On the scope of experiments in economics : comments on Siakantaris