On the Utilization of Direct as Well as Collateral Information in the Problem of Statistical Estimation [Discussion]

Attempts to formulate a general theory of statistical estimation date back as far as the eighteenth century. It was only recently, however, that a method was developed, the method of Maximum Likelihood, which, although not general in the strictest sense, takes care of a relatively large class of problems of estimation. A good number of statisticians seem reluctant, however, to accept this method in general, questioning its appropriateness in specific applications, as well as doubting the soundness of its arbitrary choice of criteria.