The essence of a "Randomized Block" design is that the experimenter forces the randomization to be done within certain groups of experimental units. Thus, if we were looking at students in the Math 169 class this term and we wanted to try two different presentations of algebraic factoring to the students. We could try a completely randomized approach (namely, number the students and then for each student randomly select the presentation that they will be given). However, because that is completely randomized, we might be concerned that there may be an over representation of daytime students getting one of the presentations. If we are so concerned that day/evening split may be a confounding factor, we could adjust our design to force an even representation across a block of day students and a block of evening students. For example, we could take some number of day sections and some number of evening sections and then, within each section randomly divide the students into two groups, one for each of the two presentations. By doing this we will be sure that there are essentially an equal number of day students getting each presentation and, likewise, an equal number of evening students getting each presentation. The design forces this distribution. Again, the aim is to eliminate the concern that the suspected difference between day and evening students might confound the experiment if we were to use a completely randomized design.