91,
identical
with that for the total group.
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
Minor forms of military training, obedience, and discipline, such as drill, marching and simple commands, should be made a part of the elementary school educational program.
12. The main threat to basic American institutions during this century has come from the infiltration of foreign ideas, doctrines, and agitators. ?
13. Present treatment of conscientious objectors, draft evaders, and enemy aliens is too lenient and mollycoddling. If a person won't fight for his . country, he deserves a lot worse than just prison or a work camp.
18. In view of the present national emergency, it is highly important to limit responsible government jobs to native, white, Christian Americans.
20. European refugees may be in need, but it would be a big mistake to lower our immigration quotas and allow them to flood the country.
23. It has become clear that the Germans and Japanese are racially war-minded and power-seeking, and the only guarantee of future peace is to wipe out most of them and to keep the rest under careful control.
26. Mexico can never advance to the standards of living and civilization of the U. S. , due mainly to the innate dirtiness, laziness, and general backwardness of Mexicans.
33? There will always be wars because, for one thing, there will always be races who ruthlessly try to grab more than their share.
toward much that he cannot personally accept for himself. He is free of rigid conformism, outgroup rejection, and imperialistic striving for power. lngroup opinions and attitudes are expressed in Items 3, 7, and 10. They are intended to express a general value for obedience and discipline, the opinion that nations are arranged hierarchically from superior to inferior, and the attitude that the superior ones should be dominant-with the assump- tion that we are one of the superior nations. The rigidity of the value for obedience is shown by the punitive attitude toward those who disobey (Item
I 3: Punishment of conscientious objectors and draft evaders).
The glorification of the national ingroup is shown further in the tendency to regard other nations as inferior when they are distant (Item 26), and threatening when they come too close (Items I 2, 20, and 2 3). W e are there- fore morally justified in excluding refugees, in "wiping out" the Germans and Japanese,2 in excluding foreigners and others from government jobs,
2 This item (23), so relevant during the war, can of course no longer be used. Ot should be pointed out that one could actively support the war without such a destructive attitude toward the enemy or such national smugness. ) I f these attitudes are correlated with rejec- tion of most other nations, then the people who made high (ethnocentric) scores on this scale may be the ones who now show similar attitudes toward our wartime allies and sup- port militaristic, imperialistic, "tough-minded policies guaranteeing American sovereignty and interests. "
? THE STUDY OF ETHNOCENTRIC IDEOLOGY 109
and in maintaining our dominant position as a nation. The cynicism about peace and the moralistic attribution of war to "ruthless, grabbing races" also indicate the sense of thre~t from outgroups and the moral righteousness of the ingroup. The fact that this theory of the cause of war is held by many college students who have been exposed to sociological explanations in terms of socioeconomic organization and conflicts raises the question: What are the inner barriers in some individuals which make them unreceptive to non- moralistic explanations? This problem, to be taken up in later chapters, concerns the personality dynamics underlying ethnocentric ideology.
If people who make high scores on this subscale are also high on the others, then it would appear that although they hold America to be superior and inviolable, they actually reject the great majority of the people in this coun- try. Item I 8 brings this out directly: it is only the native, white, Christian Americans who can be trusted. And various items from the "Minorities" subscale indicate that large sections of this population are also in the out- group category.
2. THE TOT AL ETHNOCENTRISM (E) SCALE
The total E scale is intended to measure the individual's readiness to accept or oppose ethnocentric ideology as a whole. The scale consists of 34 items3 and comprises the three subscales N, M, and P. It is presented in Table 4 (IV), with instructions to subjects, just as it was administered.
C. RESULTS: STATISTICAL ANALYSIS OF THE SCALE
The subjects were allowed the same six choices of response for each item (+3 to -3, with no neutral response), and the responses were converted into scores in the same way (-3= I point, -2 = 2 points, etc. ) as was the case with the A-S scale. All the items were regarded as pro-ethnocentric. For the 34 items, then, the total scores can range between 34 points (I point on each item, indicating strong anti-ethnocentrism) and 2 38 points (7 points on each item, strong ethnocentrism). When the total score is divided by 34 we obtain the mean score per item; thus, a total score of 5I can also be stated as
a mean per item of 1. 5.
