Go to navigation (press enter key)


Monika Hu, Duke University

Date: Monday, December 8, 2014

Time: 4:30pm

Location: Rockefeller Hall 310

Title: Generating Synthetic Household Data for the Decennial Census


Have you ever filled out surveys and worried about your submitted personal information not being protected? Such worries are sensible and legitimate, as researchers have demonstrated great risks associated with personal-level data released to the public. In this talk, I will introduce synthetic data, an approach to data dissemination which provides good data utility and sufficient privacy protection. In particular, I will present a novel two-level latent class model for generating synthetic household categorical data for the decennial census, and an illustrative example of household data synthesis of a sample of the 2012 American Community Survey public use file. 

Tea at 4:15pm in Rockefeller Hall 305