Differentially Private Confidence Intervals for Population Proportions under Stratified Random Sampling

The authors propose three differentially private algorithms for constructing confidence intervals for population proportions under stratified random sampling, providing privacy guarantees and asymptotic coverage properties.