ABSTRACT
In a comparative experiment the number of observations on each treatment is called the number of replications. A fundamental principle of such experiments is that it is essential to carry out more than one test on each treatment in order to estimate the size of the experimental error and hence to get some idea of the precision of the estimates of the treatment effects. This chapter concerns the most important basic principle of good experimentation, namely randomization. In a simple comparative experiment the residual variation is the variation not accounted for by the treatment effects. Thus the residual of a particular observation is the difference between the observation and the average observation on that particular treatment. The technique of blocking is a very useful method of increasing the precision of comparative experiments. The randomized block experiment is by far the most important type of comparative experiment.
