| Intelligence | Humor | Perseverance | Kindliness | Conceit | Courage | |
| Humor | .47 | |||||
| Perseverance | .88 | .33 | ||||
| Kindliness | .76 | .65 | .39 | |||
| Conceit | .28 | -.03 | .08 | -.56 | ||
| Courage | .89 | .43 | .79 | .72 | -.25 | |
| Deceitfulness | -.11 | -.28 | -.02 | -.69 | .66 | -.49 |
It will be seen that the intelligent, humorous, persevering, kindly, and courageous countenances tend to be the same ones, and that the faces suggesting the opposites or low degrees of these traits also tend to be very much the same ones. This is indicated by the high positive coefficients between these traits. But conceit and deceitfulness show negative or very low positive correlation with all traits except each other. In this latter case the correlation is positive and high (.66). Other interesting relations between these judgments of character can be inferred from the table of coefficients. But it should be remembered that we are not here dealing with traits as demonstrably present, but only as judged on the basis of facial characteristics and expression. The actual relation between the physiognomic details and the true character of the individual displaying them is a totally different matter. The close correlations between the several desirable traits and between the several undesirable traits, as found in this table of coefficients, seem to have a further significance and suggest that the observers do not judge each trait on the basis of particular and specific physiognomic details. They seem, rather, to get a general impression of favorableness or unfavorableness, and to rank the photographs on the basis of this general impression, no matter which trait is being judged.
It is a common practice for employers, superintendents, agencies, etc., to request the applicant for a position to send his or her photograph for inspection. The urgency of some of these requests and the emphasis placed on them seem to indicate that the photograph is believed to be valuable not only for its service in revealing the general features but also for some further and more specific indications which it affords. Very few attempts seem to have been made to test actually the value of judgments of character when they are based on photographs rather than on acquaintance. Experiments recently conducted yield some interesting preliminary data on this question. The question proposed was: "What relation exists between the judgments which strangers form, on the basis of an individual's photograph, and the judgments which acquaintances make on the basis of daily familiarity and long observation?"[5]
All the members of a group of college women were judged by twenty-four of their associates, for a number of more or less definite characteristics. The twenty-five individuals constituting the group were arranged in an order of merit for each trait, by each of the twenty-four judges. Only one arrangement, for one trait, was made by any one judge within a given week. The judgments were thus distributed over a considerable interval so that judgments for one trait might influence as slightly as possible the judgments of later traits. All these twenty-four judgments were then averaged for each trait, and the final position of each person in each trait thus determined by the consensus of opinion of the judges. This measure is then a combined estimate on the basis of actual conduct and behavior.
Photographs of all the members of the group were then secured, all of them taken by the same photographer, in the same style and size. These photographs were now judged, by a group of twenty-five men and a group of twenty-five women, all of whom were totally unacquainted with the individuals who were being judged. These strangers arranged the photographs in order of merit for the various traits of character, just as the earlier group of judges had arranged the names of the members of the group, with all of whom they were acquainted. The various arrangements of the photographs were then averaged, yielding for each photograph an average position in each trait. We thus have three measures of the group of college women: (1) the judgments of their intimate associates; (2) the judgments of twenty-five men, on the basis of photographs, and (3) the judgments of twenty-five women, on the basis of photographs. All of these measures may be compared with each other, and correlated so as to show their respective amounts of correspondence. The results are as follows:
| Trait | Judgments by Associates Compared with the Judgments of the Photographs | ||
| By 25 Men | By 25 Women | Average | |
| Neatness | .03 | .07 | .05 |
| Conceit | .10 | .27 | .19 |
| Sociability | .29 | .29 | .29 |
| Humor | .21 | .45 | .33 |
| Likeability | .30 | .45 | .38 |
| Intelligence | .42 | .61 | .51 |
| Refinement | .50 | .52 | .51 |
| Beauty | .60 | .49 | .55 |
| Snobbishness | .58 | .53 | .56 |
| Vulgarity | .61 | .69 | .65 |
| Average | .36 | .43 | .40 |
The correspondence between judgments of acquaintance and judgments of photographs is seen to vary with the trait in question. Such traits as neatness, conceit, sociability, humor, and likeability, important as they are for vocational success or failure, show very low correlation. The judgments of the photographs tell almost nothing at all of the nature of the impression which the individual makes on her acquaintances, her true character. With the remaining traits—beauty, intelligence, refinement, snobbishness, and vulgarity—the coefficients are considerably larger, and suggest that the photographs tend to be judged by the strangers in somewhat the same way as the individuals are judged by their acquaintances.
Two points of special importance should be noted in this connection. The first is that these correlations are not between the judgments of single individuals. It is the combined or group judgment of twenty-five judges which is required to yield these coefficients which even then average only about .40 correlation with the estimates of associates. The following table shows the ability of ten judges, chosen at random, to estimate these characteristics through the examination of the photographs. In securing this table the arrangement made by each individual judge was correlated with the established order as determined by the estimates of associates, in the case of the three traits—intelligence, neatness and sociability.
| Judge | Individual Correctness of Judges in Estimating | ||
| Intelligence | Neatness | Sociability | |
| I | .51 | .11 | .39 |
| II | .11 | .10 | .08 |
| III | .15 | .29 | .05 |
| IV | -.27 | .06 | .49 |
| V | .08 | .24 | .08 |
| VI | .43 | .41 | .28 |
| VII | .04 | .11 | .02 |
| VIII | .39 | -.09 | .32 |
| IX | .22 | -.08 | .00 |
| X | .30 | .02 | .55 |
| Average | .19 | .11 | .22 |