出版社：Grupo de Pesquisa Metodologias em Ensino e Aprendizagem em Ciências
摘要：The study of the relationships between the econometric components of coffee production, estimated by correlations, for example, is of great relevance. These metrics provide useful information for the decision process in the production chain of this commodity. However, the quantification and interpretation of the correlation’s magnitude do not imply direct and indirect effects applicable to the agribusiness reality. In this context, trail analysis presents as a viable alternative. The objective of this work was, through trail analysis, to determine the direct and indirect effects of econometric components on coffee production. The data used are from coffee producing municipalities in Minas Gerais, in the period from 2008 to 2013, in which coffee production was observed as the basic (dependent) variable and as independent (explanatories) variables the harvested area of the grain, the average age of the workers in the field, the average remuneration of workers in the activity, the price paid for the product and the number of producing properties per municipality. There was a strong variation with year and municipality effects, with the rest of the explanation referring more to the primary variables (planted area and number of properties in the municipality), trail coefficients 0.38 and 0.05, respectively. The other variables interfere indirectly, through the modification of both. Path analysis proved to be useful in elucidating part of the coffee production chain variability and can be used as an aid in making business decisions in the sector.
