All rating notations (high, low, presence, absence, omission, mixed) were
converted
into "high," "low," and "neutral" scores.
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
"
Sometimes the statement of the problem appears more specific (e. g. , "can't do such and such work") but the main thing about it is the idea that "one
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 92 I
cannot do one's duty" (for instance, as a good wife or mother) due to this unwelcome illness. The concern is not so much with the task itself or the feeling of failure involved in the present inability to carry it out, but rather with the violation of conventions and morals. But there is little awareness of specific conflicts, fears, frustrated desires, or life failures.
III. "Ego-alienness"
Presence (goes with high score); The symptoms and the patients' atti- tudes toward them give evidence of particularly strong repressions. The repressed problems and also some of the symptoms have a strong ego-alien quality. Certain impulses, problems, and even some symptoms are experienced as completely "foreign" to the self. They cannot be accepted or admitted. They "belong to the body," or they are the "breaking through" of a com- pletely unacceptable part of the personality. The person "cannot understand what makes me do that. " There is fear of something "devilish inside" that overwhelms the normal accepted self. Fear of "losing control. " Examples: "Fear I'll kill someone during one of my spells; fear I'll lose my mind. "
When certain impulses have broken through and have been expressed in behavior, this ego-alienness is expressed sometimes in moralistic statements and self-accusations ("Can't understand how I could do this"). Another criterion might be the blandness of the interview, stemming from the patient's inability to admit socially unacceptable impulses, either because of repression or because of fear of disapproval from the interviewer.
Absence: The symptoms are experienced as belonging to one's personality and life history. There may be severe conflict over some impulses, but the latter are admitted to consciousness and understood as part of one's self and life experiences. There is relative frankness and freedom from "moralism. "
In the case of obsessions, compulsions, delusions, and other psychotic manifestations, which usually have some ego-alien quality, the rating will have to be based on the relative emphasis on the ego-alienness itself ("Some- thing makes me do it; someone influences me; can't control it," etc. ), or on the degree to which the content of the symptom is consciously divorced from or related to the patient's past or present inner life.
IV. Externalized Theory of Onset and Causation of the Illness
Presence (goes with high E scores): Attempts at externalization of the symptoms by: (a) denying any and all precipitating factors. Tensions, de- pressions, etc. , appear "without any reason at all. " "It just appeared. " (b) dating the illness back to a very definite event or moment: an operation, a "spell," a death, a particular day-sometime during the last few weeks or years. "Before that I was well . . . happy . . . . " (c) making no attempt to relate the illness to one's past, especially not to childhood. No spontaneous references to childhood unless specifically questioned, during this first inter- vrew.
Absence: Spontaneously, or when asked about onset, states that the prob-
? THE AUTHORITARIAN PERSONALITY
lems have been present "for a long time"-for years, always or since child- hood, but perhaps in milder form.
V. Spontaneous Mention of Unhappy Childhood or Family Relationships
Presence (goes with low E scores): Patient spontaneously refers to his unhappy childhood. Often elaborates in great detail on_a history of frustra- tions (often with a clearly masochistic attitude). Complains of parents' pres- ent or past attitudes and is critical of them.
Absence: No spontaneous reference to childhood. Childhood, if discussed at all, is pictured as "happy and normal. " No criticism or other sign of hos- tility towards family members is expressed (except in "spells" and psychotic episodes), in spite of admission of friction. Sometimes there is clearly com- pensatory great concern for family members and their welfare.
VI. Cues Referring to the Patient's Character Structure
Obviously the interview fragments cannot be used to construct complete pictures of the subject's personality structure. It was found, however, that they often contained some important hints as to the nature of this structure. Formulation of categories listed below was, of course, guided by the con- cepts found most useful for distinguishing high and low scorers in the rest of the study. But only those categories were included which would be applied to the particular interview material at hand. The raters were asked to state for each case whether cues from List A or List B predominated in a record, or whether the record was "neutral" (meaning he could discern equally as many cues from "A" as from "B" or none at all).
List A (goes with high degree of ethnocentrism):
1. "Countercathectic" defenses. Men: counteraction against passivity.
Women: Anal reaction-formations. No mention of a love object or re-
lationship (other than spouse).
2. Extrapunitive or impunitive. 6 If self-criticism occurs, it is couched in
moralistic terms, "I am bad, have done wrong. " "God is punishing me. " Tendency either hysterically to dramatize certain physical symptoms, spells, etc. , or to minimize all complaints of a psychological nature.
3? Externalized superego. Religion: God is seen as an external judge of one's action. Illness is a punishment from God. Or religious standards are part of the conventional ideology. Religious reasons are given for refraining from divorce or suicide.
List B ("low" characteristics):
4? Absence of "countercathectic" defenses:
Direct expressions of orality: dependent character traits; eating, drink- ing, drugs. Dependency problems, nurturance. Open admission of weak-
6 Further research onirnpunitiveness is indicated. Here it seemed preferable to classify it "high;" it was "low" in Chapter XI (p. 406).
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 92 3
ness, passivity, femininitY in men. Love-seeking attitude. References to rejections by specific love-objects.
Sublimations: form an important part of the ego; references to achieve- ment. When symptoms interfere particularly with patient's work, this is stated in such a way as to suggest concern with the particular work he is doing. Emphasis on ideals. Concern with helping others, society; artistic activity and interest.
5? Intrapunitiveness, masochism: "I was mistreated as a child. " Identifica- tion with suffering; self-criticism ("I am a failure").
6. Internalized superego: guilt feelings, true depressions, religious ideas as part of the inner life of the person.
VII. Predominant Types of Symptoms
List A ("high"):
1. Physical anxiety symptoms and other emotional equivalents: "hyper-
ventilation syndrome"; dizziness, sweating, tingling sensations, numb- ness, tachycardia, breathlessness, fainting, tremors. Anxiety has no con- scious content.
2. Emotional outbursts, tantrums and "spells," in women.
3? Hypochondriacal fears: fear of death, heart attack, etc.
