Abstract

ESPINOZA ROMANO, Vivian. Construction and analysis of the path coefficients. RevActaNova. [online]. 2018, vol.8, n.4, pp.517-535. ISSN 1683-0789.

The present article contains an explanation of the Path coefficients, from a mathematical-statistical point of view. This method is quite useful to determine effect-cause relationships, it consists of performing a statistical analysis of cause and effect in variables that are correlated, its analysis aims to express a dependent variable " y" as a function of direct and indirect effects of variables independent "x_¡". When studying the relationships that exist between a dependent variable 'y' and a set of independent variables 'x_¡', a regression and / or correlation analysis is generally used, however, the analysis of path coefficients is a method that allows to analyze the interdependence between these variables; using in this way, the regression and correlation, only in a complementary way. For example, the indirect selection of variables related to a response variable requires the identification of simple characteristics highly associated with the dependent variable. This identification is usually based on the correlation analysis; which determines an index (correlation coefficient) or reference about the relationship between variables, but this analysis is restricted in the sense that it only provides information between variables one by one it means that, it is information between pairs of variables, so, many characteristics that apparently have no relation with the dependent variable, is due to the fact that the effects of the independent variables are not direct; but they are related indirectly and the analysis of path coefficients, is a very useful technique to determine these effect-cause relationships and the magnitude of said coefficients; they precisely provide information on the relationship, based on direct and indirect effects. This method is well known in the field of agronomy and it was already used in several crops, so, as an example its application will be briefly on the crops of the ajipa tuber (pachiryzus ajipa). Then, a regression analysis is performed first on the yield of pods in ajipa crops, rescuing the most statistically significant variables (at level of 0.05), and subsequently the correlation analysis with the remaining variables is applied since it will help in the final interpretation, corroborating the existing relationships and in this way validating the interrelation through the path analysis. The results offered by this analysis will be more precise at the application of the problem.

Keywords : Path coefficients; correlation; dependence; regression; techniques of interdependence.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )