11. Experimental Design
← 236 | 237 → CHAPTER ELEVEN
We have now stumbled upon another area in which infants have demonstrated their heightened instincts for research, the experimental design. We have to protect infants from their unbridled propensity for experimentation. They have an insatiable curiosity to discover “what would happen.” To protect them, we have to cover electric outlets, make sure sharp objects are out of reach, and lock away hazardous substances. Have you ever asked yourself, “What would happen if...?” taken some action, and then observed the consequences? If you have, then you have conducted an experiment. An experiment involves manipulating a treatment condition or initiating a cause and systematically observing the consequences. Experimentation is natural. It involves the researcher taking an action and observing the consequences in a systematic way. This chapter explores the basic types of experimental designs and assesses their strengths and limitations. It also explores threats to the internal and external validity of experimentation and how to avoid them.
The basic purpose of an experiment is to identify and explain cause-and-effect relationships. Experiments are best for topics that have well-defined concepts and are the best-suited method for hypothesis testing and identifying causation. Do you remember the criteria for causation mentioned in Chapter 1? Experimental designs are the best-suited methods for establishing causation because they are the most likely to meet all three of the major criteria for establishing causation. The studies described at the beginning of this chapter identify story...