Description
The value of individual-level data for research must be balanced against the privacy concerns of individuals. I discuss an approach for generating synthetic data sets that preserve relevant statistical properties of input data while also providing privacy guarantees for individuals in the data.