When we take into account that by the processes of selection and elimination only thirty to forty per cent of the pupils who enter the elementary school ever reach high school,[49] it is readily admitted that the high school population is a selected group, of approximately 1 in 3. Then of this number we again select less than 1 in 3 to graduate. This gives a 1 in 9 selection, let us say, of the elementary school entrants. For relatively few general purposes in life may we expect to find so high a degree of selection. Yet in this 1 in 9 group (who graduate) the percentage of the failing pupils is as high as that of the non-failing ones, and the percentage of graduates does not drop even as the number of failures rise. So far as ability is required to meet the conditions of graduation they are manifestly provided with it. Following this comparison still further, the failing pupils who do not graduate have an average number of failures that is only .6 higher than for the failing graduates (4.9-4.3); but barring those non-graduates considered in section 1 of this chapter, the average is practically the same as for the failing graduates. Moreover, the failing non-graduates continue in school, even in the face of failure, much longer than do the non-failing non-graduates. That gives evidence of the same quality to which the manager of a New York business firm paid tribute when he said that he preferred to employ a high school graduate for the simple reason that the graduate had learned, by staying to graduate, how to 'stick to' a task.