In this paper, we provide a reconstruction and elaboration of Sismondo's constructivist view on experimental truths. Some unusual features of the view are revealed. Furthermore, by indicating how to solve some potential challenges to Sismondo's constructivist view on experimental truths, we argue that this view is a natural consequence of interventionism developed from recent works on the metaphysics of structural equation modeling.