Data Confidentiality, Privacy, and Deidentification

NORC has the experienced staff to meticulously safeguard respondents’ data during data collection and to release statistics that protect the identity of individual respondents. NORC staff has experience using various techniques to limit the disclosure risk of tabular data products. This includes using readily available off the shelf software to perform complementary cell suppression as well as the development of in house complementary cell suppression algorithms to meet clients’ needs. NORC has also developed methods using log linear models to quantify the disclosure risk of a given cell suppression strategy, and guide further treatment to meet disclosure risk thresholds.

NORC has the experience and software to produce releasable data files (or products) that balance utility and the risk of disclosing individual respondent’s information. Aggregate Level Public Use File (AL-PUF) methodology is an SDL method developed at NORC, that uses low level aggregation and subsampling to produce analytic meaningful summary statistics for arbitrary domains while introducing adequate uncertainty to estimates about individuals and very small domains. In addition, NORC staff are familiar with traditional k-anonymity methods, their application and their limitations regarding both analytic utility and disclosure risk.