Toward a Psychology of Science: A Dynamic Cognitive Model of the Evolution of Scientific Theories.

Responding to a recent call for the development of a psychology of science, we examine how scientists manage empirical anomalies, a routine yet consequential challenge in scientific practice, and how their responses shape theory development. We argue that defensive anomaly-reduction strategies, which allow researchers to preserve an initial theoretical claim, can generate important secondary benefits. During the emergence phase of a theory, they foster auxiliary revisions that may reduce underfitting and place a fragile theory on a growth trajectory. During the exploitation phase of established theories, they help sustain attention to problems derived from the theory and support cumulative research. However, their accumulation may also increase the risk of overfitting, reduce predictive power, and set the theory on a degenerative trajectory. We illustrate these dynamics primarily through examples drawn from psychological theories, with a focus on responses to replication failures. Last, we derive several implications for peer review aimed at fostering scientific innovation. In particular, we propose shifting evaluation away from the consistency of findings, which may penalize emerging research programs and hinder innovation, and toward how inconsistencies are managed, thereby fostering growth trajectories, including for mature programs confronted with anomalies.