This scale was administered as part of the questionnaire which also con-
tained the initial (52 item) A-S scale. As reported in Chapter III, this ques- tionnaire was given in April, I944? to a class in Introductory Psychology at the University of California. The data presented here are based on the ques- tionnaires of the I44 women subjects, including nineteen members of major minorities.
3 Items r, 3, 4, 7, 9, 10 of the present scale were taken, in some cases with slight modifica- tions, from the "Unlabelled Fascist Attitudes" Scale of Edwards (22). Both Edwards' study and the present one have profitted from previous studies by Gundlach (46), Katz and Cantril (17), and Stagner (rn, rq). Several of the "Negro" items have been taken from Murphy and Likert (84).
? I IO THE AUTHORITARIAN PERSONALITY
TABLE 4 (IV)
THE ToTAL ETHNocENTRISM ScALE Public Opinion Questionnaire E
The following statements refer to opinions regarding a number of social groups and issues, about which some people agree and others disagree. Please mark each statement in the left-hand margin according to your agreement or disagreement, as follows:
+ r : slight support, agreement +2: moderate support, "
+3: strong support, "
- 1 : slight opposition, disagreement -2: moderate opposition, "
-3: strong opposition, "
_ _ _ 1.
___ 2.
3? 4?
5?
6.
___7?
___ 8. ___9?
___IO.
___I1. ___12. ___IJ.
_ _ _ 14.
The many political parties tend to confuse national issues, add to the expense of elections, and raise unnecessary agitation. For this and other reasons, it would be best if all political parties except the two major ones were abolished.
If there are enough Negroes who want to attend dances at a local dance hall featuring a colored band, a good way to arrange this would be to have one all-Negro night, and then the whites could dance In peace the rest of the time.
Patriotism and loyalty are the first and most important requirements of a good citizen.
Certain religious sects whose beliefs do not permit them to salute the flag should be forced to conform to such a patriotic action, or else be abolished.
The Negroes would solve many of their social problems by not being so irresponsible, lazy, and ignorant.
Any group or social movement which contains many foreigners should be watched with suspicion and, whenever possible, be investigated by the FBI.
There will always be superior and inferior nations in the world and, in the interests of all concerned, it is best that the superior ones be in con. trol of world affairs.
Negro musicians are sometimes as good as white musicians at swing music and jazz, but it is a mistake to have mixed Negro-white bands. Although women are necessary now in the armed forces and in indus- try, they should be returned to their proper place in the home as soon as the war ends.
Minor forms of military training, obedience, and discipline, such as drill, marching and simple commands, should be made a part of the elementary school educational program.
It would be a mistake to have Negroes for foremen and leaders over whites.
The main threat to basic American institutions during this century has come from the infiltration of foreign ideas, doctrines, and agitators. Present treatment of conscientious objectors, draft-evaders, and enemy aliens is too lenient and mollycoddling. If a person won't fight for his country, he deserves a lot worse than just a prison or a work camp. Negroes may have a part to play in white civilization, but it is best to keep them in their own districts and schools and to prevent too much intermixing with whites.
One main difficulty with allowing the entire population to participate
__15.
? __I6. _ _ 17. __I8.
___19.
___20. __2I.
__22.
__23.
_ _ 24.
fully in government affairs (voting, jobs, etc. ) is that such a large percentage is innately deficient and incapable.
Manual labor and menial jobs seem to fit the Negro mentality and abil- ity better than more skilled or responsible work.
It is a mistake to allow any Japanese to leave internment camps and enter the army where they would be free to commit sabotage.
In view of the present national emergency, it is highly important to limit responsible government jobs to native, white, Christian Ameri- cans.
