Describe Freud’s approach to psychoanalysisDescribe one of the projective tests identified in the course text [starting on p. 58]Identify a peer-reviewed article (Links to an external site.)in the S
THE “INTUITION QUESTIONNAIRE”: A NEW PROJECTIVE TEST BY ALEX C. SHERRIFFS University of California T HE main purpose of the investiga- tion reported here is to develop a new method of personality meas- urement, responses to which can sta- tistically and logically be shown to depend on the subjects’ measurable projections. The writer will not attempt at this time a survey of the historical and theoretical approaches to projection and projective methods.1 The present status of projective methods best can be summed up by the word “promising.” They are promising, more because of the suggestive findings of the overall research, than because of the results of crucial experiments; they are promising, in the light of the judgments of clini- cians using the techniques with profit in individual cases, rather than in the light of high statistical validity and adequate standardization. The field of unstructured personality tests is itself highly fluid in its own structure. There seems to be a real need for more studies that will help evaluate objectively methods which utilize the projective process as a route to the understanding of personality. THE “INTUITION QUESTIONNAIRE” The Test This test is of questionnaire form and includes thirty items. Each item con- sists of a very small sample of the behavior, usually out of any objectively meaningful context, of some hypotheti- 1 The reader is referred particularly to the recent provocative reviews by Sargent (8), Hunt (12), Rapaport (7), Symonds and Samuel (n), and Cattell (2). cal individual. Each test item has been chosen to represent behavior which might arise from a variety of motiva- tions. These behaviors are ones for which college women (see “The Sub- jects”) might be expected to feel some empathy. The items of the “Intuition Questionnaire” follow. ITEMS OF THE “INTUITION QUESTIONNAIRE” 1. B. is always wanting to do something she has never done before. 2. X. feels upset if she hears that people are criticizing her or blaming her. 3. A. often acts contrary to custom, or to the wishes of her parents. 4. R. said to her brother, “I hope you never marry.” 5. J. sometimes puts almost too much time and effort into neatness. 6. W. never seems to stop and let herself think. 7. P. is easily influenced by her friends; lets herself be led. 8. E. is a girl who enjoys a good hot argument. 9. T., a twenty-year-old girl, turned down an invitation to a party in order to study for an examination to be held in a week. 10. This girl said, “I give myself utterly to the happiness of someone I love.” n. L. goes off by herself every evening and watches the sunset. 12. Ra. said “Look what I’ve done!” 13. H. is a girl who prefers hiking, fishing, and flying to parties, bridge, and talking. 14. K. is a girl who says, “I work like a slave at everything I undertake until I am satisfied with the results.” 15. V. is a girl who daydreams a great deal of the time. 126 THE “INTUITION QUESTIONNAIRE” 327 16. N. is attracted by every good-looking man that she sees, 17. S. is a girl who enjoys organizing or directing the activities of a group, team, club, or committee. 18. Q. is a girl who would rather stay home alone and listen to good music than go out with a group. 19. F. is a girl who hesitated when told to choose between school, marriage, and career. 20. M. is unusually particular about her appearance and dress. 21. O. feels happy only when she remains unnoticed. 22. D. said to her father, “Why did you marry mother?” 23. Y. said to her friend, “Gee, I’d never be able to do that!” 24. Z., nineteen-year-old, attractive-looking, announced after an eight day acquaint- ance with the second man in her life that she was about to be married. 25. I. is a girl who always seems to have a great deal on her mind. 26. N. is a girl who is uneasy when put in a position of decision or leadership. 27. L. is always interested in money and handles her money very well, keeping minute account of every cent. 28. S. is a girl who said, “I think the arts are more important to me than the sciences.” 29. J. said, “I am apt to be apprehensive when I am alone in an empty house at night.” 