Aggregate-Level PUF as a New Alternative to the Traditional Unit-Level PUF for Improving Analytic Utility and Data Confidentiality

• Activity Number: 349
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Avinash Singh and Joshua M Borton*+
• Companies: NORC and NORC
• Keywords: Aggregate Level Analysis ; Calibration ; Micro-Group ; Micro-Mean ; MG-MC ; Subsampling

Abstract

In light of the seemingly bleak future of analytically useful PUFs due to increasing potential for availability of individual level data in public domains that could be used for matching purposes, we propose an aggregate level PUF (AL-PUF) as an alternative product consisting of information at aggregate or micro-group (MG) level. In the context of CMS Medicare Claims data, for each MG of beneficiaries defined by demography, geography, and enrollment; AL-PUF would contain information about MG-size, and micro-means (MMs) over the MG for various clinical profiles. Definition of MGs is based on a sample (s1) such that they satisfy a threshold of minimum size to avoid disclosure of rare subgroups but are small enough as building blocks for various analysis domains of interest. To avoid disclosure of sensitive values of individuals with rare characteristics, we propose using two successive sub