ISR Awards

Enhancing Synthetic Data Techniques for Practical Applications

The goal of the project is to enhance synthetic data techniques for practical applications. Specifically, the aims will develop novel methods to improve disclosure risk assessment, quality check verification, and population generalizability in the adjustment of complex survey design and weights. Dr. Si will offer expertise on Bayesian and survey statistics and experiences in survey operation and weighting adjustments. She will extend the methodology development on synthetic population generation and data integration to confidentiality protection. She will apply the proposed methods to enhance practical confidentiality protection through her close collaborations with colleagues from the Panel Study of Income Dynamics (PSID) and the Inter-university Consortium for Political and Social Research (ICPSR) at the University of Michigan.

Dr. Si will participate in the project for three years (Yr 1: 1.50 CM, Yr 2: 0.75 CM, and Yr 3: 0.25 CM). She will assist with all aims including data preparation, statistical methodology, software development and practical application. She is expected to lead the survey data analysis in practical applications. She will communicate the findings with survey organizers and researchers, such as from PSID and ICPSR, and improve the practice of synthetic data generation and release. Dr. Si will work closely with Dr. Reiter throughout the project with regularly scheduled meetings to generate high-quality deliverables in a steady process.

Funding:

Duke University

Funding Period:

08/15/2022 to 07/31/2025