30. C. is the only girl in her living-group who always locks her door. On the questionnaire forms, a space of one and one-half inches was left under each state- ment for the subjects to write in their responses. The questionnaire is characterized to the subjects as a method for picking out those individuals who are able to make the most effective intuitive use of small cues for predicting a total situation—an ability supposedly related to insight. The subjects are then asked to “Give a probable explanation for the behavior indicated in each of the following excerpts from life histories taken from a random sample of the population. Include the motivation underlying the behavior and the origins of the motivation.” Since the responses of the subjects are not determined by the context of the items themselves, they must necessarily be highly individual. It further ap- pears that they are definitely related to the underlying motivational systems and perception habits of the subjects (see “Results”). Administration may be group or indi- vidual. The questionnaire requires on the average 50 minutes for completion. Variables for Scoring Investigation reveals that the “Intu- ition Questionnaire” may be scored in terms of broad tension areas. These areas on which research has been done are the familiar ones: the self, the family situation, and the social environment. The responses to the questionnaires are scored for the presence or absence of such tensions in the following manner: DIRECTIONS FOR SCORING THE “INTUITION QUESTIONNAIRE” Read each item carefully and score for presence of strain in the areas of self, family, and social. In any item, strain may be indi- cated for one, two, or three of these areas; score for each area indicated. Do not score one area more than once in any item. Self: Content of response states or implies tension, strain, conflict, or insecurity in regard to the self of the subject of the item, or in regard to the life of the subject; for example, in re size, sex, physical makeup, abilities, disposition, past or present misdemeanors (guilt), “boring exist- ence,” “unsatisfying life,” worry, anxiety, etc. Example: For item i, the response: “B. has failed at every- 3*8 ALEX C. SHERRIFFS thing she has tried so far, and looks for something at which she can succeed.” . . . Family: Where response indicates in regard to parents or sibs tension, hostility, ambivalence, strain, dependence which brings on concomitant inse- curity, suppression by parents, fric- tion between parents or sibs: it should be scored as “tension family.” Example: For item i, the response: “B’s family have always held her down.” . . . Social: Tension, or lack of confidence in social techniques, social relationships, social approval, number of friends, mobility, probable success in mar- riage, friction with outside individ- uals or group, fear of “robbers,” “rape,” etc., are scored “tension social.” Example: For item i, the response: “B. has never had a date with a boy, and keeps thinking of the good times she would like to have.” . . . (This presentation of the directions is much abbreviated for the purposes of this article. The complete directions include a discussion of common problems of scoring and give standards for borderline responses.) The individual subject’s score for each tension area is obtained by sum- ming the number of items in response to which this tension was indicated. A given item response might be scored as expressing no tension or as express- ing tension in one, two, or three areas. A general, combined area, tension score is obtained by adding together the scores for each of the three areas. Item analysis indicates that the ques- tionnaire is apparently an adequate instrument for working with tension to the extent that a wide range of scores in the three areas results. Further, most of the specific items draw a variety of response in terms of the tension area indicated.2 2 The fact that a more limited variety of response than was desirable resulted for several of the items suggested some revision of the ques- tionnaire. Such revision is now in progress. EXPERIMENTAL EVALUATION OF THE “INTUITION QUESTIONNAIRE” The Subjects The subjects utilized for the major part of the evaluation of the question- naire were 93 junior and senior students enrolled in a large state university, comprising the entire female population of an advanced course in clinical psy- chology.3 The subjects ranged in age from 19 to 32 years, averaging 22.16 years. Of these, 51 per cent were psy- chology majors, 33 per cent were in related fields, such as premedical, social welfare, and child development, while 14 per cent were in other fields, such as general curriculum and English. The questionnaire was administered as a group test to the entire class three weeks before the end of the semester. The Criteria In order to gain adequate material of the kind necessary for independently determining the tensions of the subjects, it was desired to secure case history, interview, or autobiographical data for each subject. Only the autobiography proved prac- ticable, and its choice was desirable for several reasons. In the first place, it had been a tradition of this particular psychology course for a number of years for the students to write their autobiographies analytically, following a mimeographed outline which was given them early in the semester. As