ABSTRACT
In analysis of covariance, the techniques of analysis of variance and re gression analysis are combined. Reliable data must be obtained from a suit able design in order to make treatment comparisons with greatest precision. In any design, in addition to treatments, the observations are affected by several unknown factors. These factors are the main reasons for experimen tal error. To some extent, they are controlled or minimized by proper selec tion of the experimental design, and by the application of the basic princi ples of experimental design, namely, replication, randomization, and local control. The precision of the estimates can further be enhanced by using confounding, split plot, strip plot, and incomplete block techniques.
