This paper outlines why program evaluation should follow well-respected scientific standards and why it should be performed by independent researchers.