the semester developed, this outline was elaborated upon in lectures which were concerned in part with a discussion of the variables of human behavior and were aimed at increasing the students’ 8 The reason for limiting the sample to one sex was that this phase of the investigation took place in the war year, 1944, at which time the male student body of the university was markedly depleted. THE “INTUITION QUESTIONNAIRE” 329 awareness of significant areas of life experience, in order that they might integrate their own experiences mean- ingfully in the writing of the autobiog- raphies. Not until the middle of the semester were they to begin the actual writing, for which seven weeks were allowed. Since this writing was tra- ditional, the students would in no way feel that they were singled out for par- ticular experimental study, as they were simply fulfilling the requirements of the course.4 In addition, the autobiography seemed desirable because there has been sup- port in recent literature for the thesis that quantification of such material is practicable. [Vide Allport (i), Stouf- fer(io); Cavan, Hauser, and Stouf- fer(3);Stagner(9).]B The autobiographies, which were typed, averaged on their completion 50 pages in length. The shortest contained 28 pages, the longest, 108. The autobiographies of the 93 sub- jects were rated for the degree of ten- sion that the rater felt to be present in each of the three areas. Also, an overall tension rating was made. The five-point scales used for these ratings are modifications of the scales success- fully used in the University of Califor- nia Child Guidance Study (6). These rating scales follow. * To make it possible for the students to pursue their self-analysis with greater freedom, and to insure objectivity in scoring the questionnaires and in rating the autobiographies, a coding scheme is employed in the course which provides for anonymity except at the time of the actual tran- scribing of grades for the group as a whole. Students are permitted to disclose their identity only after the course is formally over if at that time they wish to discuss with a clinician material that they have uncovered. 5 After the investigation had been concluded, intensive interviews with 18 of these subjects indi- cated to the satisfaction of the investigator the essential validity of the autobiographies for this particular sample at least. RATING SCALES OF TENSION AREAS: SELF, FAMILY, SOCIAL AND COMBINED AREA Self: [Adjustment to size, sex, physical makeup, abilities, disposition, past or present misdemeanors (guilt), “bor- ing existence and unsatisfying life,” worry, anxiety, etc.] 1. Extreme in tension in regard to self; extreme lack of confidence, excessive in self-depreciation and unfavorable comparison—or com- pensatory and marked bragga- docio. Perhaps intense guilt feel- ings, or feelings of inadequacy. 2. Marked unacceptance of one or two items of makeup, but not as extreme as i. (If to point of dominating whole personality, classify under i.) May have many mild points of discomfort. 3. Lacfe confidence in some particu- lars, but for the most part is unself-conscious—or unaware—of assets or liabilities. Occasional tension or conflict about self, but not enough to be considered as characteristic. 4. Without much tension; this per- son recognizes handicaps but accepts them realistically; or un- aware, and not bothered. Little evidence for compensatory be- havior to be found. 5. Full confidence; accepts limita- tions and assets easily. Family: (Codings on severity of tension, whether it involves one or more family members, including step- parents, or parent substitutes in the home. Watch for tension produc- ing relationships whether shown openly, or by withdrawing or in compensatory demonstrativeness, or relationships so ingrown that nor- mal, friendly relations to outsiders are hampered or precluded.) 1. Extreme tension 2. Same trends, but not so extreme as i 3. For most part, low-grade tension 4. Absence of real strain 5. Completely secure in family affection ALEX C. SHERRIFFS Social: (The area of the person’s feeling of tension or confidence and satisfaction in social relations.) 1. Acute tension in this area 2. Definite but less acute tension than i 3. Taes social relationships fairly much in stride 4. Very little tension 5. Confident Combined Area: (An overall evaluation of the person’s total adjustment; the general security-insecurity which is die dynamic resultant of all aspects of the individual’s life-situation.) 