In a community of I,ooo whites and 50 Negroes, a drunken Negro shoots and kills an officer who is trying to arrest him. The white pop- ulation should immediately drive all the Negroes out of town.
European refugees may be in need, but it would be a big mistake to lower our immigration quotas and allow them to flood the country.
The many faults, and the general inability to get along, of the Okla- homans ("Okies"), who have recently flooded California, prove that we ought to send them back where they came from as soon as condi- tions permit.
The people who raise all the talk about putting Negroes on the same level as whites and giving them the same privileges are mostly radical agitators trying to stir up conflicts.
It has become clear that the Germans and Japanese are racially war- minded and power-seeking, and the only guarantee of fumre peace is to wipe out most of them and to keep the rest under careful control.
A large-scale system of sterilization would be one good way of breed- ing out criminals and other undesirable elements in our society and so raise its general standards and living conditions.
THE STUDY OF ETHNOCENTRIC IDEOLOGY I I I
_ _ 25. An occasional lynching in the South is a good thing because there
_ _ _ 26.
_ _ 27. _ _ _ 28.
_ _ _ 29.
_ _ 30. - - 3 ! . - - 3 2 ? - - 3 3 ? - - 3 4 ?
is a large percentage of Negroes in many communities and they need a scare once in a while to prevent them from starting riots and dismrb- ances.
Mexico can never advance to the standards of living and civilization of the U. S. , due mainly to the innate dirtiness, laziness, and general backwardness of Mexicans.
Filipinos are all right in their place, but they carry it too far when they dress lavishly, buy good cars, and go around with white girls.
It would be best to limit Negroes to grammar and trade school educa- tion since more schooling just gives them ambitions and desires
which they are unable to fulfill in white competition.
Zootsuiters demonstrate that inferior groups, when they are given too much freedom and money, just misuse their privileges and create disturbances.
The most vicious, irresponsible, and racketeering unions are, in most cases, those having largely foreigners for leaders.
There is something inherently primitive and uncivilized in the Negro, as shown in his music and his extreme aggressiveness.
W e are spending too much money for the pampering of criminals and the insane, and for the education of inherently incapable people. There will always be wars because, for one thing, there will always be races who ruthlessly try to grab more than their share.
Most Negroes would become officious, overbearing, and disagreeable if not kept in their place.
? II2 THE AUTHORITARIAN PERSONALITY
1. RELIABILITY
Data indicating the reliability and related statistical properties of the E scale and its subscales are given in Table 5(IV).
TABLE 5 (IV)
RELIABILITY OF THE E. "I'HNOCENTRIS'If (E) SCALE AND ITS SUBSCALES
Property E Scale Reliabilitya . 91
Negroes
? 91 12
2. 72 2. 65 2. 78
1. 25 1. 25 1. 42
1. 0-5. 6
Minorities
? 82 12
3. 32 3. 23 3. 40
1. 21 1. 37 1. 28
1. 0-6. 0
Patriotism
. 80 10
3. 53 3. 88 3. 18
1. 26 1. 26 1. 46
1. 0-6. 1
Number? of iterns Mean (total) b Mean (odd half) Mean (even half) S. D. (total)b S. D. (odd half) S. D. (even half)
Rangeb
34
3. 17
3. 02 3. 32
t. 15 1. 17 1. 21
1. 2-5. 6
Total Subscales
s. . rhe split-half reliability of each scale was obtained by correlating the sum of the scores on the odd items with the sum of the even items. and correcting this value by the Spearman-Brown formula,
brhe means, s. D. ? s, and ranges are given in terms of mean score per
item on the scale or subscale in question, If this value is multiplied by the number of items in the scale o~ subscale, it is converted into mean total score,
The split-half reliability of the total E scale is . 91, a value which meets accepted statistical standards. 4 The odd and even halves were roughly equiv- alent in the sense that they contained about equal numbers of items from the three subscales. The lower mean of the odd half seems due to the slight over- weighting with low-mean Negro items. The obtained range covered most of the possible range, with the exception of the extremely high end. The absence of very high scores (averages of over 6 points per item) is also reflected in the relatively low group mean of 3? I7, as compared with the neutral point of 4. 0 per item. The distribution of scores is very symmetrical-the mean divides the range in half, and the median is 3. 2-but platykurtic, with very little clustering of scores around the mean.
