Steve Levitt and John List, writing about behavioural economics in Science:
Most of this research eschews a narrow conception of rationality, while continuing to embrace precisely stated assumptions that produce a constrained optimization problem. A less “scientific,” and in our view less productive line of research in this area approaches the problem from the opposite direction: Observing an unexpected pattern of behavior (e.g., lower stock markets on rainy days in New York City), one looks for a psychological theory consistent with that behavior (in this case, seasonal affective disorder). Given the wide array of psychological explanations from which to choose, however, a researcher undertaking such a task has virtually unlimited freedom to explain any observed behavior ex post facto.
Perhaps the greatest challenge facing behavioral economics is demonstrating its applicability in the real world. In nearly every instance, the strongest empirical evidence in favor of behavioral anomalies emerges from the lab. Yet, there are many reasons to suspect that these laboratory findings might fail to generalize to real markets. We have recently discussed several factors, ranging from the properties of the situation–such as the nature and extent of scrutiny–to individual expectations and the type of actor involved. For example, the competitive nature of markets encourages individualistic behavior and selects for participants with those tendencies. Compared to lab behavior, therefore, the combination of market forces and experience might lessen the importance of these qualities in everyday markets.
Recognizing the limits of laboratory experiments, researchers have turned to “field experiments” to test behavioral models. Field experiments maintain true randomization, but are carried out in natural environments, typically without any knowledge on the part of the participant that their behavior is being scrutinized. Consequently, field experiments avoid many of the important obstacles to generalizability faced by lab experiments.