Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Statistical methods of grouping means are used to classify experimental data and to separate the means in groups. The present work had as objective to develop an algorithm to separate means using the non parametric test Scott-Knott, and to implement and to validate a system capable to accomplish all the necessary stages of the test, from the analysis of variance to the separation of the means. The system was developed with Borland Delphi to platform PC using Windows 95 or above. The system was validate comparing the generated results with results calculated in a manual way and with results generated by the Duncan's and Tukey's tests. The system uses recursive functions and has a larger precision and a larger speed. When compared to the other methods of mean separation, the Scott-Knott method facilitates the interpretation of the results. The data were classified in different groups and there is not overlapping among the found groups. That overlapping is characteristic of other methods quite used as Duncan's and Tukey's methods. The developed system makes possible to use the test of Scott-Knott in the analysis of experiments with a great data set. The system is been used to vegetal genotype screening. This kind of experiments have usually more than a hundred treatments and need a good way of ranking the material.