The high reliabilities of the subscales are noteworthy, especially in view of the small number of items in each.
In terms of reliability, equivalence of halves, and form of distribution,
4On the chance that the 19 minority group members might be atypical in some way, a separate reliability was computed for the 125 remaining subjects. The obtained value was .
91, identical with that for the total group.
? Negroes Minorities Patriotism
Negroes
. 74 . 76
Minorities . 74
. 83
Patriotism
? 76 ? 83
Total E
. 90 . 91 . 92
THE STUDY OF ETHNOCENTRIC IDEOLOGY IIJ
then, it seems safe to conclude that the E scale and its subscales provide ade- quate measuring instruments. To the extent that the scale is valid, it provides a measure of ethnocentrism, in most of its generality and complexity. It may be claimed that the higher an individual's score, the greater his acceptance of ethnocentric propaganda and the greater his disposition to engage m ethnocentric accusatiops and programs of one form or another.
2. INTERCORRELATIONS AMONG THE SUBSCALES
The reliability data support the hypothesis that there is such a thing as general ethnocentric ideology and that people can be roughly ranked ac- cording to the strength of their acceptance or rejection of it. Support for this hypothesis is also given by the high intercorrelations among the sub- scales, as shown in Table 6(1V).
TABLE 6 (IV)
CORRELATIONS OF THE E SUBSCALES WITH EACH 01'H. rn AND WITH THE TOTAL E SCALEa
aThese are the raw correlation coefficients. If they were corrected for attenuation to give the maximal value theoretically obtainable (with perfectly reliable instruments), they would all be . 9 or over.
The subscale intercorrelations, which range from ? 74 to . 83, are of con- siderable significance. The fact that they involve items dealing with so great a variety of groups and ideas suggests again that ethnocentrism is a general frame of mind, that an individual's stand with regard to one group such as Negroes tends to be similar in direction and degree to his stand with regard to most issues of group relations.
The intercorrelations of . 90 to . 92 between each subscale and the total E scale make the same point; an individual's score on any one subscale per- mits one to predict very closely his score on the entire E scale. Or, to put it in another way: While almost every subject shows some variability in his responses to the individual items (as will be shown below), almost every one demonstrates a general degree of pro- or anti-ethnocentrism which is relatively consistent from one group or type of group to another. And ethno- centric hostility toward outgroups is highly correlated with ethnocentric idealization of ingroups.
? THE AUTHORITARIAN PERSONALITY
3. INTERNAL CONSISTENCY: STATISTICAL ANALYSIS OF THE INDIVIDUAL ITEMS
The functions of item analysis, and the procedures involved, have been discussed in the previous chapter. The data on the item analysis of the E scale are presented in Table 7(IV). Each item is identified by a key word or phrase, and the letters N, M, and P refer to the subscales Negroes, Minori- ties, and Patriotism respectively. It will be recalled that the Discriminatory Power (D. P. ) equals the mean for the high quartile minus the mean for the low quartile. The total group mean is, of course, based on all four quartiles.
