David L. Sallach

If social models are to be used to inform policies, their validation is a vital topic. However, the validation techniques of the natural sciences are not sufficient to validate social models. The present paper considers the concept of contextual validation and how it can address these additional requirements. The discussion illustrates how contextual validation is applicable to both substantive and computational social science research, and how propensity theory can be used to achieve its benefits.

Download pdf