Multi-Factor Policy Evaluation and Selection in the One-Sample Situation

Firms nowadays need to make decisions with fast information obsolesce. In this paper I deal with one class of decision problems in this situation, called the “one-sample” problems: we have finite options and one sample of the multiple criteria with which we use to evaluate those options. I develop evaluation procedures based on bootstrapping DEA (Data Envelopment Envelopment) and the related decision-making methods. This paper improves the bootstrap procedure proposed by Simar and Wilson (1998) and shows how to exploit information from bootstrap outputs for decision-making.