ECHO Technical Notes TN #50 Statistical Analysis of Simple Agricultural Experiments
Use of a new crop variety or production technique may dramatically increase food production in a given area. Alternatively, an innovation successful in North America may utterly fail in the tropics. The goal of “adaptive research” is to evaluate a particular innovation for its usefulness
under local conditions. This technical note is written for those who want to improve the quality of their experiments but who have little or no background in statistics. It supplements the 81st issue of ECHO Development Notes with step-by-step instructions on how to manually calculate
statistics for the most commonly used experimental designs. A little persistence, very basic math skills, and perhaps a calculator are all you need to do the calculations. If you have a computer equipped with statistical software, doing a set of calculations by hand or with a calculator will
help you to understand how to use the software and interpret the output. Examples are given for experiments in which only one factor (e.g. crop variety) is tested. A limited amount of statistical terminology is woven into the text for the benefit of those interested in further study.