The use of the iterated Bootstrap to find estimators that have the correct expectations is now standard. However when the distributions are skewed, or without means, the median makes more sense to us. This paper is primarily concerned with an algorithm that produces estimators whose median equals the unknown parameter. The method is illustrated by its application to four troublesome parametric estimation problems and a dataset.
All Science Journal Classification (ASJC) codes
- Statistics and Probability
- Statistics, Probability and Uncertainty
- Applied Mathematics
- Estimating equations