Joshua Snoke is a statistician at RAND. He researches statistical data privacy, algorithmic fairness, and workforce development.
He has expertise broadly in statistical data privacy and has published technical papers and policy reports on topics such as synthetic datasets, model estimation across multiple databases, and differentially private algorithms. He focuses mainly on evaluating and developing practical applications of privacy-preserving methodologies to identify technologies which increase access to administrative or survey data while maintaining privacy and confidentiality.
Outside of privacy, he works on algorithmic fairness and is currently working on applying these ideas to applications in the Department of Defense (DoD) and healthcare. He also works on evaluating programs and policies for workforce development. He focuses on DoD and DHS manpower and personnel questions, such as evaluating recruitment and promotion strategies, evaluating predictors of success in military personnel, and evaluating barriers to diversity in the workforce.
His work, both privacy-related and otherwise, involves a variety of administrative and survey data, such as data from the U.S. Census Bureau, the Internal Revenue Service, the UK Administrative Data Research Network, the Defense Manpower Data Center, and the American Educator Panels. His broader methodological interests include estimation, model selection, non-parametric modeling and classification, and machine learning.
He received his Ph.D. in statistics with a graduate minor in social data analytics from the Pennsylvania State University. He received his B.S. in mathematics and economics from Wheaton College.
Selected Publications
Cabreros, Irineo, Joshua Snoke, Osonde A. Osoba, Inez Khan, and Marc N. Elliott, Advancing Equitable Decisionmaking for the Department of Defense Through Fairness in Machine Learning, RAND Corporation (RR-A1542-1), 2023
Bowen, C. M. and J. Snoke, "Comparative Study of Differentially Private Synthetic Data Algorithms from the NIST PSCR Differential Privacy Synthetic Data Challenge," Journal of Privacy and Confidentiality, 11(1), 2021
Matthews, Miriam, Andrew R. Morral, Terry L. Schell, Matthew Cefalu, Joshua Snoke, and R. J. Briggs, Organizational Characteristics Associated with Risk of Sexual Assault and Sexual Harassment in the U.S. Army, RAND Corporation (RR-A1013-1), 2021
Snoke, J. and C. M. Bowen, "How Statisticians Should Grapple with Privacy in a Changing Data Landscape," CHANCE, 33(4), 2020
Snoke, J., G. M. Raab, B. Nowok, C. Dibben, and A. Slavković, "General and specific utility measures for synthetic data.," Journal of the Royal Statistical Society: Series A (Statistics in Society), 181(3), 2018
Snoke, Joshua, T. R. Brick, A. Slavković, and M. D. Hunter, "Providing accurate models across private partitioned data: Secure maximum likelihood estimation.," The Annals of Applied Statistics, 12(2), 2018
Snoke, J. and A. Slavković, "pMSE Mechanism: Differentially Private Synthetic Data with Maximal Distributional Similarity.," Privacy in Statistical Databases, 2018
Robson, Sean, Maria C. Lytell, Matthew Walsh, Kimberly Curry Hall, Kirsten M. Keller, Vikram Kilambi, Joshua Snoke, Jonathan W. Welburn, Patrick S. Roberts, Owen Hall, and Louis T. Mariano, U.S. Air Force Enlisted Classification and Reclassification: Potential Improvements Using Machine Learning and Optimization Models, RAND Corporation (RR-A284-1), 2022