ABSTRACT
An experiment that is inadequately designed is severely limited in what it can tell us about the world. Even worse, however, is that strong conclusions drawn from such an experiment may confuse and mislead the experimenter as to the next appropriate steps in the research program. Of primary concern in experimental testing is the "typology trap," in which the one stimulus is assumed to be representative or typical of all other stimuli from the same class. In the most extreme form of this problem, multiple responses from many subjects to one stimulus are used in statistical testing as if each response were actually to a different stimulus from the same class. This chapter illustrates that the hypotheses actually tested have frequently not been the ones that have been intended. For experimental research in which the responses of animals are tested with various types of stimuli, sampling designs are often inappropriate to test intended hypotheses.