1. Extremely tense 2. Markedly tense 3. Moderately tense 4. Low-grade tensions only 5. Absence of real strain (The definitive statements which accom- pany the codings on these scales are omitted here. In general, they are comparable to the specifications regarding Tensions in Self.) Results 6 i. Reliability. Several tests of the reliability of the questionnaire have been made. These include: (i) The correlations between the ratings of two different raters. (These were calcu- lated for both Rater A vs. “Rater B and Rater A vs. Rater C. Rater A discussed the rating scales with Rater B, while Rater C had access only to the printed directions.) (2) The correlations be- tween the ratings of one rater, Rater A, at two different times. This particular test of reliability was desired since 6 The findings on reliability and validity are based upon Pearson product-moment correlation coefficients. The variables correlated were the unweighted scores for the questionnaires and the values on the rating scales for the criteria. Three individuals contributed to the scoring on which measures of reliability and validity were based: Rater A, the writer, who at the time of the study was lecturer in clinical psychology at the University of California; Rater B, Frances Sherriffs, who earlier as a graduate student at Stanford University has been appointed clinical fellow; Rater C, Sydney Smith, who was acting as teaching assistant in clinical psychology at the University of California at the time of the investigation. validity coefficients were based entirely on the ratings of Rater A (see below). (3) Correlations between scores result- ing on test-retest administration with nine months intervening. These ques- tionnaires were scored by Rater A. An estimate of the reliability of the criterion has been achieved by corre- lating the ratings made by Raters A and B on a sample of the autobiographies. The results obtained by the various techniques employed to test reliability are summarized in Table i. 2. Intercorrelations. It is important to know to what degree the three defined areas of tension under consid- eration are independent. Theoretically, it is conceivable that the three areas could be merely three expressions of the same thing—namely, general tension. Therefore, intercorrelations were calcu- lated between the three scales of the questionnaire and between the three areas in the autobiographies. These intercorrelations are low enough so that we may conclude that in scoring the questionnaires or in the ratings for the three areas of the auto- biographies we are getting at sufficiently independent variables. The intercorre- lations based on the entire sample of 93 subjects are presented in Table 2. 3. Questionnaire Validity. The pri- mary measures accepted for validity in this investigation are the correlations between the autobiography ratings and the scores on the questionnaire. Since only Rater A scored all the data in this study, validity coefficients are based on his ratings. These correlations, based on the entire group of 93, appear in Table 3.7 SUPPLEMENTARY FINDINGS ON VALIDITY Although the primary measures ac- cepted for validity in this investigation 7 For a discussion of the relatively low validity of the questionnaire score for Family Tension, see below, discussion of certainty of criterion. THE “INTUITION QUESTIONNAIRE” TABLE i FINDINGS ON RELIABILITY * 331 QUESTIONNAIRE Rater A vs. Rater B Rater A vs. Rater C Rater A vs. Rater A Test-Retest (9 mos.) (Rater A vs. Rater A) CRITERION Rater A vs. Rater B SELF .90 • 73 • 93 .76 .64 FAMILY .96 .87 •93 .76 •7i SOCIAL •73 .60 •73 .76 •49 COMBINED AREA .89 •73 .80 •75 •77 * These correlations are based on random samples of 39 cases and are all significant at the i-per-cent level. are the correlations between autobiog- raphy ratings and the scores on the questionnaire, these correlations are given added meaning when they are considered along with certain further information. Information on the fol- lowing points is therefore presented: (i) the correlations between question- naire scores and criteria consisting of interview ratings, (2) the effect on the validity coefficients of the certainty with which the criteria ratings were made, (3) the correlations between the ques- tionnaire scores and the subjects’ self- ratings on the tension areas, (4) the effect of the subject’s conscious desire to disguise on his responses to the questionnaire, and (5) the effect on questionnaire scores of membership in a known group, i. Naive Subjects with Interview Criteria. There were two major rea- sons for enlarging the investigation to include an additional group of subjects. First, it was desirable to verify on a psychologically nai’ve group the results which had been obtained on the clini- cally sophisticated group. It was im- portant to learn whether na’ive subjects could respond to a questionnaire in terms of “underlying motivations” and whether they would project in a man- ner similar to that of the more sophisti- cated. Second, it was important to determine whether the relationships found between the questionnaire and the autobiography would hold when the questionnaire was measured against some other method of clinical evaluation. A sample of 38 women was chosen from a class of 325 lower-division stu- TABLE 2 INTERCORRELATIONS BETWEEN TENSION AREAS TENSION Self vs. Family Self vs. Social Family vs. Social AUTOBIOGRAPHY * • 34 • 53 .27 QUESTIONNAIRE t .10 •27 .01 * These correlations arc significant at the i-per-cent level. tOf these correlations, self vs. social is significant at the i-per-cent level; the remaining are not significant at the 5-per-cent level. 