/NIA R01AG057800; PI: Yang)” (Project website).
- Extensive life course designs integrated four U.S. population-based panel studies of over 50,000 individuals from most 20th century birth cohorts followed for up to 25 years.
- Integrated data creation: public code documentation provides open-source Stata code as an example for users to create the dataset that combines data from these four individual population-based longitudinal surveys.
- “Consortium for Longitudinal Behavioral and Social Science Data Integration and Coordination (CLASSIC) (NIA U24AG081810; PI: Neupert)” (Project website).
- Infrastructure support through a web portal of meta-data sharing to promote collaboration and coordination among longitudinal studies for cross-study comparison and CDA.
- Methodological support through consulting and methodological workshops on conducting advanced longitudinal data analysis.
Representative Publications:
Wang, L., & Yang, Y. C. (2025). The impact of marital trajectories on late-life social isolation: A U.S.-China comparison. Longitudinal and Life Course Studies, 138–172.
Yang, Y. C., Walsh, C. E., Shartle, K., Stebbins, R. C., Aiello, A. E., Belsky, D. W., Harris, K. M., Chanti-Ketterl, M., & Plassman, B. L. (2024). An early and unequal decline: Life course trajectories of cognitive aging in the United States. Journal of Aging and Health, 36(3–4), 230–245.
Yang, Y. C., Zhang, M., Shartle, K., Aiello, A. E., & Harris, K. M. (forthcoming). Social disconnection and cognitive aging across the life course: Coordinated longitudinal analysis of three national cohort studies. The Journals of Gerontology: Series B.
Joint Modeling of Longitudinal Trajectory and Time-to-event Data
This approach focuses on linking trajectory and transition as interrelated longitudinal measures of change within datasets. Specifically, I employ the joint latent class mixed models (JLCMM) recently developed in biostatistics that merge growth models, event history models, and latent class models for simultaneously modeling of longitudinal trajectory data and time-to-event data as joint temporal processes. This family of models not only can capture population heterogeneity in temporal progression of aging but also reduce bias in estimates of trajectory parameters and event risks.
Representative Publications:
Walsh, C. E., Yang, Y. C., Oi, K., Aiello, A. E., Belsky, D. W., Harris, K. M., & Plassman, B. L. (2022). Age profiles of cognitive decline and dementia in late life in the Aging, Demographics and Memory Study (ADAMS). The Journals of Gerontology: Series B, 77, 1880–1891.
Yang, Y. C., & Zhang, M. (in press). Advances in quantitative aging research: Data linkage and methodological approaches to modeling life course processes. In M. Shafer, D. Carr, J. Angel, & R. Settersten (Eds.), The handbook of sociology of aging (Chap. 5). Springer.