Completely Randomized Design

Advantages

• Extremely easy to construct

• Easy to analyze, even for unbalanced data

• Can be used for any number of treatments

Disadvantages

• Best suited for relatively few treatments

• Homogeneous experimental units are needed

Randomized Block Design

Advantages

• For comparing several means with one nuisance variable

• Easy to construct

• Simple analysis

• Can be used for any number of treatments and blocks

Disadvantages

• Best for relatively few treatments

• Additional nuisance variables cause inefficiency

• Treatment effects must not vary from block to block