4? Hysterical conversions. 7
5? "Rigid compulsive rumination": repetitious complaints, self-accusa-
tions, self-reassurances, "thinking around and around in a circle. "
? 6. Depersonalization (sense of estrangement from self and world) in a person who emphasizes that heretofore he had had no tendency toward
timidity and withdrawal.
7. Suspiciousness, fear of people or aversion to people is stated in some-
what externalized terms: "They irritate me," "I can't stand them. They make me nervous. " (To be distinguished from hypersensitiveness and withdrawal when described in a more intraceptive way. )
8. Psychopathic tendencies, not in the sense merely of unconventional behavior (as the term is sometimes used) but rather in the sense of a really defective and not sufficiently internalized superego. Antisocial and destructive behavior, callousness, emotional shallowness.
9? Stomach ulcers in men, especially in subjects who emphasize their masculinity.
List B:
ro. Depressed mood, hopelessness, lack of self-confidence, verbalized feel- ings of inadequacy, suicidal ideas, guilt. Often patients complain of
7 This item was included in the manual used by our raters; but later analysis of case material suggested that there may be more frequent hysterical conversions in low scorers. The trends are as yet not clear. Perhaps there is also a sex difference here. Two of the low-quartile men had conversion symptoms.
? II.
"depression," but the "true" depressions as described above have to be distinguished from the more schizoid type of mood disturbance. Tiredness. "Neurasthenia. "
Dismenorrhea.
THE AUTHORITARIAN PERSONALITY
I2.
I 3?
I4? In men, expression of traits opposite to the culturally emphasized
I 5?
masculine pattern. Withdrawal, timidity, shyness, sometimes coupled with feelings of unreality or with physical anxiety symptoms or hys- terical conversion. The patient's attitude toward all these "weaknesses" is to some degree accepting.
Conscious anxiety and conflicts.
3. THE METHOD OF QUANTIFICA TION
Conscious rejection of feminine role by women.
Inspection of the statements of complaint led to the impression that the high and low quartile groups were clearly differentiated with regard to the variables just discussed. In order to check this impression in a more rigorous manner, it was decided to use a method of "blind ratings," similar to that used on the interview, T. A. T. and Projective Question Test material (Parts II and III). At least two judges who are not acquainted with the subjects (and in this case did not know which were high and which were low scorers)
independently rate each subject on certain characteristics. These ratings can then be compared with other data on the subjects. The problem of blind ratings has been discussed in other chapters dealing with interview scoring and projective questions. If these ratings turn out to be highly correlated with some trait of the subjects (in this case ethnocentrism) on which the judges had no information, it is very probable that relationships between the former and latter traits exist in reality and could be demonstrated by other methods. (This holds only if the rating procedure is so controlled as to pre- vent the raters from utilizing cues other than those to be experimentally tested. ) There are, however, various possible pitfalls inherent in the rating method. Such "errors of measurement" could arise, for instance, from material which did not contain sufficient information from which to judge the subject on a certain trait. Other errors might arise from the manner in which the characteristics were defined and described for the judges; from the ability of the judges to apply the instructions to the material to be judged -depending in turn on the judges' training, theoretical bias, and personality- and from other factors. Use of a rating technique, therefore, requires not only the determination of certain relationships between the ratings and other data, but also a careful consideration of various factors which could have influenced the results either in the positive or in the negative direction.
a. THE RATERS. The two primary raters were staff members of the major study and will be referred to as raters (or judges) A and B. In addition to much clinical training and experience, these two raters had a strong psycho-
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 925
analytic orientation. Both were familiar with all concepts, hypotheses, data,
and results of the total study. Both had had opportunity to interview high-
and low-scoring subjects, and were therefore acquainted with the behavior
I
and material usually obtained from such subjects in interview situations. Each of the judges independently rated each record on all seven cate- gories and then assigned an "over-all" intuitive estimate of "highness" or
"lowness. "
After these primary ratings had been completed, 7 independent raters
(they are referred to hereafter as the control raters, their ratings as control ratings) were used, each rating one category only. 8 Six of the control raters were clinical psychologists (of these, r was a senior clinician, 5 junior clini- cians at the level of internes, working at the Langley Porter Clinic). They were not acquainted with the concepts and data of the over-all study. This was important for reasons to be discussed below. One of the more narrowly defined categories (Onset and cause of illness are explained by subject in externalized terms) was rated by our office secretary who had had no formal training in psychology or psychiatry, but who had much intuitive psycho- logical insight and who had absorbed a great deal of the research material and hypotheses. The 7 judges varied greatly in age, training, and theoretical orientation.
b. THE RATING TECHNIQUE. The instructions for the control raters were as follows:
The material to be rated consists of "Statements of Complaint" by Langley Porter Clinic patients in a first intake interview or in the first interview with a therapist. Only the section "patient's story" or "chief complaint" was included. The inter- views are here reproduced verbatim, although a few have been slightly condensed by the writer. Each numbered paragraph refers to one case. There are 26 men, 33 women.
Each case is to be rated on one variable (or syndrome) as described in the manual. Each rater will be assigned one variable and will not know about other variables until he has completed his ratings.
The ratings are to be made in terms of presence (v) or absence (-) of the trait. A few of the variables permit of a "mixed" judgment. Assignment of such a "mixed" (M) rating should be avoided if possible. But occasionally it may have to be used. Sometimes (due to the fragmentary way in which some of these interviews are
recorded) there will be insufficient material to rate. In this case mark (o).
Each primary rater (A and B) first rated each record in terms of all seven individual categories. They knew which categories were expected to be related to high or to low ethnocentrism. They therefore tried to assign each record a "high" or "low" rating for each category. Often a record did not contain enough material to permit the rater to reach a decision on a given variable, e. g. , a given topic was not discussed, or there were few cues permit-
s W e wish to thank Dorothy Bomberg, Janet Gist, Carole Home, Virginia Patterson, Dr. Claire W. Thompson, Anne Vollmar, and Elaine Wesley Barron for the patience and care with which they carried out, on short notice, the task of doing the control ratings.
