Ebook: Statistical Procedures for Agricultural Research

2nd ed. — John Wiley & Sons, Inc. 1984. — 680 p. — ISBN: 978-0-471-87092-0.For most agricultural research institutions in the developing countries, the presence of trained statisticians is a luxury. Of the already small number of such statisticians, only a small fraction have the interest and experience in agricultural research necessary for effective consultation. Thus, we feel the best alternative is to give agricultural researchers a statistical background so that they can correctly choose the statistical technique most appropriate for their experiment. The major objective of this book is to provide the developing- country researcher that background.
This is a second edition of an International Rice Research Institute publication with a similar title and we made extensive revisions to all but three of the original chapters. We added four new chapters. The primary emphases of the working chapters are as follows:
Chapters 2 to 4 cover the most commonly used experimental designs for single-factor, two-factor, and three-or-more-factor experiments. For each design, the corresponding randomization and analysis of variance procedures are described in detail.
Chapter 5 gives the procedures for comparing specific treatment means: LSD and DMRT for pair comparison, and single and multiple d.f. contrast methods for group comparison.
Chapters 6 to 8 detail the modifications of the procedures described in Chapters 2 to 4 necessary to handle the following special cases:Experiments with more than one observation per experimental unit
Experiments with missing values or in which data violate one or more assumptions of the analysis of variance
Experiments that are repeated over time or siteChapters 9 to 11 give the three most commonly used statistical techniques for data analysis in agricultural research besides the analysis of variance. These techniques are regression and correlation, covariance, and chi-square. We also include a detailed discussion of the common misuses of the regression and correlation analysis.
Chapters 12 to 14 cover the most important problems commonly encountered in conducting field experiments and the corresponding techniques for coping with them. The problems are:Soil heterogeneity
Competition effects
Mechanical errorsChapter 15 describes the principles and procedures for developing an appropriate sampling plan for a replicated field experiment.
Chapter 16 gives the problems and procedures for research in farmers’ fields. In the developing countries where farm yields are much lower than experiment-station yields, the appropriate environment for comparing new and existing technologies is the actual farmers' fields and not the favorable environment of the experiment stations. This poses a major challenge to existing statistical procedures and substantial adjustments are required.
Chapter 17 covers the serious pitfalls and provides guidelines for the presentation of research results. Most of these guidelines were generated from actual experience.