Critics of social science have long pointed out–and social scientists regretted–the difficulty of doing proper controlled experiments in their field.