Executive Summary : | Surveys often collect data on stigmatized characteristics, such as drug abuse, sexual deviancy, and tax fraud. To protect participants' privacy, special survey techniques are essential. Warner (1965) introduced the Randomized Response Technique (RRT), which uses randomization devices like coins, dice, and decks of cards to elicit responses from interviewees. The results consist of statements related to the sensitive characteristic and sometimes non-sensitive ones. There is a need for more effective, simple, and easy-to-understand RRTs that provide increased anonymity. The project aims to design RRTs for rare sensitive characteristics under Poisson approximation, multivariate estimation of sensitive characteristics, and multistage RRTs for increased privacy. Simulation data will be used to study the efficiency of the proposed RRTs and compare them with contemporary RRTs in terms of privacy protection measures. The project will involve collecting and reviewing relevant literature, identifying practical problems related to real-life situations, identifying appropriate parameters, developing effective RRT models, and generating data under suitable software frameworks to simulate real-world scenarios. Empirical investigations will be conducted using statistical software such as STATISTICA, SPSS, MATLAB, or R. Recommendations for real-life problem applications will be made to survey statisticians and practitioners, and manuscripts will be communicated to national and international journals. The RRTs developed under this project will be of interest to survey statisticians and researchers working in various fields, such as drug abuse, alcohol addiction, sexually transmitted diseases, abortions, rapes, and criminal activities. |