B1288
Title: Advances on permutation tests for non-monotonic alternatives
Authors: Stefano Bonnini - University of Ferrara (Italy)
Michela Borghesi - Università degli Studi di Ferrara (Italy)
Massimiliano Giacalone - University of Campania Luigi Vanvitelli (Italy) [presenting]
Abstract: Sample surveys and related inferential statistical methods are suitable tools to support decision-making practices for policymakers. For instance, to achieve the environmental sustainability goals connected to Circular Economy and product and process innovation, Governments and Parliaments should evaluate and stimulate the propensity of companies towards such virtuous behaviors. This practice becomes fundamental in a country like Italy, with an economic system mainly based on small and medium enterprises. The focus is on complex tests of hypotheses for investigating the effects of important factors, such as company size, on the propensity of enterprises towards a Circular Economy. Frequently, the complexity is given by the V-shaped or U-shaped effects of ordinal factors in multisample problems. In other words, in the tested hypothesis, the propensity towards Circular Economy is decreasing up to a pick point and then increasing with respect to the ordinal factor under study. The proposed solution is based on a nonparametric method that consists in a combined permutation test. The good performance of the method and its usefulness in real data problems are proved through a Monte Carlo simulation study and the application to data concerning a survey on the Circular Economy, respectively.