? THE AUTHORIT ARIAN PERSONALITY
ting inferences regarding character structure. In this case no rating was assigned. When there were about an equal number of cues pointing in the high and in the low directions, a "mixed" rating was assigned. The raters then went over the records a second time, trying to guess in each case whether the subject had made a high or low score on theE scale. The guesses were to be based on the decisions reached regarding the individual categories. However, no mechanical formula was set up to convert the individual ratings into "over-all" ratings. The raters arrived at the latter by a new rating process in which any or all of the categories could be used and weighted as the rater saw fit.
Two types of data were obtained from this rating material: (I) Inter- rater scoring agreement for each category and for the over-all ratings. (2) Relationship between ratings and scores on the E scale. These will now be discussed.
4. T H E RELIABILITY OF T H E MEASURES
When several persons agree considerably more than half of the time that certain subjects in a group do, others do not, possess a given trait, the chances are good that these various raters knew what they were supposed to look for, had a similar conception of the trait, understood this concept, and could clearly recognize something in the interview data to which this concept could be applied; and that personality, training, and other differences be- tween the raters influenced the ratings only to a relatively small degree.
All rating notations (high, low, presence, absence, omission, mixed) were converted into "high," "low," and "neutral" scores. E. g. , a rating of "pres- ence" on variable I-Main emphasis on subject's physical complaints-was- considered a "high" score, "absence" a "low" score; "mixed" notations and omissions were considered "neutral" scores.
Scoring reliability was then obtained by computing the percentage of times 2 raters had assigned the same scores to the same records. Whenever both raters had assigned exactly the same score (high, low, or neutral) to the same record, this was considered one agreement. When one of the raters had given either a high or a low, the other a neutral score, this was considered one-half an agreement. When one rater gave a high score, the other a low, this was counted a full disagreement. The number of agreements, divided by the total number of records rated, yielded the percentage agreement between 2 raters. There were very few instances in which both judges gave a neutral score.
Table 9(XXII) shows the percentage agreements between Raters A and B as well as the scores on which these figures are based. All of the percentage agreements, except one (category III, "lows"), are above 70, statistically higher than could have been obtained by chance (I per cent level).
Raters A and B agreed best, 9I per cent, on variable V (Subject spon-
? Over-all rating of highness or lowness on E Scale
Single Variables:
I. Main emphasis on somatic
complaints
II. Intraception
III. Ego-alienness
Low 4. . 3 High 1551 .
Low
51 82. 1 86. 4 90. 3
45 62. 1 76. 3. 71. 0
48 78. 6 83. 1 83. 9
41. 5 76. 8 70. 3 64. 5
45 75. 0 76. 3 n. 4
1 .
13 1 28. 5 91. 9
IV. Externalized theory of onset High and causes Low
13 83 4 26
Low322 1
High 15'72 1 Low 1 4 3 1
-
2 2
25 914
. 3 184
2 10 10
. 2 9 8
23
22
22 26
21. 5 20
21 24
High 14111 Low321 5
.
. . . 2
'7 2 2 2 . 2
VII. Predominant type of symptoms High 18
Low 4 - 1 4
v. Spontaneous mention o t IJI! haPPY High
childhood and family relations ! LOW 15
VI. Cues regarding character High
structure Low 3
2
25 53. 5 89. 3 90. 7
23 2
10
2 20
520 46 '71. 4 '78. 0 4 26 83. 9
TABLE 9 (XXII)
DIE AMQUNT OF AGJ! EE)IENT BETWEEII TWO RATmS IN ESTIMATING A SUBJECT' S STANDING ON DIE E SCALE FROM AN ANALYSIS OF HIS INTAKE INTERVIEW
PSYCHIATRIC CLINIC PATIENTS: MEN AND WOMEN COMBINED (N = 59; High scorers: N = 28; Low Scorers: N = 31)
No. scored High (H) No. scored Neutral (N No. scored Low (L)
Interrater Agreementa
No. of Agreements Percent Agreements
Actual
Standing (! )Rated (2)Rated H (3)Rated N (4)Rated H (5)Rated L (6)Rated L ('! )High (8)H and L (9)High (IO)H and L
onE H by by One; N by Both by One;\ L by Both by One: N or Combined or Combined
Scale Both by Other bY Other High 21 - - 5
by Other Low 2 . 23
413
1
Mean agreement on the variables High 15. 4 6. 3 1. 4 1. 4 0. 9 2. 6 22. 2 Low 4. 8 I. 7 1. 9 2. 4 14. 0 6. 3 24. 7
! ! Number of agreements for a given variable is the sum of the values, opposite that variable, in Columns 1. 3, and 5.
24 . 28
1 22. 5 19 3 25. 5
48 80. 3 81. 4 82. 3
'79. 3 79. 5 79. 7
? THE AUTHORITARIAN PERSONALITY
taneously and explicitly mentions unhappy childhood and family relation- ships). This result was to be expected because this variable is so unambiguous and requires little subjective evaluation. Besides, most subjects did not bring up this subject when first asked about their symptoms.
The next highest agreement was reached on the "over-all" guess regarding the subject's standing on ethnocentrism; then came Intraception, Types of Symptoms, and Character Traits, with agreements around So per cent.
The average agreement for all seven variables was So per cent.
There are several possible reasons why the over-all rating had so high a reliability. One reason is that the instructions prohibited "neutral" ratings in this instance. Another is that the category "over-all" highness-lowness, is a broad one, and the raters are thus given the opportunity to utilize a great variety of explicit or nonexplicit cues and impressions; that they should do this was favored by the fact that both A and B had had experience inter- viewing high and low scorers, possibly developing thereby a "feeling for" a general "high" or "low" personality factor.
Analysis of the ratings assigned by the two judges showed that rater B had a relatively greater number of omissions (meaning "I can't tell from the data given") whereas rater A had tried hard to come to a decision, even when the data offered only one subtle cue. As can be seen from Table 9(XXII), what lowered the agreements between A and B were usually instances in which one rater gave a neutral score; there were very few cases in which one gave a high, the other a low rating.
