DETECTING OUTLIERS IN DETERMINISTIC NONPARAMETRIC FRONTIER MODELS WITH MULTIPLE OUTPUTS

This article provides a statistical methodology for identifying outlie rs in production data with multiple inputs and outputs used in determi nistic nonparametric frontier models. The methodology is useful in ide ntifying observations that may contain some form of measurement error and thus merit closer scrutiny. When data checking is costly, the meth odology may be used to rank observations in terms of their dissimilari ty to other observations in the data, suggesting a priority for furthe r inspection of the data.