340 – SPEED: Applications of Advanced Statistical Techniques in Complex Survey Data Analysis: Small Area Estimation, Propensity Scores, Multilevel Models, and More
Prisoners Are People Too: Statistical Disclosure Control in the 2016 Survey of Prison Inmates
Nicole Mack
RTI International
Marcus Berzofsky
RTI International
Stephanie Zimmer
RTI International
Organizations releasing public or restricted use files attempt to minimize disclosure risks for participants while maintaining data utility. This presentation reviews the disclosure risk assessment for the 2016 Survey of Prison Inmates (SPI). SPI routinely collects identifiable and sensitive information such as health histories and offense histories from prisoners. Surveys of confined populations, such as prisoners, pose higher disclosure risk challenges than general population surveys. Populations in known, confined locations - like prisons - are more vulnerable to identification. Thus, we sought to determine the methods that best encapsulate the tradeoff between data utility and minimization of disclosure risk. Through crosstabulations we assessed risk and discerned how easily one can identify certain groups and individuals. We considered various disclosure avoidance methods to decrease disclosure risk such as coarsening, perturbation, and suppression. For each method, we describe disadvantages and advantages, evaluate the tradeoff, and propose how methods presented here could also be applied to other surveys with hierarchal groups such as schools or hospitals.