Identity Disclosure Control in Microdata Release by Post-Randomization

• Activity Number: 354 - SPEED: Big Data, Small Area Estimation, and Methodological Innovations Under Development, Part 2
• Type: Contributed
• Date/Time: Tuesday, July 30, 2019 : 10:30 AM to 11:15 AM
• Sponsor: Survey Research Methods Section
• Author(s): Xiaoyu Zhai* and Tapan Nayak
• Companies: and George Washington University
• Keywords: identity disclosure control; PRAM; data utility; Public-use Micro-data

Abstract

Motivated by RR methods in sampling, Gouweleeuw et al. (1998) introduced Post Randomization (PRAM) Method to protect categorical data from disclosure. PRAM can be applied independently to multiple categorical variables or jointly to the cross-classification variable. Perturbation probabilities for the cross-classification are needed for proper inferences and measuring confidentiality. We adopt the identity disclosure concept in previous literatures, which is considered the most serious confidentiality disclosure, and propose a new structure to control identity disclosure risk for a prefixed upper bound. We derived and explored properties of this structure and examined the trade-off between confidentiality disclosure and data utility. At last, we described the methodology in full details and applied it to a Public-use Micro-data Sample. In the empirical study, we showed that we achieved the disclosure control goals and maintained a sound data utility.