A new unrelated question randomized response model

A Multi-proportion Randomized Response Model Using the Inverse Sampling

“…From the innovative paper of Warner (1965) a lot of improvements have been implemented and a lot of more refined techniques have been developed. Multiple randomized response devices (see Mangat and Singh 1990;Singh 2002;Gjestvang and Singh 2006), modified randomized response techniques (see Mangat 1994;Kuk 1990), models taking into account the probability of lying (see Mangat and Singh 1995), models with one or more scrambled variables (see Gupta et al 2002;Gjestvang and Singh 2006) and models with unrelated questions (see Singh and Mathur 2006;Pal and Singh 2012)-just to report some examples-are all models or techniques originated by the intuition of Warner. In particular, an interesting extension to the case where the aim is the estimation of the proportions of t-related mutually exclusive population groups, with at least one and at most t 1 of which exhibit sensitive characteristics, is due to Abul-Ela et al (1967).…”