This chapter introduced us to Randomization Design, which provides the scheme of how treatment levels can be assigned to experimental units. The specific designs discussed are CRD, RCBD, and Latin Square Design. An RCBD is employed to account for a blocking factor, or a nuisance variable, which is not of interest but may have an impact on the response. Likewise, a Latin square design is helpful in the presence of two such blocking variables. In an RCBD, with no replicates, the interaction between the treatment and the blocking variable is assumed to be negligible and the Mean Square(MS) value of this interaction serves as the estimate of the error variance which turns out to be the denominator of the \(F\)-statistic for testing treatment significance. The next chapter will introduce us to another widely used design called split-plot design.