ABSTRACT
Failure to completely randomize an experiment does not mean the randomization concept is ignored. Rather, complete randomization is replaced with separate randomizations within and between certain treatment combinations. Such separate randomizations sometimes occur naturally, e.g., with factors like days or plots that represent a time or location component, and sometimes are used to make the experiment quicker, easier, or less expensive to run. This chapter explores how to identify restrictions on randomization. It discusses how to modify the mathematical model to account for restrictions on randomization and how the restrictions affect the analysis of the data. A mathematical proof of the effect of restrictions on randomization is included for those having a strong statistical background. The proof can be skipped without loss of understanding of the principle. The chapter summarizes the conditions under which randomization is and is not important.
