Respondent-specific randomized response technique to estimate sensitive proportion

Article Type

Research Article

Publication Title

Statistics in Transition New Series

Abstract

In estimating the proportion of people bearing a stigmatizing characteristic in a community of people, randomized response techniques are plentifully available in the literature. They are implemented essentially using boxes of similar cards of two distinguishable types. In this paper, we propose a more general procedure using five different types of cards. A respondent-specific randomized response technique is also proposed, in which respondents are allowed to build up the boxes according to their own choices. An immediate objective for this change is to enhance, sense of protection of privacy of the respondents. But as by-products, higher efficiency in terms of actual coverage percentages of confidence intervals and related features are demonstrated by a simulation study, and superior jeopardy levels against divulgence of personal secrecy are also reported to be achievable. AMS subject classification: 62D05.

First Page

53

Last Page

70

DOI

https://10.59170/stattrans-2023-051

Publication Date

9-1-2023

Comments

Open Access, Gold

This document is currently not available here.

Share

COinS