Past Workshops

Yajuan Si

Applying Weights to Longitudinal Survey Data

September 15, 2022, 1-5pm

Yajuan Si

Research Associate Professor in the Survey Methodology Program

Longitudinal surveys collect rich data about individual characteristics and enable trajectory modeling to understand the developmental course and lifespan. Examples include the Monitoring the Future (MTF) panel study, the Panel Study of Income Dynamics (PSID), and the Midlife in the United States (MIDUS) Series. However, the quality of estimates can be attenuated by panel attrition. Moreover, sample selection procedures for longitudinal studies are often complex in nature, including design features such as stratification, cluster sampling, and survey weights for probability samples. A failure to account for these selection features in estimation could affect the inferential validity and generalizability of descriptive summary measures and estimates of trajectory models. Weighting approaches have been popular in the survey statistics literature to simultaneously adjust for these complex sample design features and panel attrition. However, there is no clear consensus in the literature on whether or how to apply weighting adjustments for attrition in longitudinal trajectory modeling. In this workshop, we will provide an overview of the main methodological issues and focus on hands-on analysis with survey weights using statistical software (e.g., R). We will illustrate the workflow with applications to data from the MTF, PSID, and MIDUS studies.

Workshop Zoom recording

Workshop materials (Zip file)