This paper presents a modification of Warner’s [8] Randomized Response model. According to O’Muircheartaigh et al [7], non-response is inevitable in a survey; in view of this, our model further reduces the non-response bias by further sampling for the non-respondents. In this paper we performed an empirical practice of our model and we also performed the empirical comparison of our model with Warner [8] model. We discovered that our model is more efficient than the Warner [8] model.