Selecting the proper experimental design depends on the situation presented and the experimental units given. You need to be able to identify when it is appropriate to utilize a certain type of experiment. Make sure you understand the differences between the types of experiments and how to create the treatment groups randomly.
While completely randomized designs are the simplest statistical designs for experiments, there are times when the simplest method for yielding precise results is not completely randomized. There are three main designs that you can use after selecting a sample for your experiment. Don't mix up language of experiment and the language of sample surveys!
Completely Randomized Design - Experimental units are randomly assigned to treatments equally by chance.
Blocking Design - Sort groups of individuals that are known to be similar in some way that can expect different results. Do not randomize those groups together. USE ALL BLOCKS in a blocked experiment.
Matched Pairs Design - A common type of block for comparing two treatments with similar individuals or identical treatment methods.