Because the ratings of A and B were so similar, and because of B's con- sistently greater number of neutrals, which lowered all reliability (and validity) figures somewhat, only rater A's ratings were compared with those made by the 7 control raters.
Table w(XXII) shows the percentage agreements between the ratings by
A and those made by the 7 control judges, each of whom rated only one category. As the control raters made no "over-all" guesses of highness or lowness on E, no agreement with A's over-all rating could be obtained. In- stead, a composite "high" or "low" score for each subject was derived from the 7 control ratings of single variables. A rating of "high" on a given variable was counted as one point, a "neutral" rating was given ? 2 point, a "low" ? rating, o points. By adding the points for each subject, scores ranging from o
to 7 were obtained. All subjects receiving such a composite score of 4 or more were then classed as "high," those with scores below 4 as "low. " The agreement between A's "over-all" rating with these composite ratings is shown near the bottom of Table w(XXII).
The composite high-low score agreed with A's over-all estimate of high- ness or lowness S5 per cent of the time. Practically the same figure was obtained when comparing A's and B's over-all guesses.
The average agreement between A and the control raters for the single
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 929
variables was only very slightly lower (77 per cent) than the average agree- ment between A and B (So per cent). On the whole, the reliabilities obtained were quite acceptable.
In general, those variables which had the highest reliability when the ratings of A and B were compared also had a relatively high percentage agreement in the comparison between A and the control raters, and conversely with the variables of relatively low reliability. The two categories in which this relationship was almost reversed were: categories I (Emphasis on Physical Symptoms) and II (Intraception). The reliability of the former ranked sec- ond best in the case of A and the control raters, about fifth in the case of A and B. The reliability of the latter ranked sixth with A and the controls, second best with A and B.
The relatively low agreement between A and B on category I was caused not so much by disagreements but by a relatively large proportion of "neu- tral" scores (omissions) on the part of B (see Table 9(XXII)). While the control raters and A felt able to judge the presence or absence of certain cues in the record, rater B frequently felt that the subjects' attitudes toward their symptoms were not sufficiently brought out in the interviews.
The rela"tively low agreement between the control rater and A on Intra- ception was due mainly to a larger number of disagreements. The control rater was in this case particularly dissatisfied with her ratings, feeling that she did not have a sufficient grasp of the concept of intraception nor enough acquaintance with the cues by which the trait could be recognized. In the case of A and B the concept and manifestations of intraception had been made the object of special study and played an important part in their thinking. Rater A thought that her ratings of any given case-on over-all highness- lowness and on all other variables-had been more influenced by her impres- sion of the subject's intraceptiveness than by any other cue. It is reasonable to assume that this difference in training is the cause of the difference in reli- ability between the two sets of raters. This becomes even more probable when the reliability figures are compared with the agreements between rat- ings and actual E score. Here, A's and B's ratings of presence or absence of intraception were related to high and lowE score (in the expected direction) So to S3 per cent of the time, while the control rater's judgment agreed with E only 65 per cent of the time (see Table 12 (XXII)).
The highest reliabilities (91 per cent and 95 per cent) were obtained for variable V (Patient Mentions Unhappy Childhood, etc. ). It is clear that there is not much room for disagreement here.
The lowest reliabilities were obtained for variable III (Ego-alienness). Here the agreement between A and the control rater was only 65 per cent, t)lat between A and B 70 per cent. The former figure barely meets the stand- ard of acceptable reliability of measurement. There was an unusually large discrepancy between A's and the control rater's estimate for the high scorers,
? 93?
THE AUTHORITARIAN PERSONALITY TABLE 10 (XXII)
THE AMOUNT OF AGREEMENT BETWEEN A SINGLE RATER (A) AND SEVEN OTHER RATERS IN ESTIMATING VARIABLES IN INI'AKE INI'ERVIEWS
PSYCHIATRIC CLINIC PATIENTS: MEN AND WOMEN COMBINED (N = 59)
Variable
I. Main emphasis on somatic complaints
II. Intraception
III. Ego-alienness
IV. Externalized theory of onset and causes
v. Spontaneous mention of unhappy childhood and family relations
Hby N by
4 1
1 2
4 2
One; Other
VI. CUes regarding charac- High ter structure Low
Agreement between A's over-
all H-L rating and composite
score based on ratings of 7
independent raters, each High 21 rating a single variable LOW 5
VII. Predominant type of High
22 --
Actual Standing Rated
on E Hby Scale Both
High 12 Low 4
High 18 Low 4
High 9 Low 4
Rated
Rated Rated Nby Hby One; Both L by Other
- 2 1 4
2 6 1 8
1 10 2 5
1 4 4 2
- 1 - 1
1 4 1 2
High 14 6 Low 4 4
High 26 - LOW 17 1
3 symptoms Low41-7
Averages High 16. 0
2. 9
11 5 3 2
0. 7 ? Low 5. 7 1. 9 1. 3 ?
5 4
? Rated Rated
No. of Agreements
T otal Percentage Agreement
81. 4
72. 0
65. 3
73. 7
L by L by Both N by
6 4 20 1
2 2 15 3
1 2 10 7
1 - 11 1
2 5 16 7
12 181
2 22
Percentage Agreement
18 + 4 = 22
25 + 1 = 26 83. 9
-
20 + 1 = 21 42. 5
20+1. 5=21. 5 69. 4
1 15 1
75. 0
1. 9 2. 3 15. 0 3. 0
23 27
21. 1 24. 5
50
45. 5
82. 1 84. 7 87. 1
75. 5 77. 2 78. 8
PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 93 I
PSYCHIATRIC CLINIC PATIENTS: MEN AND WOMEN COMBINED (N =59)
One Other
H and L Respectively
H and L Combined
12 + 3 = 15 38. 5
21+2. 5=23. 5 75. 8
16+4=20 43. 5 18+5. 5=23. 5
71. 4 75. 8
27 29
56 96. 4 94. 9 93. 5
48 78. 6
14+5=19 43. 5 67. 9 73. 7 20+4. 5=24. 5 79. 0
244785.