332 ALEX C. SHERRIFFS TABLE 3 VALIDITY OF THE QUESTIONNAIRE FOR TENSIONS TENSION Self Family Social Combined area CORRELATION * .76 •57 * These correlations are significant at the i-per- cent level. They compare favorably with meas- ures of validity reported for projective and other personality tests. In considering their magnitude relative to the degree of reliability of the criterion and of the questionnaire scoring, it is necessary to take account of the fact that the ratings utilized for validity estimates were those made by Rater A, and that, therefore, the pertinent reliability figures are the ones based on the ratings of this rater. dents. The course in which these subjects were enrolled was a beginning survey course in psychology, open only to those who did not intend to take further courses in psychology. This group was comprised of college fresh- men and sophomores who ranged in age from 17 to 24, with a mean age of 18.99. ^n tnis grouP there were no psychology majors, 24 per cent were majors in related fields, and 76 per cent were in other fields. (For comparison with the clinical group see presentation above.) The questionnaire was administered to this group at the beginning of the semester and again ten weeks later, thus giving material for computing short- term test-retest reliability. The criteria in the case of this group were ratings made on the basis of a 50- to 6o-minute interview with each subject. During this interview material was sought by which the interviewer, Rater A, could judge the subjects on the three areas of tension and on general insecurity. It is obvious that in one hour any attempt to obtain full and accurate information on highly personal areas could not be entirely successful. The interviews were conducted from four to six weeks after the first admin- istration of the questionnaire. For the questionnaire, two estimates of reliability were obtained, Rater A vs. Rater B, and test-retest. It proved im- possible to obtain material by which to estimate the reliability of the criteria. The correlations between the inter- viewer’s ratings and the questionnaire scores on the tension areas for these subjects served as an index of validity. For these estimates of reliability and validity see Table 4. The findings on reliability for this group suggest that “naive” subjects are able to respond to the questionnaire in terms of “underlying motivations” to the extent that their responses can be scored with as much consistency be- tween raters as had been found for the responses of the “sophisticated” sub- jects (Table i). The magnitude of the ten-week test- retest reliability correlations reinforces the conclusions suggested by the nine- month reliability that there is consider- able temporal stability of tension areas as indicated by questionnaire scores. The validity coefficients run lower than do those for the clinical psychol- ogy group, as was expected because of the nature of the criteria, though other factors may also contribute to the lower correlations. The above relationships and the fact that the two groups represent a fairly wide range of age, academic status, and fields of interest, although these vari- ables were not systematically sampled, suggest that further research might prove the test to be capable of wide application in college populations. 2. The Effect on the Validity Corre- lations of the Certainty with Which the Criteria Ratings Were Made. It is obvious that one factor which may be THE “INTUITION QUESTIONNAIRE” 333 expected to influence markedly the validity coefficients is that of the ade- quacy of the criteria. In order to evaluate the importance of this factor we have treated separately those subjects for whom the autobiography seemed most adequate and those for whom this criterion seemed least adequate. The method by which the autobiog- raphies were grouped as to most and least adequate was by division in terms of academic grade on the papers, sort- Correlations between questionnaire scores and the autobiography ratings for each of these two groups vary in the main from the correlations based on the total clinical psychology group in the expected directions. This finding and the results of tests of the significance of the differences between the correlations of the “Certain” and “Uncertain” groups appear in Table 5. The trends represented in the above findings and the significant