In general the D. P. 's in Table 7(IV) are very satisfactory,5 averaging 2. 97? For the 34 items, 5 D. P. 's are over 4. 0, 13 are between 3. 0 and 3. 9, and 10 are between 2. 0 and 2. 9; only 3 are between r. o and r. 9, and 3 less than 1. 0. Furthermore, all 6 items with D. P. 's of less than 2. 0 have group means of less than 3. 0, so the D. P. is more significant than it appears. 6
The three least discriminating items are I9, 25, and 28, all in subscale N. They are also the only three items with group means of less than 2. 0. Their low means indicate almost unanimous disagreement by all subjects. This is to be expected, since the items are particularly violent and repressive: Ne- groes should be driven out of town, lynched, kept ignorant and uneducated. But these data show the advantage of permitting three degrees of agreement and of . disagreement, and they also reveal a subtle receptiveness in the high-scoring subjects to openly antidemocratic programs. Of the 36 low scorers only one responded with -2 (on Item 28), all other responses on all three items being a firm -3 (and thus a low mean of r. oo). The high quartile, on the other hand, had a mean of I. 8 on each of the three items; nearly half of them responded with -2 or above. One might ask if this is an indication of potential response during a period in which fascism had become a real power. Not all those who score high onE, certainly, are receptive to violent antidemocracy; the task of determining the deeper psychological forces which make for potential receptiveness or opposition to fascism- the ultimate in ethnocentrism-is one which follows the first task of measur- ing ethnocentrism in its presently existing form.
The item analysis indicates that the N, M, and P subscales contributed about equally to the total differentiation between the high and low quartiles on the total scale, the average D. P. being 3. 0, 2. 9, and 3? I respectively. Apart from items I9, 25, and 28, discussed above, the Negro items were highly dis- criminating. Ethnocentrists and anti-ethnocentrists, as measured by the total scale, are clearly differentiated with respect to most of the ideas contained
5 The D. P. 's would be even higher if the "range of talent" in this group included more extreme ethnocentrists. This is shown by results on subsequently tested groups.
6 While correlations between items or between each item and the total scale have not been computed for this group, later data on similar scales suggest that the average correla- tion between single items is about . 4, while between each item and the sum of the remain- ing items the average correlation is about . 6.
? THE STUDY OF ETHNOCENTRIC IDEOLOGY TABLE 7 (IV)
MEANS AND DISCRIMINATORY PO\VmS OF THE E-SCALE ITEMS
No. Item
1. (M: political parties)
2. (N: dance)
3. (P: patriotism)
4. (M: religious sects)
5. (N: lazy)
6. (M: foreign groups)
7. (P: superior nations)
8. (N: bands)
9. (M: women)
10. (P: military training)
Low Quartile
2. 17 1. 97 3. 86 1. 61 1. 53 1. 69 1. 25 1. 25 3. 75 2. 47 1. 69 1. 22 1. 44 1. 72 2. 33 1. 58 1. 92 1. 08 1. 00 3. 50 1. 81 1. 08 1. 50 2. 03 1. 00 1. 06 1. 22 1. 03 1. 39
4. 08 1. 17 3. 72 1. 17 3. 22 1. 53 5. 89 2. 64 4. 75 '1. 06
4. 70 1. 73 4. 34 1. 34 4. 76 1. 89 5. 07 2. 00
11. (N:
12. (P:
13. (P:
14. (N:
15. (M:
16. (N:
17. (M: Japs in army)
18. (P: native white Americans)
19. (N: drive out)
20. (P: refugees)?
21. (M: Okies)
22. (N: agitators)
23. (P: Germans and Japs)
foremen) 6. 05 foreign ideas) 4. 86 conscientious object~rs) 4. 64 districts) 6. 33 voting) 5. 06 menial jobs)
sterilize)
lynch)
Mexico) 3. 69 Filipinos) 5. 64 granunar schools) 1. 86 motsuiters) 5. 58
24. (M:
25. (N:
26. (P:
27. (M:
28. (N:
29. (M:
30. (M: foreigners, unions)
31. (N: primitive)
32. (M: pamper criminals)
33. (P: always war)
34. (N: overbearing)
Means: Total scale Subscale N Subscale . M
Subscale P
Number: T otal H. Q. L. Q.
group
=
144
Range of scores:
Total group= 1. 2-5. 6 H. Q. =4. 2-5. 6 L. Q. =1. 2-2. 2
= 36 = 36
High Quartile
3. 72 6. 17 6. 48 5. 08 3.
