A COMPARATIVE STUDY OF THE STATISTICAL METHODS MOST COMMONLY USED IN AGRICULTURAL RESEARCH

Abstract

This paper comprises a brief and simple explanation of the fundamental principles underlying the use of the statistical methods most commonly used in agricultural research. Among the methods compared are Bessel's method, the deviation from the mean method, Fisher's method for the Analysis of Variance (including the following phases: Unrestricted method, method of Randomized Blocks, method of Latin Square), and "Student's" method for Paired Observations. An example for the application of each method and new tables devised to test exclusively whether a difference is statistically significant or not are also presented.