Abstract:
|
In this paper, we propose a new unrelated question model for estimating the proportion of a sensitive attribute in a population on the same lines of Odumade and Singh (2009) model of efficient use of two decks while collecting in information from respondents in a sample. The proposed estimator is shown to be more efficient and protective than the Warner (1965), Odumade and Singh (2009), Singh and Sedory (2011, 2012) and Greenberg et al. (1969) unrelated question models.
|