12. The main threat to basic American institutions during this century has come from the infiltration of foreign ideas, doctrines, and agitators. ?
13. Present treatment of conscientious objectors, draft evaders, and enemy aliens is too lenient and mollycoddling. If a person won't fight for his . country, he deserves a lot worse than just prison or a work camp.
18. In view of the present national emergency, it is highly important to limit responsible government jobs to native, white, Christian Americans.
20. European refugees may be in need, but it would be a big mistake to lower our immigration quotas and allow them to flood the country.
23. It has become clear that the Germans and Japanese are racially war-minded and power-seeking, and the only guarantee of future peace is to wipe out most of them and to keep the rest under careful control.
26. Mexico can never advance to the standards of living and civilization of the U. S. , due mainly to the innate dirtiness, laziness, and general backwardness of Mexicans.
33? There will always be wars because, for one thing, there will always be races who ruthlessly try to grab more than their share.
toward much that he cannot personally accept for himself. He is free of rigid conformism, outgroup rejection, and imperialistic striving for power. lngroup opinions and attitudes are expressed in Items 3, 7, and 10. They are intended to express a general value for obedience and discipline, the opinion that nations are arranged hierarchically from superior to inferior, and the attitude that the superior ones should be dominant-with the assump- tion that we are one of the superior nations. The rigidity of the value for obedience is shown by the punitive attitude toward those who disobey (Item
I 3: Punishment of conscientious objectors and draft evaders).
The glorification of the national ingroup is shown further in the tendency to regard other nations as inferior when they are distant (Item 26), and threatening when they come too close (Items I 2, 20, and 2 3). W e are there- fore morally justified in excluding refugees, in "wiping out" the Germans and Japanese,2 in excluding foreigners and others from government jobs,
2 This item (23), so relevant during the war, can of course no longer be used. Ot should be pointed out that one could actively support the war without such a destructive attitude toward the enemy or such national smugness. ) I f these attitudes are correlated with rejec- tion of most other nations, then the people who made high (ethnocentric) scores on this scale may be the ones who now show similar attitudes toward our wartime allies and sup- port militaristic, imperialistic, "tough-minded policies guaranteeing American sovereignty and interests. "
? THE STUDY OF ETHNOCENTRIC IDEOLOGY 109
and in maintaining our dominant position as a nation. The cynicism about peace and the moralistic attribution of war to "ruthless, grabbing races" also indicate the sense of thre~t from outgroups and the moral righteousness of the ingroup. The fact that this theory of the cause of war is held by many college students who have been exposed to sociological explanations in terms of socioeconomic organization and conflicts raises the question: What are the inner barriers in some individuals which make them unreceptive to non- moralistic explanations? This problem, to be taken up in later chapters, concerns the personality dynamics underlying ethnocentric ideology.
If people who make high scores on this subscale are also high on the others, then it would appear that although they hold America to be superior and inviolable, they actually reject the great majority of the people in this coun- try. Item I 8 brings this out directly: it is only the native, white, Christian Americans who can be trusted. And various items from the "Minorities" subscale indicate that large sections of this population are also in the out- group category.
2. THE TOT AL ETHNOCENTRISM (E) SCALE
The total E scale is intended to measure the individual's readiness to accept or oppose ethnocentric ideology as a whole. The scale consists of 34 items3 and comprises the three subscales N, M, and P. It is presented in Table 4 (IV), with instructions to subjects, just as it was administered.
C. RESULTS: STATISTICAL ANALYSIS OF THE SCALE
The subjects were allowed the same six choices of response for each item (+3 to -3, with no neutral response), and the responses were converted into scores in the same way (-3= I point, -2 = 2 points, etc. ) as was the case with the A-S scale. All the items were regarded as pro-ethnocentric. For the 34 items, then, the total scores can range between 34 points (I point on each item, indicating strong anti-ethnocentrism) and 2 38 points (7 points on each item, strong ethnocentrism). When the total score is divided by 34 we obtain the mean score per item; thus, a total score of 5I can also be stated as
a mean per item of 1. 5.
