 2019 Royal Statistical Society 1369–7412/20/82269 J. R. Statist. Soc. B (2020) 82, Part 1, pp. 269–271 Correction: ‘A new randomized response model’ Lovleen Kumar Grover and Amanpreet Kaur Guru Nanak Dev University, Amritsar, India [Received July 2019. Revised September 2019] Summary. We point out a minor mistake in ‘A new randomized response model’, published in 2006, which has been cited by various researchers, though no one has pointed out the mistake. Keywords: Direct survey model; Randomized response model; Scrambled responses; Sensitive population proportion; Unbiased estimator 1. Introduction Gjestvang and Singh (2006) suggested a new efficient randomized response model. They found an unbiased estimator of the sensitive population proportion and obtained its variance. Further, they showed that their suggested model can easily be adjusted more efficiently than the Warner (1965), Mangat and Singh (1990) and Mangat (1994) models by the suitable selection of certain parameters of the suggested randomized response model. In section 2 of Gjestvang and Singh (2006) (see page 525), they claimed that all the three models due to Warner (1965), Mangat and Singh (1990) and Mangat (1994) are special cases of their suggested model. They specified the following three points. (For convenience, we use exactly the same notation and symbols as those of Gjestvang and Singh (2006).) (a) If p.1 + β 1 θ 1 / + .1 - p/.1 - α 1 θ 1 / = p 0 and T β 2 θ 2 - .1 - T/α 2 θ 2 = 1 - p 0 , then their new model reduces to Warner’s (1965) model. (b) If p.1 + β 1 θ 1 / + .1 - p/.1 - α 1 θ 1 / = .1 - p 0 /.1 - T 0 / and T β 2 θ 2 - .1 - T/α 2 θ 2 = 1 - .1 - p 0 /.1 - T 0 /, then their new model reduces to Mangat and Singh’s (1990) model. (c) If p.1 + β 1 θ 1 / + .1 - p/.1 - α 1 θ 1 / = 1 and T β 2 θ 2 - .1 - T/α 2 θ 2 = 1 - p 0 , then their new model reduces to Mangat’s (1994) model. However, these claims are wrong. We justify our assertion theoretically as follows in the next section. 2. Theoretical justification Taking the following equations after simplifying, we note that p.1 + β 1 θ 1 / + .1 - p/.1 - α 1 θ 1 / = α 1 α 1 + β 1 .1 + β 1 θ 1 / +  1 - α 1 α 1 + β 1  .1 - α 1 θ 1 / = 1, T β 2 θ 2 - .1 - T/α 2 θ 2 = α 2 α 2 + β 2 β 2 θ 2 -  1 - α 2 α 2 + β 2  α 2 θ 2 = 0, always: .1/ Address for correspondence: Lovleen Kumar Grover, Department of Mathematics, Guru Nanak Dev University, Amritsar 143005, Punjab, India. E-mail: lovleen 2@yahoo.co.in