A hypothesis that has received much support. In practice, theory is often used interchangeably with the word “hypothesis.” Theory can be a dangerous term because it is often misused to mean “complicated and obscure arguments.” Also, “theory” is often added to a paper after the study has been completed. A good theory should have predictive validity. To demonstrate how to test the predictive validity of theories, Armstrong (1991) examined theories about consumer behavior. In the Journal of Consumer Research, authors generally begin their papers by describing theories. Knowledge of such theories should lead one to make better forecasts. Sixteen academics in this field, presumably familiar with the theories, were asked to predict the outcomes of 20 studies with 105 hypotheses. All of these studies had been published in the Journal of Consumer Research, but the academics in the sample reported that they could not remember seeing them. As it turned out, their predictions were less accurate than those made by 43 high school students. Thus, contrary to the hypothesis, academic theories in consumer behavior did not have predictive validity.