Most of the data discussed in this volume are based on experiments; most of the illustrations and examples, however, are derived from current social problems—including prejudice, propaganda, war, alienation, aggression, unrest, and political upheaval. This duality reflects two of my own biases—biases that I cherish. The first is that the experimental method is the best way to understand a complex phenomenon. It is a truism of science that the only way to really know the world is to reconstruct it. That is, to truly understand what causes what, we must do more than simply observe; rather, we must be responsible for producing the first “what” so that we can be sure that it really caused the second “what.” My second bias is that the only way to be certain that the causal relations uncovered in experiments are valid is to bring them out of the laboratory and into the real world. (Location 85)
Although, like the amateurs, we usually begin with an observation, we do not stop there. We do not need to wait for things to happen so that we can observe how people respond; we can make things happen. We can conduct an experiment in which we subject scores of people to particular events (Location 271) Moreover, we can do this in situations in which everything can be held constant, except the particular factors being investigated. We can then draw conclusions based on data that are more precise and numerous than those available to the amateur, who must depend on observing events that occur randomly and under circumstances in which many things are happening at once. (Location 274)