The Duhem‐Quine thesis and experimental economics: A reinterpretation