In the present state of our knowledge concerning methods discretion is needed in the selection of tests. Those primarily adapted to one group may not be valid for another social or age group. We have just finished an interesting comparison of the results of a certain performance test in which college young women did worse on the average than younger persons. We have all too little proof that tests worked up for children are equally valid for adults. It is proposed that we render decisions upon, for instance, findings by the Binet tests, and yet it does not seem likely by the sort of results we ourselves have been getting that they could be as freely applied to adults as to children for the purposes of social diagnosis, of discriminating those who are bound to be unsuccessful. We must remember that no one as yet has given us the results of these tests as applied to hundreds of ditch diggers, or section hands, who in their lowly spheres form most useful members of society. When it comes to the interpretation of tests we need to exercise much discretion. It is most dangerous to proceed to render diagnosis or prognosis without knowledge of the individual’s background in heredity, developmental history and social environment. Such items as previous illnesses, present physical condition, debilitating habits, and educational opportunities need to be noted. In our work the question of epileptic variations alone is frequently before us. We see very clearly that grave injustice to the situation may be done without taking such possible features of the case into account. Very different phases of psychological work with offenders properly are taken up from the viewpoint of human conduct in general being the province of the student of human life. Quite the minority of offenders show mental peculiarities which can be learned by testing and then related to their delinquencies. Let us look at some comparisons of offenders as we have grouped them by most careful study. Our court work has been in the main to see the problem cases. Undoubtedly most of the suspected psychotic or defective cases have been selected and brought for study. We have made a series of 1000 recidivists, repeated offenders, the average age of whom is about 15 years. These have been graded by mental tests most carefully and we have found the following: