Definition

Internal and external validity

Internal validity is whether the study really shows what it claims within its setting; external validity is whether the result generalizes beyond it.

Internal validity asks: could anything other than the manipulation explain the result — confounding, selection, attrition, measurement artifacts? External validity asks: would it hold with other people, settings, times, and operationalizations?

The two trade off: the laboratory control that maximizes internal validity strips away the ecology that supports generalization. Strong programs of research triangulate — tight experiments plus field studies — rather than optimizing one study for both.