# Some new quantitative randomized response models using optional and partial scrambling for sensitive data

## TL;DR

This paper introduces four new models for handling sensitive data that improve privacy and accuracy in surveys.

## Contribution

The paper proposes novel randomized response models with optional and partial scrambling for better efficiency and privacy.

## Key findings

- The proposed models outperform existing methods in terms of relative efficiency.
- They offer improved privacy protection for sensitive quantitative data.
- A new weighted score confirms the superior performance of the models.

## Abstract

This research proposes four new optional and partial quantitative randomized response models to be used for the estimation of mean and sensitivity level of quantitative variables. These models are constructed based on the current quantitative scrambling and randomization methods and seek to produce unbiased estimators with better efficiency and privacy. We compare the proposed models based on standard comparison measures, such as relative efficiency, privacy protection, and a new weighted score. The results show that proposed models provide better performance compared to the current methods and are, therefore, very appropriate for surveys dealing with sensitive information.