With an emphasis on the problems of control of extraneous variables and threats to internal and external validity, the arrangement or design of experiments is discussed. The purpose of experimentation in an educational institution, and the principles governing true experimentation (randomization, replication, and control) are presented, as are three pre-experimental, three true experimental, and ten quasi-experimental designs with the advantages, disadvantages, and appropirate statistical treatments for each. This discussion follows Campbell and Stanley's chapter on experimental design in Gage's "Handbook of Research on Teaching". (MC)