Sometimes the statement of the problem appears more specific (e. g. , "can't do such and such work") but the main thing about it is the idea that "one
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 92 I
cannot do one's duty" (for instance, as a good wife or mother) due to this unwelcome illness. The concern is not so much with the task itself or the feeling of failure involved in the present inability to carry it out, but rather with the violation of conventions and morals. But there is little awareness of specific conflicts, fears, frustrated desires, or life failures.
III. "Ego-alienness"
Presence (goes with high score); The symptoms and the patients' atti- tudes toward them give evidence of particularly strong repressions. The repressed problems and also some of the symptoms have a strong ego-alien quality. Certain impulses, problems, and even some symptoms are experienced as completely "foreign" to the self. They cannot be accepted or admitted. They "belong to the body," or they are the "breaking through" of a com- pletely unacceptable part of the personality. The person "cannot understand what makes me do that. " There is fear of something "devilish inside" that overwhelms the normal accepted self. Fear of "losing control. " Examples: "Fear I'll kill someone during one of my spells; fear I'll lose my mind. "
When certain impulses have broken through and have been expressed in behavior, this ego-alienness is expressed sometimes in moralistic statements and self-accusations ("Can't understand how I could do this"). Another criterion might be the blandness of the interview, stemming from the patient's inability to admit socially unacceptable impulses, either because of repression or because of fear of disapproval from the interviewer.
Absence: The symptoms are experienced as belonging to one's personality and life history. There may be severe conflict over some impulses, but the latter are admitted to consciousness and understood as part of one's self and life experiences. There is relative frankness and freedom from "moralism. "
In the case of obsessions, compulsions, delusions, and other psychotic manifestations, which usually have some ego-alien quality, the rating will have to be based on the relative emphasis on the ego-alienness itself ("Some- thing makes me do it; someone influences me; can't control it," etc. ), or on the degree to which the content of the symptom is consciously divorced from or related to the patient's past or present inner life.
IV. Externalized Theory of Onset and Causation of the Illness
Presence (goes with high E scores): Attempts at externalization of the symptoms by: (a) denying any and all precipitating factors. Tensions, de- pressions, etc. , appear "without any reason at all. " "It just appeared. " (b) dating the illness back to a very definite event or moment: an operation, a "spell," a death, a particular day-sometime during the last few weeks or years. "Before that I was well . . . happy . . . . " (c) making no attempt to relate the illness to one's past, especially not to childhood. No spontaneous references to childhood unless specifically questioned, during this first inter- vrew.
Absence: Spontaneously, or when asked about onset, states that the prob-
? THE AUTHORITARIAN PERSONALITY
lems have been present "for a long time"-for years, always or since child- hood, but perhaps in milder form.
V. Spontaneous Mention of Unhappy Childhood or Family Relationships
Presence (goes with low E scores): Patient spontaneously refers to his unhappy childhood. Often elaborates in great detail on_a history of frustra- tions (often with a clearly masochistic attitude). Complains of parents' pres- ent or past attitudes and is critical of them.
Absence: No spontaneous reference to childhood. Childhood, if discussed at all, is pictured as "happy and normal. " No criticism or other sign of hos- tility towards family members is expressed (except in "spells" and psychotic episodes), in spite of admission of friction. Sometimes there is clearly com- pensatory great concern for family members and their welfare.
VI. Cues Referring to the Patient's Character Structure
Obviously the interview fragments cannot be used to construct complete pictures of the subject's personality structure. It was found, however, that they often contained some important hints as to the nature of this structure. Formulation of categories listed below was, of course, guided by the con- cepts found most useful for distinguishing high and low scorers in the rest of the study. But only those categories were included which would be applied to the particular interview material at hand. The raters were asked to state for each case whether cues from List A or List B predominated in a record, or whether the record was "neutral" (meaning he could discern equally as many cues from "A" as from "B" or none at all).
List A (goes with high degree of ethnocentrism):
1. "Countercathectic" defenses. Men: counteraction against passivity.
Women: Anal reaction-formations. No mention of a love object or re-
lationship (other than spouse).
2. Extrapunitive or impunitive. 6 If self-criticism occurs, it is couched in
moralistic terms, "I am bad, have done wrong. " "God is punishing me. " Tendency either hysterically to dramatize certain physical symptoms, spells, etc. , or to minimize all complaints of a psychological nature.
3? Externalized superego. Religion: God is seen as an external judge of one's action. Illness is a punishment from God. Or religious standards are part of the conventional ideology. Religious reasons are given for refraining from divorce or suicide.
List B ("low" characteristics):
4? Absence of "countercathectic" defenses:
Direct expressions of orality: dependent character traits; eating, drink- ing, drugs. Dependency problems, nurturance. Open admission of weak-
6 Further research onirnpunitiveness is indicated. Here it seemed preferable to classify it "high;" it was "low" in Chapter XI (p. 406).
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 92 3
ness, passivity, femininitY in men. Love-seeking attitude. References to rejections by specific love-objects.
Sublimations: form an important part of the ego; references to achieve- ment. When symptoms interfere particularly with patient's work, this is stated in such a way as to suggest concern with the particular work he is doing. Emphasis on ideals. Concern with helping others, society; artistic activity and interest.
5? Intrapunitiveness, masochism: "I was mistreated as a child. " Identifica- tion with suffering; self-criticism ("I am a failure").
6. Internalized superego: guilt feelings, true depressions, religious ideas as part of the inner life of the person.
VII. Predominant Types of Symptoms
List A ("high"):
1. Physical anxiety symptoms and other emotional equivalents: "hyper-
ventilation syndrome"; dizziness, sweating, tingling sensations, numb- ness, tachycardia, breathlessness, fainting, tremors. Anxiety has no con- scious content.
2. Emotional outbursts, tantrums and "spells," in women.
3? Hypochondriacal fears: fear of death, heart attack, etc.
