A well-designed experiment should have which of the following characteristics?

1. Subjects assigned randomly to treatments
2. A control group or at least two treatment groups
3. Replication

A good treatment should have randomization. Two groups are often designated in experiments. An experiment should be able to be replicated.