RRTCS: An R Package for Randomized Response Techniques in Complex Surveys.

Area of Science:

• Survey Methodology
• Statistics
• Data Privacy

Background:

• Randomized response (RR) techniques improve data reliability and confidentiality for sensitive topics.
• Standard RR methods are typically designed for binary responses and simple random sampling.
• Existing methods face challenges with quantitative sensitive data and complex survey designs.

Purpose of the Study:

• To develop a method for estimating sensitive characteristics in complex survey designs.
• To address limitations of standard RR techniques with quantitative variables and complex sampling.
• To provide a computational tool for improved parameter estimation in sensitive surveys.

Main Methods:

• Development of a new computer program for parameter estimation.
• Adaptation of RR techniques for complex sampling designs (stratification, clustering).
• Consideration of second-order inclusion probabilities for variance estimation.

Main Results:

• The new program facilitates estimation of sensitive characteristic parameters under complex sampling.
• The method accounts for departures from simple random sampling assumptions.
• Improved accuracy in estimating finite population parameters from sensitive data is achievable.

Conclusions:

• A novel computational approach enhances the estimation of sensitive data in complex surveys.
• The developed program offers a practical solution for handling sensitive quantitative variables.
• This advancement supports more reliable data collection and analysis in sensitive research areas.