ABSTRACT
This chapter introduces the analysis of variance (ANOVA) techniques to compare three or more sample means simultaneously. This comparison is done to determine if any statistically significant differences exist between the means of the populations. The chapter explains when and how ANOVA is used, and the underlying assumptions. It describes the techniques used in the one-factor and two-factor ANOVA with equal or unequal observations. The chapter provides the Latin square design in agricultural research. It utilizes the ANOVA techniques in testing hypotheses concerning the means of three or more populations. The objective of the two-factor ANOVA is to isolate the effect of not one, but two variables of interest in an experiment. The two types of two-factor ANOVA are: randomized block design and completely randomized design. The Latin square design permits us to reduce the number of subjects required, and assesses the relative effects of various treatments when a two-directional blocking restriction is imposed on the experimental units.
