In order to obtain reliable estimates, generated through path analysis, it is important to define the necessary number of experiments. The objective of this work was to determine the number of experiments necessary for the path analysis the number of pods per plant, number seed per pod and weight of 100 grains on grain yield the common bean cultivars (Phaseolus vulgaris L.). Thirteen common bean cultivars were evaluated in nine experiments conducted in Santa Maria, in Rio Grande do Sul State, Brazil, in the growing seasons between 2000/2001 and 2004/2005. We measured the characters grain yield, number of pods per plant, number seed per pod and weight of 100 grains. Were planned 511 data matrices (13 cultivars and four characters). Analyses of correlation, diagnoses of multicollinearity, path analysis and scatter plots were carried out. The number of experiments necessary for the path analysis, was determined by the estimates of the parameters of the quadratic response plateau model. In the central depression region of Rio Grande do Sul State, Brazil, seven experiments are sufficient for the path analysis in common bean.
Phaseolus vulgaris L.; correlation analysis; quadratic response plateau model
