In experiments for determining the ability of computing-machine operators various tests were used. One of the most important was a mental-arithmetic test. This was designed to determine the applicant’s fundamental knowledge of arithmetic. Another was a numerical substitution test. In each of the tests conducted the scores of the applicants were compared with the rankings made previously by department heads, and in most instances there was an agreement of sufficient approximation to indicate the value of the tests. Although still in its infancy, as it were, so far as its practical application in industry goes, the scientific method of mental measurement, wherever and whenever applied in accordance with true psychological principles and by standards and methods devised by trained psychologists, has so completely demonstrated its economic value and social usefulness that its general adoption, as these facts become more generally known, seems inevitable. CHAPTER IX HOW TO USE THE MENTIMETER TESTS The Mentimeter tests differ from the Alpha tests, or from the Beta test of the United States Army, from the Otis test, or from any other system of tests now available, chiefly in their flexibility. Rather than present to the public a certain fixed and invariable group of eight or ten tests which are to be used wherever a measure of general intelligence is to be employed, as has been done in other cases, the present authors have chosen to present a wide variety of tests from which each reader may select those for his use which actually give the best results. It is not probable that exactly the same tests would select men of high intelligence in the graduate work of a university as would be needed to select the intelligent men in a logging camp in the wilds of Canada or our own Northwest. The present authors do not profess to know just how much of each mental trait is required to make up a perfect superior intelligence, and for that reason they have not attempted to propose any single group of tests as the best measure of intelligence. The reader is asked to “try out” such tests in the Mentimeter series as seem to him to offer greatest promise of usefulness, and then to make up his own “team of tests” in such manner as will best reveal the kind of intelligence in which he is interested. For the benefit of those who wish some suggestions as to the tests which would probably be most useful in the main lines of work to which intelligence tests may be applied, the authors here propose certain tentative or suggestive lists which would seem to them to offer great promise of successful use. For the classification of clerical workers in business and industry, the following tests should at least be given thorough trial: