ChileanJAR

Wheat yield trials of the Carillanca Experiment Station and international yield trials,sown in different dates and locations, are analized by means of additive, linear and multiplicative models, with indication of the conditions that allows to pass from one to another. The multiplicative model implies the concept of family of curves, and permits fo replace the slopes of the regression lines of varieties on the mean yield of locations, by a value a, common to all these lines.