Joint Regression Analysis and Incorporation of Environmental Variables in Stochastic Frontier Production Function: An Application to Experimental Data of Winter Rye

Abstract

This chapter joins the main properties of two specific regression techniques, joint-regression analysis (JRA) and stochastic frontier approach (SFA) in the analysis of experimental data sets from a breeding program of winter rye (Secale cereale L.), conducted in Poland, Research Center for Cultivar Testing de Słupia Wielka, over the period 1997–1998. With JRA, a meta-model, based on several linear regressions, had been estimated in order to analyze multilocation trials of winter rye production and to select the best cultivars (more productive) for a related stratum (locality/genotype). With SFA, another regression model had been investigated to predict production rankings of cultivars, through individual efficiency estimates. These measures resulted from a stochastic production frontier on experimental data of production and different climate conditions. Both techniques show similar dominant cultivars for the same environments.