Statistically Valid Inferences from Privacy-Protected Data

Unprecedented quantities of data that could help social scientists understand and ameliorate the challenges human society are presently locked away inside companies, governments, other organizations, in part because privacy concerns. We address this problem with a general-purpose access analysis system mathematical guarantees for research subjects, statistical validity researchers seeking science insights. build on standard “differential privacy,” correct biases induced by privacy-preserving procedures, provide proper accounting uncertainty, impose minimal constraints choice methods estimated. illustrate replicating key analyses from two recent published articles show how we can obtain approximately same substantive results while simultaneously protecting privacy. Our approach is simple to use computationally efficient; also offer open-source software implements all our methods.