difference TABLE 4 RELIABILITY AND VALIDITY ESTIMATES BASED ON NAIVE SUBJECTS ‘ RELIABILITY OF QUESTIONNAIRE Rater A vs. Rater B Test-Retest (10 weeks) VALIDITY Questionnaire vs. Interview SELF .87 .68 .64 FAMILY .92 .90 .56 SOCIAL •75 .70 .42 COMBINED AREA •73 •78 •54 * These correlations are all significant at the i-per-cent level. ing the autobiographies into those which had originally been graded “A” in quality and those which had received grades lower than “A.” The letter grade “A” had originally been defined as representing an autobiography which had been written at a dynamic rather than a descriptive level, which had covered all the material asked for in the outline, and which was complete enough to give an understanding of its author and her life-space. Papers less adequate in these qualities were given grades lower than “A.” These grades had been assigned to the autobiog- raphies before the present aspect of the research problem had been anticipated. In the total group of 93 subjects, 31 had written autobiographies which had been graded “A”; this group was con- sidered as the “Certain Group.” The remaining 62 cases were considered as the “Uncertain Group.” found for the area of family when con- sidering only the “Certain” cases sub- stantiate the belief that inadequacies of the autobiography as criterion do have a depressing influence on the validity coefficients. The estimates of validity based on the “Certain Group” would appear the more reasonable ones by which to measure the predictive value of the questionnaire.8 3. Questionnaire Scores vs. Subjects’ Self-ratings on the Tension Area. It 8 The original findings based on the entire group of 93 subjects which had resulted in a rela- tively high reliability coefficient for ratings on the family area of the criteria and yet yielded the lowest coefficient of validity (Tables i and 3) might be explained by the fact that the raters had similar viewpoints as to the meaning of such phenomena as, for example, the omission of any discussion of one member of the family, the over- protestation of love or hate for any family mem- ber, or the lengthy defense of a parent’s cultural background. This meant that it might be possible that two raters would make similar ratings on some papers which really yielded insufficient data for valid ratings. 334 ALEX C. SHERRIFFS will be objected by some that tension scores constitute biographical data easily obtained by self-report, whereas such constructs as needs are dynamic, funda- mental to the individual, and relatively unconscious. The investigator subscribes to the contrary belief that subjects are not fully conscious of their own tensions. He conceives of tensions as representing dynamic, fundamental aspects of per- sonality—often basic to needs which relationship was expected to hold true in the case of correlations between self- ratings and the clinical ratings derived from the autobiographies in the one case and from the interviews in the other. The actual results are presented in Table 6. Though the more mature subjects rated themselves more in agreement with the questionnaire scores and with the clinical ratings than did the begin- ning subjects, the correlations for both TABLE 5 EFFECT ON VALIDITY COEFFICIENTS OF “CERTAINTY” OF CRITERION RATINGS CORRELATIONS BETWEEN QUESTIONNAIRE AND AUTOBIOGRAPHY * “Certain” Group “Uncertain” Group All cases combined (See Table 3) SIGNIFICANCE OF DIFFERENCE BETWEEN CORRELATIONS f “Certain” vs. “Uncertain” Groups SELF •7i •74 .76 0.04 FAMILY .82 •34 • 57 4.62 SOCIAL .78 .65 • 7i 1.65 * All correlations in this table are significant at the i-per-cent level. tThe critical ratio based on the difference in regard to family surpasses the i-per-cent level o£ confidence, the remaining critical ratios do not reach the 5-per-cent level. themselves sometimes seem best de- scribed as constructs representing chan- nels by which an individual attempts partially to relieve his tensions. This appears to be true especially when those needs have not yet become autonomous in their functioning. For information as to the validity of subjects’ self-ratings, such reports were obtained from a sample of 36 of the clinically sophisticated subjects and from the entire group of 38 “naive” subjects. It was to be expected because of the greater maturity and insight and broader psychological background of the more sophisticated subjects that there would be a higher correlation between their self-ratings and question- naire scores than would be the case with the more naive subjects. The same groups appear to be low enough to refute the objection that tension areas are variables on which subjects can and will consciously report accurately. 