We present a new Monte-Carlo method for finding the solution of an estimating equation that can be expressed as the expected value of a 'full data' estimating equation in which the expected value is with respect to the distribution of the missing data given the observed data. Equations such as...