Regression Models in Complex Survey Sampling for Sensitive Quantitative Variables

Randomized response (RR) techniques are widely used in research involving sensitive variables, such as drugs, violence or crime, especially when a population mean prevalence must be estimated. However, they not generally applied to examine relationships between variable and other characteristics. This type of technique was initially qualitative studies later showed that logistic regression may performed with RR data. Since many the variables considered this context quantitative, were extended these cases estimate values required. Regression analysis is valuable statistical tool for exploring among establishing associations responses covariates. In article, we propose design-based complex sample designs based on unified approach. We present estimators coefficients, study their theoretical properties consider different ways variance. The estimation simulated using various quantitative randomized models. method proposed also analyse findings from real-world survey.