scored by me, the usual tabulations. made, and the profiles constructed. With all identifying information except age eliminated, they were then arranged in random order. Two clinicians, who are also experts in Ror. schach, analyzed each of the 60 proto cols separately in this order. Because of the importance of knowing how, by what process, using what evidence in the Rorschach, a judge arrived at his rating or judgment in each of the categories, each judge was urged to describe as much as he could of the procedure he was using, the conclu sions arrived at, and the evidence used; and the whole process was re corded by Audograph" Let it be said here that the task which the judges were asked to perform, that of analyzing 60 records in succession and of verbalizing the whole process, was a monumental one It demanded not only a devotion to science "beyond the call of duty" but also an admir. able willingness to expose one's falli. bility. My success in persuading Dr. Klopfer and Dr. Mortimer Meyer, for the Rorschach, and Dr. Shneidman, for the TAT and MAPS, to give so generously of themselves in this project was primarily due to their belief in its importance and to their eager ness to see a unique body of material and to engage in what they antici. pated to be a rewarding learning experience.

The purpose of the Rorschach analysis was two-fold: (1) to obtain an unbiased judgment (that is, without knowledge of homosexual or heterosexual identification of subjects and without life-history materials) of per. sonality structure and overall adjust ment of the subjects in both groups; (2) to determine the accuracy with which expert clinicians who are Rorschach workers can differentiate homosexual from heterosexual records. Each judge was asked, in addi. tion to the overall adjustment rating, to analyze the Rorschach protocol in terms of a number of categories, such

as methods of handling aggression, affection and dependency needs, methods of impulse control, and clinical label, if any. These judgment categories were used because of their theoretical importance in current ap. proaches to homosexuality. The ad justment rating was on a five-point scale from 1, superior, to 5, maladjusted; with 3 representing average adjustment. adjustment. The norm which the judges used was, of course, a subjective one, of average adjustment in the population at large, not just in this group. Assigning an adjustment rat ing to a Rorschach protocol is diffi cult, as all of us know. The meanings of the five points of the rating scale were defined as follows: (1) superior, or top adjustment; better than the average person in the total popula tion; evidence of superior integration of capacities, both intellectual and emotional, ease and comfort in relation to the self and in functioning effectively in relation to the social environment; (8) as well-adjusted as the average person in the total popu lation; nothing conspicuously good or bad; (5) bottom limit of normal group and or maladjusted, with signs of pathology. Ratings 2 and 4 are self-evident, 2 being better-than-average but not quite superior, and 4 being worse-than-average, or the bottom limit of the average group. These ratings are very difficult to objectify, and it is very difficult to be sure that they were used in the same way by the two judges.

One further comment about proced. ure, before discussing the results of the judging on adjustment: each judge, before he began, knew that some records were homosexual and some were heterosexual. Most clinici. ans in the Los Angeles area are fam iliar with the project, and it would have been impossible to secure experts without some knowledge of it. The judge was told that the opportunity to distinguish homosexual from heterosexual records would

mattachine REVIEW

TABLE II-Ratings on Overall Adjustment-Rorschach

Ratings

(Top)

Group

F-nsgws|

(Bottom)

Judge "A" Homosexual. Heterosexual Total

Judge "B" Homosexual. Heterosexual. Total.

TITIT

superior adjustment for the homosexual group. By the method of "grand medians", chi square for Judge "A" is zero for the differences in adjustment between heterosexuals and homosexuals and for Judge "B" the difference is 231, which is insignifi

cant.

come later and that the present task was that of telling me as much as he could about what he thought the subject to be like in personality struc ture and adjustment. If anything impressed him about the pattern of sex. ual adjustment, he should say it, but this was not the primary purpose of this stage of the analysis. The task of the judges was broken down into two steps. (1) The protocols were anal yzed, with overall adjustment ratings given and summary judgments made, in the categories already described: and (2) each judge was then presentation ed with 30 pairs of protocols, matched for age, education, and IQ, the task being to distinguish the homosexual record in each pair.

The results of the judging of adjust ment from the Rorschach protocols are presented in Table II.

It will be noted that there are no significant differences between the number of homosexuals and hetero sexuals having a rating of 3 and better for each judge; two-thirds of each group are assigned an adjustment rating of 3 or better. There are apparent differences between judges. For Judge "B" there is a greater unwillingness to assign a top rating. In fact, for Judge "B", there is a slight but insignificant trend in the direction of

0 (exact agreement).

1 rating step............

2 rating steps.

3 rating steps.

Differences

The immediate question is the de gree of agreement between the two judges. Although a Tschuprow coef. ficient between the ratings of Judge "A" and Judge "B" is only 0.33, it is important to point out that the situ

is not as bad as this low coef ficient would seem to indicate. Table III shows that the two judges agreed exactly in 19 of the 60 cases, 8 being homosexual and 11 hetero sexual. In 23 cases they disagreed by one rating step, 12 of these being homosexual and 11 heterosexual. This means that in 42 out of the 60 cases there was either exact agreement or disagreement by only one step. So it is safe to say that in two-thirds of the total distribution there is high agreement. An additional fact that may be pointed out is that 14, or ap proximately one-half, of the homo sexuals were placed either in Adjust ment Rating 1 or 2 by both judges.

How is one to interpret this find ing? Is one to take it at face value and TABLE III

Number of Subjects Homosexual

Heterosexua 11

Total

273