With the objective of describing the characteristics of the works published in the Ciência Rural between 1971 and 2000 and to identify deficiencies in the employment of statistical analysis, 1788 works were classified, being 879 of the animal area and 909 of the vegetable. Inside of each area it was obtained the simple relative frequency of the experimental desing and of the statistical tests of complemental analysis applied for qualitative and quantitative treatments and, the maximum values, minima and medium for numbers of treatments and repetitions and for the level of used error , showing that in the animal area, the profile of the works is of using 3 treatments, 4 repetitions, level of 5% of probability error, test of Tukey for qualitative treatments and analysis of simple lineal regression for quantitative treatments. Already in the vegetable area, the profile was maybe of using the delineamento in blocks to the, with 6 treatments, 5 repetitions, level of 5% of probability error, test of Duncan for qualitative treatments and analysis of simple lineal regression for quantitative treatments. In the classification of the adaptation of the tests of complemental analysis, in the animal area it was Appropriate = 63%, Inadequate = 26%, and vegetable of Approppriate = 61%, Inadequate = 25%, showing that most of the works possesses appropriate use of the tests in the complemental analyses, could give a good reliability to the results and conclusions obtained in these works.
Regression analysis; averages comparison test; ex-perimental planning; statistical analysis
