too good a psychologist to attempt to "sell" this idea to an academic audience.

In a time when many things seem uncertain, and there are some reasons for grave anxiety, hope and courage are found in the idea of the university, a center of research, of scientific idealism, of professional pride, and of loyalty to the community entering the field of public health. Here men and women are to be trained to serve their fellows, to help to bring in a better social order in which health shall be interpreted in ever wider and nobler ways. We cling still to the dream of Francis Bacon, the vision of a people served by a brotherhood of scholars who give themselves gladly that knowledge may enrich and bless the lives of all. May the American university strive always to deserve the verdict of the citizens of the New Atalantis upon Salomon's House, "the noblest foundation, as we think, that was ever upon the earth, and the lantern of this kingdom. GEORGE E. VINCENT

THE MEASUREMENT AND UTILIZA-
TION OF BRAIN POWER IN
THE ARMY, II.

Military Applications of Mental Ratings. -The sample distribution curves of Fig. 1 indicate the value of mental ratings for the identification and segregation of different kinds of military material. The illiterate group of this figure was examined by means of Beta, all other groups by means of Alpha.

Comparison of various military groups distinguished from one another by actual attainment in the service shows that the psychological tests discriminate between these groups with definiteness. This point may be illustrated by reference to the percentages of men of different groups making A and B grades in Examination Alpha: officers, 83.0 per cent.; officers' training

school candidates, 73.2 per cent.; sergeants, 53.4 per cent.; corporals, 39.7 per cent.; literate privates, 18.8 per cent. The comparison of measures of central tendency reveals equally striking differences. Moreover, within the officer group itself significant differences appear for different branches of the service.

The relation of success or failure in officers' training schools to intelligence ratings is exhibited by Fig. 2, in which it is to be noted that elimination through failure in the school increases rapidly for ratings below C. Of men rating above C+, 8.65 per cent. were eliminated; of those below C, 52.27 per cent. The data for this figure were obtained from three schools with a total enrollment of 1,375

men.

Similarly Fig. 3 shows the relation between success or failure in non-commissioned officers' training schools and intelligence ratings. The elimination increases rapidly for grades below C +. Of men rating above C, only 18.49 per cent. were eliminated; of men rating below C, 62.41 per cent. The results presented in this figure were obtained from four schools with a total enrollment of 1,458 men.

Increasingly extensive and effective use has been made of the psychological rating as an aid in the selection of men for officers' training schools, non-commissioned officers' training schools and other lines of training or service which require special ability. It has been convincingly demonstrated that the data of psychological examinations can readily be used to diminish the necessary elimination during training and thus to increase the efficiency of the schools.

The extreme differences in the intellectual status of army groups are fairly indicated by Fig. 4, which presents the data for groups whose military impor

FIG. 4. Proportion of low, average and high grade men in typical groups.

Numerous varieties of evidence indicate the extreme military importance of the prompt recognition of low grade men. The percentages of men ranking below the average in psychological examinations are notably large for the disciplinary group, men having difficulties in drill, men reported as "unteachable" and men designated by their officers as "poorest" from the standpoint of military usefulness.

The comparison of negro with white recruits reveals markedly lower mental ratings for the former. A further significant difference based on geographic classification has been noted in that the northern negroes are mentally much superior to the southern.

The relation between officers' judgments

Officers'

Rating

Very

Poor

28

Poor

51

Fair

70

Good

75

Best

99

FIG. 5. Median intelligence scores of groups designated as "best, ," "good," "fair," "poor" and "very poor'' in military value.

group is 374. The men were selected from twelve different companies, approximately thirty men in each company being ranked by an officer in serial order from "best" to "poorest. The rank order for each company was then correlated by the psychological examiner with the rank order supplied by psychological examination. In seven of the twelve companies the correlations ranged from .64 to .75. The average correlation was .536. These correlations are high, considering the large number of factors which may influence a man's value to the service.

