Estimating positive surveys from negative surveys

The negative survey is an emergent survey method, which could protect sensitive data and individual privacy. Because positive survey results are needed in most situations, it is essential to estimate positive surveys from negative surveys. However, the traditional method for reconstructing positive surveys from negative surveys could return negative values (i.e. less than zero), and obviously this is impractical. In this paper, two novel methods to estimate positive surveys from negative surveys are proposed. Both methods can return nonnegative and proper values, and their results are identical. Simulation experiments demonstrate that the proposed two methods return more reasonable results than the traditional method.