4. The Effect of the Subject’s Con- scious Aim to Disguise His Responses to the Questionnaire. If the question- naire is to be valuable as a projective method, either to reveal important aspects of personality or to serve as a control against which to study the variables that influence projection, then it is important to know how much its validity depends on the subject’s co- operation and honesty. Also, it is of importance to know what aspects about himself a subject can disguise in his responses to the test, if motivated to do so. The possibility of disguise would be maximal if the subject had insight both as to what variables the THE “INTUITION QUESTIONNAIRE” 335 test was intended to tap and as to the function of these variables in his own makeup. One of the important criteria of project!ve tests is that “The subject should be unaware of the purpose of the task and of the inferences to be drawn from his productions” (5, p. 289). It is assumed further, in connec- tion with such tests, that the important variables projected will be ones into which the subject has little insight. It was “explained” that, when the ques- tionnaire was administered a second time, those subjects who were insightful would now be able to respond to it in ways which would not reveal their own personalities. Those with less insight would have more difficulty. The moti- vation of college students to display insight is well known, and it was felt that this would inspire maximum effort. TABLE 6 SELF-RATINGS ON TENSION AREAS vs. QUESTIONNAIRE SCORES AND CLINICAL RATINGS QUESTIONNAIRE SCORES vs. SELF-RATINGS “Sophisticated” Subjects “Nai’ve” Subjects CLINICAL RATINGS vs. SELF-RATINGS “Sophisticated” Subjects “Nai’ve” Subjects SELF .36* .00 .46t .32 FAMILY — . 10 — .13 • 42t .21 SOCIAL •32 .00 .27 .22 * This correlation is significant at the 5-per-cent level. t These correlations are significant at the i-per-cent level. No other correlation is significant at the 5-per-cent level. Our hypothesis, then, was that those variables which could not be disguised were ones which were an integral part of basic personality and/or were vari- ables of which the subjects were unaware. To test the disguisability of tensions on the questionnaire, the following pro- cedure was employed. The question- naire was administered under regular instructions to a new group of 38 students in a beginning psychology class. After the papers had been col- lected, the subjects were told that this questionnaire was in fact a project! ve test in which one revealed one’s own personality and the things important to one through the responses given con- cerning the motivations of the hypo- thetical individuals whose situations were described in the questionnaire. Five minutes after the first question- naires had been collected, the subjects were given fresh copies and were told that they were now to attempt in every way to disguise themselves in their responses to the items. In addition, they were asked to make their dis- guises plausible, so that the person who did the scoring would not be aware of their intent to distort the picture of themselves. This latter condition was imposed because it was felt that some subjects might otherwise leave items unanswered. The correlations between the scores on the first and second responses to the questionnaire were: Tension Self=.6i, Tension Family=.88, and Tension iSo«’fl/=.68. These correlations are all significant at the i-per-cent level. These correlations closely approxi- 336 ALEX C. SHERMFFS mate the test-retest reliabilities result- ing after a ten-week interval (Table 4). Their magnitude suggests that, even under motivation for disguise, the ques- tionnaire is a reasonably adequate in- strument for the measurement of tensions, and further suggests that these tensions represent underlying aspects of personality of which the sub- jects are unaware. These findings corroborate the conclusions drawn from the study of the relationship of self- ratings to questionnaire scores. 5. The Effect on Questionnaire Scores These results seem to indicate that the questionnaire yields measures which are related to maladjustment, defined in terms of an objective outside criterion— namely, the fact of seeking psychiatric help. The differences are all in the same direction, and the combined area score significantly differentiates the two groups. (It will be recalled that this score is found to be correlated with ratings of degree of “overall” security in the autobiographies and interviews.) Certain facts may operate to depress these critical ratios. First, there is the TABLE 7 COMPARISON OF QUESTIONNAIRE SCORES: PSYCHIATRIC ASSISTANCE GROUP vs. “NORMAL” GROUP TENSION Self Family Social Combined Area PSYCHIATRIC ASSISTANCE GROUP MEAN 18.31 7.25 10.00 35-63 “NORMAL” GROUP MEAN 15.17 6.30 8.95 30.56 CRITICAL RATIO 2.48 •97 1.19 3.06 of Membership in a Known Group. No systematic effort has yet been made to procure standardization data from outside criterion groups. However, the