The median score for the "very poor"

group of Fig. 5 is 28 points in an examination whose maximal score is 212 points. By contrast with this, the median score of the "best" group of privates is 99 points. ! The commanding officers of ten different. organizations, representing various arms of the service, in a certain camp were asked to designate (1) the most efficient men in their organizations, (2) the men of average ability and (3) men so inferior that they are "barely able" to perform their duties.

The officers of these organizations had been with their men from six to twelve months and knew them exceptionally well. The total number of men rated was 965, about equally divided among "best," "average," and "poorest." After the officers' ratings had been made, the men were given the usual psychological test.

D-D CC- C C+ B A FIG. 7. Men of "low military value" (····) (147), compared with a complete draft quota 12,341, (-).

Comparison of test results with officers' ratings showed:

(a) That the average score of the "best" group was approximately twice as high as the average score of the "poorest" group.

(b) That of men testing below C-, 70 per cent. were classed as "poorest" and only 4.4 per cent. as "best." (c) That of men testing above C+, 15

per cent. were classed as "poorest" and 55.5 per cent. as "best."

(d) That the man who tests above C+ is about fourteen times as likely to be classed "best" as the man who tests below C

(e) That the per cent. classed as "best" in the various groups increased steadily from 0 per cent. in D- to 57.7 per cent. in A, while the per cent.

classed as "poorest" decreased steadily from 80 per cent. in D- to 11.5 per cent. in A.

In an infantry regiment of another camp were 765 men (regulars) who had been with their officers for several months. The company commanders were asked to rate these men as 1, 2, 3, 4 or 5 according to "practical soldier value," 1 being highest and 5 lowest. The men were then tested, with the following results:

(a) of 76 men who earned the grade A or B, none was rated "5" and only 9 were rated "3" or "4."

(3) Of 238 "D" and "D-"men, only
one received the rating "1," and only
7 received a rating of "2."
(c) Psychological ratings and ratings of
company commanders were identical
in 49.5 per cent. of all cases. There
was agreement within one step in 88.4
per cent. of cases, and disagreement of
more than two steps in only .7 per
cent. of cases.

Fig. 6 exhibits a striking contrast in the intelligence status and distribution of "best" and "poorest" privates. The personal judgment data for this figure were obtained from sixty company commanders who were requested to designate their ten "best" and their ten "poorest" privates. Of the "poorest," 57.5 per cent. graded D or D-; less than 3 per cent. graded A or B. The results suggest that intelligence is likely to prove the most important single factor in determining a man's value to the military service.

In one training camp excellent opportunity was offered to compare a group of soldiers selected on the basis of low military value with a complete draft quota. In the "low value" group there were 147 men, in the complete draft quota 12,341 men. The distributions of intelligence ratings for these two military groups appear

as Fig. 7, from which it is clear that if all men with intelligence ratings below Chad been eliminated, the "low value" group would have been reduced by at least half.

In a certain training camp 221 inapt soldiers, belonging to a negro regiment of Pioneer Infantry, were referred by their commanding officer for special psychological examination. Nearly one half (109) of these men were found to have mental ages of seven years or less. The army nevertheless had been attempting to train these men for military service. In justice to the Psychological Service it should be stated that these negroes had been transferred from camps where there were no psychological examiners. For this reason they had not been examined before being assigned to an organization for regular training.

In another instance some 306 soldiers from organizations about to be sent overseas were designated by their commanding officers as unfit for foreign service. They were referred for psychological examination with the result that 90 per cent. were discovered to be ten years or less in mental age, and 80 per cent. nine years or less.

It has been discovered that when soldiers are assigned to training units without regard to intelligence, extreme inequalities in the mental strength of companies and regiments appear. This fact is strikingly exhibited by Figs. 8 and 9, of which the former shows the proportions of high grade and of illiterate or foreign soldiers in the various companies of an infantry regiment. Compare, for example, the intelligence status of C and E companies. The former happens to have received only 3 per cent. of A and B men along with 38 per cent. of illiterates and foreigners, the latter received by contrast 29 per cent. of high grade men with only 9 per cent. of men who are as a rule difficult to train. It is needless to at