Epistemology of Replicability

Summary

Replicability, the property of scientific results whereby they can be obtained again under similar conditions, is an indicator of the reliability, objectivity, and trustworthiness of science. It is assessed through replication studies, which may be direct, partial, or conceptual, depending on their methodological similarity to the original experiments. Key factors undermining replicability include low statistical power, misuse of methodological flexibility, failures of representativeness, and weak or poorly specified theories. Although replicability is widely regarded as essential for scientific self-correction and the production of reliable results, this status is occasionally contested. Proposed measures to promote replicability include preregistration, the use of well-powered designs, and theory formalization. Strategies for selecting and conducting replication studies will also be discussed.