4? Hysterical conversions. 7
5? "Rigid compulsive rumination": repetitious complaints, self-accusa-
tions, self-reassurances, "thinking around and around in a circle. "
? 6. Depersonalization (sense of estrangement from self and world) in a person who emphasizes that heretofore he had had no tendency toward
timidity and withdrawal.
7. Suspiciousness, fear of people or aversion to people is stated in some-
what externalized terms: "They irritate me," "I can't stand them. They make me nervous. " (To be distinguished from hypersensitiveness and withdrawal when described in a more intraceptive way. )
8. Psychopathic tendencies, not in the sense merely of unconventional behavior (as the term is sometimes used) but rather in the sense of a really defective and not sufficiently internalized superego. Antisocial and destructive behavior, callousness, emotional shallowness.
9? Stomach ulcers in men, especially in subjects who emphasize their masculinity.
List B:
ro. Depressed mood, hopelessness, lack of self-confidence, verbalized feel- ings of inadequacy, suicidal ideas, guilt. Often patients complain of
7 This item was included in the manual used by our raters; but later analysis of case material suggested that there may be more frequent hysterical conversions in low scorers. The trends are as yet not clear. Perhaps there is also a sex difference here. Two of the low-quartile men had conversion symptoms.
? II.
"depression," but the "true" depressions as described above have to be distinguished from the more schizoid type of mood disturbance. Tiredness. "Neurasthenia. "
Dismenorrhea.
THE AUTHORITARIAN PERSONALITY
I2.
I 3?
I4? In men, expression of traits opposite to the culturally emphasized
I 5?
masculine pattern. Withdrawal, timidity, shyness, sometimes coupled with feelings of unreality or with physical anxiety symptoms or hys- terical conversion. The patient's attitude toward all these "weaknesses" is to some degree accepting.
Conscious anxiety and conflicts.
3. THE METHOD OF QUANTIFICA TION
Conscious rejection of feminine role by women.
Inspection of the statements of complaint led to the impression that the high and low quartile groups were clearly differentiated with regard to the variables just discussed. In order to check this impression in a more rigorous manner, it was decided to use a method of "blind ratings," similar to that used on the interview, T. A. T. and Projective Question Test material (Parts II and III). At least two judges who are not acquainted with the subjects (and in this case did not know which were high and which were low scorers)
independently rate each subject on certain characteristics. These ratings can then be compared with other data on the subjects. The problem of blind ratings has been discussed in other chapters dealing with interview scoring and projective questions. If these ratings turn out to be highly correlated with some trait of the subjects (in this case ethnocentrism) on which the judges had no information, it is very probable that relationships between the former and latter traits exist in reality and could be demonstrated by other methods. (This holds only if the rating procedure is so controlled as to pre- vent the raters from utilizing cues other than those to be experimentally tested. ) There are, however, various possible pitfalls inherent in the rating method. Such "errors of measurement" could arise, for instance, from material which did not contain sufficient information from which to judge the subject on a certain trait. Other errors might arise from the manner in which the characteristics were defined and described for the judges; from the ability of the judges to apply the instructions to the material to be judged -depending in turn on the judges' training, theoretical bias, and personality- and from other factors. Use of a rating technique, therefore, requires not only the determination of certain relationships between the ratings and other data, but also a careful consideration of various factors which could have influenced the results either in the positive or in the negative direction.
a. THE RATERS. The two primary raters were staff members of the major study and will be referred to as raters (or judges) A and B. In addition to much clinical training and experience, these two raters had a strong psycho-
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 925
analytic orientation. Both were familiar with all concepts, hypotheses, data,
and results of the total study. Both had had opportunity to interview high-
and low-scoring subjects, and were therefore acquainted with the behavior
I
and material usually obtained from such subjects in interview situations. Each of the judges independently rated each record on all seven cate- gories and then assigned an "over-all" intuitive estimate of "highness" or
"lowness. "
After these primary ratings had been completed, 7 independent raters
(they are referred to hereafter as the control raters, their ratings as control ratings) were used, each rating one category only. 8 Six of the control raters were clinical psychologists (of these, r was a senior clinician, 5 junior clini- cians at the level of internes, working at the Langley Porter Clinic). They were not acquainted with the concepts and data of the over-all study. This was important for reasons to be discussed below. One of the more narrowly defined categories (Onset and cause of illness are explained by subject in externalized terms) was rated by our office secretary who had had no formal training in psychology or psychiatry, but who had much intuitive psycho- logical insight and who had absorbed a great deal of the research material and hypotheses. The 7 judges varied greatly in age, training, and theoretical orientation.
b. THE RATING TECHNIQUE. The instructions for the control raters were as follows:
The material to be rated consists of "Statements of Complaint" by Langley Porter Clinic patients in a first intake interview or in the first interview with a therapist. Only the section "patient's story" or "chief complaint" was included. The inter- views are here reproduced verbatim, although a few have been slightly condensed by the writer. Each numbered paragraph refers to one case. There are 26 men, 33 women.
Each case is to be rated on one variable (or syndrome) as described in the manual. Each rater will be assigned one variable and will not know about other variables until he has completed his ratings.
The ratings are to be made in terms of presence (v) or absence (-) of the trait. A few of the variables permit of a "mixed" judgment. Assignment of such a "mixed" (M) rating should be avoided if possible. But occasionally it may have to be used. Sometimes (due to the fragmentary way in which some of these interviews are
recorded) there will be insufficient material to rate. In this case mark (o).
Each primary rater (A and B) first rated each record in terms of all seven individual categories. They knew which categories were expected to be related to high or to low ethnocentrism. They therefore tried to assign each record a "high" or "low" rating for each category. Often a record did not contain enough material to permit the rater to reach a decision on a given variable, e. g. , a given topic was not discussed, or there were few cues permit-
s W e wish to thank Dorothy Bomberg, Janet Gist, Carole Home, Virginia Patterson, Dr. Claire W. Thompson, Anne Vollmar, and Elaine Wesley Barron for the patience and care with which they carried out, on short notice, the task of doing the control ratings.
