Statistics under incomplete knowledge of data
Jorg Pfleiderer, P. Krommidas · Monthly Notices of the Royal Astronomical Society · 1982
A different and generalized formulation and interpretation is proposed of the method used by Avni et al. to determine population distributions from preselected random samples with insufficient knowledge of the data, e.g. from intensity measurements with definite flux density results as well as upper limits. The information contained in inconclusive or inexact results is fully used. A set of implicit equations for the estimation of the population frequencies or probabilities is derived in the non-parametric case. The set can be solved by iteration. The errors in the resulting probabilities are also estimated. The parametric case leads to another set of implicit equations.