Science teachers ought to examine closely the common practice of asking students to formulate a conclusion at the end of every five-minute demonstration or forty-minute experiment. These activities concern but a limited sample of all the phenomena governed by the principle under consideration. At the end of these experiences students should have the opportunity to choose among formulating a generalization with various qualifications, stating that they have only learned something about the particular operation at hand, or stating that they could make no sense of the data.