? THE AUTHORIT ARIAN PERSONALITY
ting inferences regarding character structure. In this case no rating was assigned. When there were about an equal number of cues pointing in the high and in the low directions, a "mixed" rating was assigned. The raters then went over the records a second time, trying to guess in each case whether the subject had made a high or low score on theE scale. The guesses were to be based on the decisions reached regarding the individual categories. However, no mechanical formula was set up to convert the individual ratings into "over-all" ratings. The raters arrived at the latter by a new rating process in which any or all of the categories could be used and weighted as the rater saw fit.
Two types of data were obtained from this rating material: (I) Inter- rater scoring agreement for each category and for the over-all ratings. (2) Relationship between ratings and scores on the E scale. These will now be discussed.
4. T H E RELIABILITY OF T H E MEASURES
When several persons agree considerably more than half of the time that certain subjects in a group do, others do not, possess a given trait, the chances are good that these various raters knew what they were supposed to look for, had a similar conception of the trait, understood this concept, and could clearly recognize something in the interview data to which this concept could be applied; and that personality, training, and other differences be- tween the raters influenced the ratings only to a relatively small degree.
All rating notations (high, low, presence, absence, omission, mixed) were converted into "high," "low," and "neutral" scores. E. g. , a rating of "pres- ence" on variable I-Main emphasis on subject's physical complaints-was- considered a "high" score, "absence" a "low" score; "mixed" notations and omissions were considered "neutral" scores.
Scoring reliability was then obtained by computing the percentage of times 2 raters had assigned the same scores to the same records. Whenever both raters had assigned exactly the same score (high, low, or neutral) to the same record, this was considered one agreement. When one of the raters had given either a high or a low, the other a neutral score, this was considered one-half an agreement. When one rater gave a high score, the other a low, this was counted a full disagreement. The number of agreements, divided by the total number of records rated, yielded the percentage agreement between 2 raters. There were very few instances in which both judges gave a neutral score.
Table 9(XXII) shows the percentage agreements between Raters A and B as well as the scores on which these figures are based. All of the percentage agreements, except one (category III, "lows"), are above 70, statistically higher than could have been obtained by chance (I per cent level).
Raters A and B agreed best, 9I per cent, on variable V (Subject spon-
? Over-all rating of highness or lowness on E Scale
Single Variables:
I. Main emphasis on somatic
complaints
II. Intraception
III. Ego-alienness
Low 4. . 3 High 1551 .
Low
51 82. 1 86. 4 90. 3
45 62. 1 76. 3. 71. 0
48 78. 6 83. 1 83. 9
41. 5 76. 8 70. 3 64. 5
45 75. 0 76. 3 n. 4
1 .
13 1 28. 5 91. 9
IV. Externalized theory of onset High and causes Low
13 83 4 26
Low322 1
High 15'72 1 Low 1 4 3 1
-
2 2
25 914
. 3 184
2 10 10
. 2 9 8
23
22
22 26
21. 5 20
21 24
High 14111 Low321 5
.
. . . 2
'7 2 2 2 . 2
VII. Predominant type of symptoms High 18
Low 4 - 1 4
v. Spontaneous mention o t IJI! haPPY High
childhood and family relations ! LOW 15
VI. Cues regarding character High
structure Low 3
2
25 53. 5 89. 3 90. 7
23 2
10
2 20
520 46 '71. 4 '78. 0 4 26 83. 9
TABLE 9 (XXII)
DIE AMQUNT OF AGJ! EE)IENT BETWEEII TWO RATmS IN ESTIMATING A SUBJECT' S STANDING ON DIE E SCALE FROM AN ANALYSIS OF HIS INTAKE INTERVIEW
PSYCHIATRIC CLINIC PATIENTS: MEN AND WOMEN COMBINED (N = 59; High scorers: N = 28; Low Scorers: N = 31)
No. scored High (H) No. scored Neutral (N No. scored Low (L)
Interrater Agreementa
No. of Agreements Percent Agreements
Actual
Standing (! )Rated (2)Rated H (3)Rated N (4)Rated H (5)Rated L (6)Rated L ('! )High (8)H and L (9)High (IO)H and L
onE H by by One; N by Both by One;\ L by Both by One: N or Combined or Combined
Scale Both by Other bY Other High 21 - - 5
by Other Low 2 . 23
413
1
Mean agreement on the variables High 15. 4 6. 3 1. 4 1. 4 0. 9 2. 6 22. 2 Low 4. 8 I. 7 1. 9 2. 4 14. 0 6. 3 24. 7
! ! Number of agreements for a given variable is the sum of the values, opposite that variable, in Columns 1. 3, and 5.
24 . 28
1 22. 5 19 3 25. 5
48 80. 3 81. 4 82. 3
'79. 3 79. 5 79. 7
? THE AUTHORITARIAN PERSONALITY
taneously and explicitly mentions unhappy childhood and family relation- ships). This result was to be expected because this variable is so unambiguous and requires little subjective evaluation. Besides, most subjects did not bring up this subject when first asked about their symptoms.
The next highest agreement was reached on the "over-all" guess regarding the subject's standing on ethnocentrism; then came Intraception, Types of Symptoms, and Character Traits, with agreements around So per cent.
The average agreement for all seven variables was So per cent.
There are several possible reasons why the over-all rating had so high a reliability. One reason is that the instructions prohibited "neutral" ratings in this instance. Another is that the category "over-all" highness-lowness, is a broad one, and the raters are thus given the opportunity to utilize a great variety of explicit or nonexplicit cues and impressions; that they should do this was favored by the fact that both A and B had had experience inter- viewing high and low scorers, possibly developing thereby a "feeling for" a general "high" or "low" personality factor.
Analysis of the ratings assigned by the two judges showed that rater B had a relatively greater number of omissions (meaning "I can't tell from the data given") whereas rater A had tried hard to come to a decision, even when the data offered only one subtle cue. As can be seen from Table 9(XXII), what lowered the agreements between A and B were usually instances in which one rater gave a neutral score; there were very few cases in which one gave a high, the other a low rating.