questionnaire responses have been evaluated in relation to certain readily available information concerning some of the subjects. Nine subjects in the advanced psy- chology group and seven in the begin- ning psychology group were known by the investigator to have sought during the year some sort of professional psy- chiatric help. The questionnaire scores of these 16 subjects were compared with those of the rest of their respective groups. Tests for the significance of the differences between the means of the scores on the questionnaire scales for these two groups yielded the results presented in Table 7. fact that an unduly large proportion of the group of 16 belonged to the begin- ning psychology group, which had on the average a slightly lower mean score for the scales.9 Secondly, there may well have been students in the “nor- mal” group who, unbeknown to us, were themselves seeking psychiatric help. The scale differences between the two groups are in line with clinical experience on the basis of which it would be predicted that overall (total) tension and self tension would be most indicative of felt need for psychiatric help. 9 The mean for combined area (total tension) was i.80 points lower, and the mean for self tension was 3.38 points lower, and the mean for family tension and social tension were respectively .370 and i.08 points higher. THE “INTUITION QUESTIONNAIRE” 337 COMMENT ON TOTAL (COMBINED AREA) TENSION SCORE After tension has been analyxed into the three areas, self, family, and social, and after the interrelationships between these three areas have been found to be very low, the problem arises as to the meaning of the total tension score. On the one hand, it appears to be the sum of three variables which statisti- cally have little relationship to one another. It is difficult to interpret the meaning of a score apparently based entirely on quite discrete variables. On the other hand, this score does corre- late, and relatively highly, with what to the raters is a meaningful concept when judging the autobiographies and the subjects of interviews. The real meaning of the total tension score must in the long run rest on its relationships to outside criteria. One such relationship has been presented, showing that on this variable alone students seeking psychiatric help are significantly differentiated at the i-per- cent level from the rest of their class. CONCLUSION Evidence is presented to substantiate the belief that a new projective method, the “Intuition Questionnaire,” ade- quately meets certain important cri- teria. It (r) yields relatively quanti- fiable, valid, and reliable results, (2) is easy, rapid, and standard in adminis- tration, (3) provides for standard and systematic scoring, (4) is disguised in a plausible manner so that it is not obvious to the subjects that it is a per- sonality test, (5) provides for common strong motivation, and (6) is aimed at a relatively stable level of personality. This investigation leads logically to problems for further research, such as revision and standardization of the “Intuition Questionnaire” and further experimentation with it and the vari- ables on which it is scored. In addition, the writer believes that development of this instrument will pave the way for valuable research on individual differ- ences in projection and the variables underlying such differences. REFERENCES 1. ALLPORT, G. W. The use of personal docu- ments in psychological science. New York: Social Science Research Council, 1942. 2. CATTEIX, R. B. Projection and the design of projective tests of personality. Character & Pers., 1944, 12, 177-194- 3. CAVAN, R. S., HAUSER, P. M., & STOUFFER, S. A. Note on the statistical treatment of life history material. Soc. Forces, 1930, 9, 200—203. 4. FRANK, L. K. Projective methods for the study of personality. /. PsycAol., 1939, 8, 389-413- 5. HAGGARD, E. A. A projective technique using comic strip characters. Character & Pers., 1942, 10, 289-295. 6. MACFARLANE, JEAN W. Studies in child guidance: I. Methodology of data collec- tion and organization. Monogr. Soc. Res. Child Developm., 1938, 3, No. 6. 7. RAPAPORT, D. Diagnostic psychological test- ing. Vol. II. Chicago: The Year Book Publishers, Inc., 1946. 8. SARGENT, HELEN. Projective methods: Their origins, theory, and application in person- ality research. PsycAol. Bull., 1945, 42, 257-293. 9. STAGNER, R. The psychology of personality. New York: McGraw-Hill, 1937. 10. STOUFFER, S. A. An experimental compari- son of statistical and case history methods of attitude research. (Unpublished.) 11. SYMONDS, P. M., & SAMUEL, E. A. Projective methods in the study of personality. Rev. educ. Res., 1941, n, 80-93. 12. WHITE, R. W. The interpretation of imagi- native productions. In Hunt, J. McV., Personality and the behavior disorders. New York: Ronald Press, 1944. Pp. 214- 254.