Finally, another indirect scheme to determine the personal qualities needed for economic efficiency may be suggested by the psychology of the typical correlations of human traits. We have seen that group psychology proclaims that a certain individual probably has certain traits because he belongs to this or that nation or to this or that otherwise well-known group. Correlation psychology proclaims that a particular individual possesses or does not possess certain traits because he shows or does not show some other definite qualities. A correlation, for instance, which the commercial world often presupposes, may exist between individual traits and the handwriting. Graphologists are convinced that a certain loop or flourish, or the steepness or the length of the letters, or the position of the _i_ dot, is a definite indication that the writer possesses certain qualities of personality; and if just these qualities are essential requirements for the position, the impression of the handwriting in a letter may be taken as a sufficient basis for appointment. The scientist has reason to look upon this particular case of graphological correlation with distrust. Yet even he may acknowledge that certain correlations exist between the neatness, carefulness, uniformity energy, and similar features of the letter, and the general carefulness, steadiness, neatness, and energy of the personality. However, the laboratory psychologists nowadays have gone far beyond such superficial claims for correlations of symptoms. With experimental and statistical methods they have gathered ample material which demonstrates the exact degree of probability with which we have a right to expect that certain qualities will occur together. Theoretically we may take it for granted that those traits which are always present together or absent together ultimately have a common mental root. Yet practically they appear as two independent traits, and therefore it remains important to know that, if we can find one of them, we may be sure that the other will exist there too. Inasmuch as the one of the two traits may be easily detected, while the other may be hidden and can be found out only by long careful tests, it would be valuable, indeed, for the employment manager to become acquainted with such correlations as the psychologist may discover: as soon as he becomes aware of the superficially noticeable symptom, he can foresee that the other disposition is most probably present. To give an illustration: in the interest of such measurements of correlations we have studied in the Harvard laboratory the various characteristics of attention and their mutual dependence.[18] We found that typical connections exist between apparently independent features of attention. Persons who have a rather expansive span of attention for acoustical impressions have also a wide span for the visual objects. Persons whose attention is vivid and quick have on the whole the