Because the ratings of A and B were so similar, and because of B's con- sistently greater number of neutrals, which lowered all reliability (and validity) figures somewhat, only rater A's ratings were compared with those made by the 7 control raters.
Table w(XXII) shows the percentage agreements between the ratings by
A and those made by the 7 control judges, each of whom rated only one category. As the control raters made no "over-all" guesses of highness or lowness on E, no agreement with A's over-all rating could be obtained. In- stead, a composite "high" or "low" score for each subject was derived from the 7 control ratings of single variables. A rating of "high" on a given variable was counted as one point, a "neutral" rating was given ? 2 point, a "low" ? rating, o points. By adding the points for each subject, scores ranging from o
to 7 were obtained. All subjects receiving such a composite score of 4 or more were then classed as "high," those with scores below 4 as "low. " The agreement between A's "over-all" rating with these composite ratings is shown near the bottom of Table w(XXII).
The composite high-low score agreed with A's over-all estimate of high- ness or lowness S5 per cent of the time. Practically the same figure was obtained when comparing A's and B's over-all guesses.
The average agreement between A and the control raters for the single
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 929
variables was only very slightly lower (77 per cent) than the average agree- ment between A and B (So per cent). On the whole, the reliabilities obtained were quite acceptable.
In general, those variables which had the highest reliability when the ratings of A and B were compared also had a relatively high percentage agreement in the comparison between A and the control raters, and conversely with the variables of relatively low reliability. The two categories in which this relationship was almost reversed were: categories I (Emphasis on Physical Symptoms) and II (Intraception). The reliability of the former ranked sec- ond best in the case of A and the control raters, about fifth in the case of A and B. The reliability of the latter ranked sixth with A and the controls, second best with A and B.
The relatively low agreement between A and B on category I was caused not so much by disagreements but by a relatively large proportion of "neu- tral" scores (omissions) on the part of B (see Table 9(XXII)). While the control raters and A felt able to judge the presence or absence of certain cues in the record, rater B frequently felt that the subjects' attitudes toward their symptoms were not sufficiently brought out in the interviews.
The rela"tively low agreement between the control rater and A on Intra- ception was due mainly to a larger number of disagreements. The control rater was in this case particularly dissatisfied with her ratings, feeling that she did not have a sufficient grasp of the concept of intraception nor enough acquaintance with the cues by which the trait could be recognized. In the case of A and B the concept and manifestations of intraception had been made the object of special study and played an important part in their thinking. Rater A thought that her ratings of any given case-on over-all highness- lowness and on all other variables-had been more influenced by her impres- sion of the subject's intraceptiveness than by any other cue. It is reasonable to assume that this difference in training is the cause of the difference in reli- ability between the two sets of raters. This becomes even more probable when the reliability figures are compared with the agreements between rat- ings and actual E score. Here, A's and B's ratings of presence or absence of intraception were related to high and lowE score (in the expected direction) So to S3 per cent of the time, while the control rater's judgment agreed with E only 65 per cent of the time (see Table 12 (XXII)).
The highest reliabilities (91 per cent and 95 per cent) were obtained for variable V (Patient Mentions Unhappy Childhood, etc. ). It is clear that there is not much room for disagreement here.
The lowest reliabilities were obtained for variable III (Ego-alienness). Here the agreement between A and the control rater was only 65 per cent, t)lat between A and B 70 per cent. The former figure barely meets the stand- ard of acceptable reliability of measurement. There was an unusually large discrepancy between A's and the control rater's estimate for the high scorers,
? 93?
THE AUTHORITARIAN PERSONALITY TABLE 10 (XXII)
THE AMOUNT OF AGREEMENT BETWEEN A SINGLE RATER (A) AND SEVEN OTHER RATERS IN ESTIMATING VARIABLES IN INI'AKE INI'ERVIEWS
PSYCHIATRIC CLINIC PATIENTS: MEN AND WOMEN COMBINED (N = 59)
Variable
I. Main emphasis on somatic complaints
II. Intraception
III. Ego-alienness
IV. Externalized theory of onset and causes
v. Spontaneous mention of unhappy childhood and family relations
Hby N by
4 1
1 2
4 2
One; Other
VI. CUes regarding charac- High ter structure Low
Agreement between A's over-
all H-L rating and composite
score based on ratings of 7
independent raters, each High 21 rating a single variable LOW 5
VII. Predominant type of High
22 --
Actual Standing Rated
on E Hby Scale Both
High 12 Low 4
High 18 Low 4
High 9 Low 4
Rated
Rated Rated Nby Hby One; Both L by Other
- 2 1 4
2 6 1 8
1 10 2 5
1 4 4 2
- 1 - 1
1 4 1 2
High 14 6 Low 4 4
High 26 - LOW 17 1
3 symptoms Low41-7
Averages High 16. 0
2. 9
11 5 3 2
0. 7 ? Low 5. 7 1. 9 1. 3 ?
5 4
? Rated Rated
No. of Agreements
T otal Percentage Agreement
81. 4
72. 0
65. 3
73. 7
L by L by Both N by
6 4 20 1
2 2 15 3
1 2 10 7
1 - 11 1
2 5 16 7
12 181
2 22
Percentage Agreement
18 + 4 = 22
25 + 1 = 26 83. 9
-
20 + 1 = 21 42. 5
20+1. 5=21. 5 69. 4
1 15 1
75. 0
1. 9 2. 3 15. 0 3. 0
23 27
21. 1 24. 5
50
45. 5
82. 1 84. 7 87. 1
75. 5 77. 2 78. 8
PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 93 I
PSYCHIATRIC CLINIC PATIENTS: MEN AND WOMEN COMBINED (N =59)
One Other
H and L Respectively
H and L Combined
12 + 3 = 15 38. 5
21+2. 5=23. 5 75. 8
16+4=20 43. 5 18+5. 5=23. 5
71. 4 75. 8
27 29
56 96. 4 94. 9 93. 5
48 78. 6
14+5=19 43. 5 67. 9 73. 7 20+4. 5=24. 5 79. 0
244785.
