A technique for eliciting truthful answers to sensitive questions in a survey, used to estimate an unknown proportion of population belonging to the respective sensitive group. For example, a researcher is interested in estimating the proportion of students who use illicit drugs at least once a week. Before answering ‘Yes’ or ‘No’ to the question ‘Are you using illicit drugs at least once a week?’ the students in the sample are asked to flip a coin and answer truthfully if the coin shows heads, otherwise to answer ‘Yes’, without revealing the outcome of flipping a coin. Then the estimate of the true proportion of non-users is twice the proportion of the respondents whose answer was ‘No’.