Synthesized randomized response techniques

Abstract

Reducing response bias in survey research is important in ensuring that the data collected accurately represents the population of interest. This study proposes the synthesized random response technique (SRRT) estimator as an efficient method to reduce response bias in this regard. We compared the proposed SRRT estimator with other conventional techniques, such as the Hybrid Tripartite Randomized Response Technique (HTRRT), the Alternative Tripartite Randomized Response Technique (ATRRT), and the Two-Stage Unrelated Question Randomized Response Model for Estimating Prevalence of Stigmatized Attributes (TSUQRRMEPSA). Our results show that the proposed SRRT estimator consistently produces more efficient estimates than other conventional techniques as sample sizes increase. As a result, the proposed SRRT estimator have the potential to enhance the accuracy of survey data, which is sensitive for informed decision-making. Hence, the proposed method is more efficient in reducing response bias than other conventional methods.