No Detail Found!

Introducing a weighted measure of privacy and efficiency for comparison of quantitative randomized response models Muhammad Azeem

Author(s)
Muhammad Azeem Department of Statistics, University of Malakand,Khyber Pakhtunkhwa,Pakistan
Abstract
In sample surveys on sensitive variables, the randomized response technique is commonly employed by researchers to ensure the respondents' privacy protection and hence minimizing the non-response rate. The privacy protection level and efficiency are the two important indicators of the quality of any randomized response technique. The existing unified measure of model efficiency and privacy level for comparison of randomized response models assigns equal weightage to respondents' privacy protection and model's efficiency. In practice, some situations may prefer privacy over efficiency whereas in other situations, efficiency may be more important than privacy. This paper introduces a unified weighted metric of privacy protection level and efficiency for comparison of quantitative randomized response techniques. The practical usefulness of the proposed measure is that it enables the researchers to assign different weights to privacy and efficiency as per requirement of the situation. Two existing randomized response models are compared using the proposed measure and the results are discussed.
Publication Details
Page(s) 377-385
DOI DOI not available
Published Journal: Pakistan Journal of Statistics, Volume: 39, Issue: 3, Year: 2023

Request the full text from PASTIC

Keywords
privacy protection, Model efficiency, randomized response models, weighted measure., sensitive surveys
0

Citations

0

Downloads

39

Views