Utilizing stratified double response techniques in public health investigations to optimize privacy and efficiency via sensitive data estimation

This study examines the complex issue of specifically quantifying sensitive quantitative variables while ensuring respondents' privacy and maintaining data integrity. The present study presents two novel optional stratified double response models (OSDRMs) that integrate simultaneously additive and subtractive frantically mechanisms to improve privacy protection while maintaining efficiency. The models utilize two scrambling variables for each response and employ stratified random sampling (SRS) to divide the population into similar subgroups (strata), ensuring that each subgroup is representative. In every stratum, simple random sampling with replacement (SRSWR) is employed, resulting in more precise estimates. The OSDRMs develop unbiased and viable estimates by assessing the mean as well as the sensitivity level for highly sensitive information, thereby reducing the likelihood of response bias and social desirability effects. The mathematical derivations and comprehensive simulation studies indicate notable enhancements in relative efficiency when contrasted with conventional randomized response methods. We investigated the suggested models in actual scenarios by conducting a cross-sectional survey in the Lahore, Faisalabad, and Sargodha districts of Punjab, Pakistan. The results demonstrate that OSDRMs are an effective, privacy-protecting technique for collecting sensitive data, resulting in improved precision and dependability of data. This strategy is significant for public health management because it promotes ethical data gathering while also improving monitoring and policy formulation in circumstances involving sensitive data.