Simple Comparative Experiments

This chapter looks at a method for making simple comparisons of two or more "treatments", such as varying brands of toothpaste, and testing their efficacy for preventing cavities. The F-test compares the variance among the treatment means versus the variance of individuals within the specific treatments. The F-test is a vital tool for any kind of design of experiments, not just simple comparisons, so it’s important to understand as much as one can about it. The F-test uses variance as its underlying statistic. Using least significant difference without the protection of a significant F-test could be a real “downer” because of the risk of getting false positive differences. The famous child psychologist, Jean Piaget, suggested that thinking is a flexible, trial-and-error process. He observed that children, at their earliest stage of development, experiment with objects and learn from the experience.