[:es]Título: Permutation tests to check assumptions in DEA as well as perform second stage analysis
Ponente: Mette Almild (University of Copenhagen)
Organizador: Jesús Pastor
Fecha: Martes 24 de marzo a las 12:00 horas.
Lugar: Sala de seminarios del CIO en el Edificio Torretamarit, Universidad Miguel Hernández (Campus de Elche)
Resumen: Recently, permutation tests have been proposed for statistical inference within the domain of DEA. These tests do not rely on asymptotic properties and are therefore useable for the limited sample sizes encountered in most practical applications. Permutation tests can be used both to check some of the assumptions underlying the use of DEA, but also for second-stage analysis of the corresponding results. The former includes tests for returns to scale and for comparison of production frontiers across independent groups of production units. The latter includes tests for hypotheses on the Malmquist Index and its components. We here provide an overview of the permutation tests developed to date, explain how they can be used, and provide examples of empirical applications of some of the tests.[:]
Categories: Novedades