This scale was administered as part of the questionnaire which also con-
tained the initial (52 item) A-S scale. As reported in Chapter III, this ques- tionnaire was given in April, I944? to a class in Introductory Psychology at the University of California. The data presented here are based on the ques- tionnaires of the I44 women subjects, including nineteen members of major minorities.
3 Items r, 3, 4, 7, 9, 10 of the present scale were taken, in some cases with slight modifica- tions, from the "Unlabelled Fascist Attitudes" Scale of Edwards (22). Both Edwards' study and the present one have profitted from previous studies by Gundlach (46), Katz and Cantril (17), and Stagner (rn, rq). Several of the "Negro" items have been taken from Murphy and Likert (84).
? I IO THE AUTHORITARIAN PERSONALITY
TABLE 4 (IV)
THE ToTAL ETHNocENTRISM ScALE Public Opinion Questionnaire E
The following statements refer to opinions regarding a number of social groups and issues, about which some people agree and others disagree. Please mark each statement in the left-hand margin according to your agreement or disagreement, as follows:
+ r : slight support, agreement +2: moderate support, "
+3: strong support, "
- 1 : slight opposition, disagreement -2: moderate opposition, "
-3: strong opposition, "
_ _ _ 1.
___ 2.
3? 4?
5?
6.
___7?
___ 8. ___9?
___IO.
___I1. ___12. ___IJ.
_ _ _ 14.
The many political parties tend to confuse national issues, add to the expense of elections, and raise unnecessary agitation. For this and other reasons, it would be best if all political parties except the two major ones were abolished.
If there are enough Negroes who want to attend dances at a local dance hall featuring a colored band, a good way to arrange this would be to have one all-Negro night, and then the whites could dance In peace the rest of the time.
Patriotism and loyalty are the first and most important requirements of a good citizen.
Certain religious sects whose beliefs do not permit them to salute the flag should be forced to conform to such a patriotic action, or else be abolished.
The Negroes would solve many of their social problems by not being so irresponsible, lazy, and ignorant.
Any group or social movement which contains many foreigners should be watched with suspicion and, whenever possible, be investigated by the FBI.
There will always be superior and inferior nations in the world and, in the interests of all concerned, it is best that the superior ones be in con. trol of world affairs.
Negro musicians are sometimes as good as white musicians at swing music and jazz, but it is a mistake to have mixed Negro-white bands. Although women are necessary now in the armed forces and in indus- try, they should be returned to their proper place in the home as soon as the war ends.
Minor forms of military training, obedience, and discipline, such as drill, marching and simple commands, should be made a part of the elementary school educational program.
It would be a mistake to have Negroes for foremen and leaders over whites.
The main threat to basic American institutions during this century has come from the infiltration of foreign ideas, doctrines, and agitators. Present treatment of conscientious objectors, draft-evaders, and enemy aliens is too lenient and mollycoddling. If a person won't fight for his country, he deserves a lot worse than just a prison or a work camp. Negroes may have a part to play in white civilization, but it is best to keep them in their own districts and schools and to prevent too much intermixing with whites.
One main difficulty with allowing the entire population to participate
__15.
? __I6. _ _ 17. __I8.
___19.
___20. __2I.
__22.
__23.
_ _ 24.
fully in government affairs (voting, jobs, etc. ) is that such a large percentage is innately deficient and incapable.
Manual labor and menial jobs seem to fit the Negro mentality and abil- ity better than more skilled or responsible work.
It is a mistake to allow any Japanese to leave internment camps and enter the army where they would be free to commit sabotage.
In view of the present national emergency, it is highly important to limit responsible government jobs to native, white, Christian Ameri- cans.
In a community of I,ooo whites and 50 Negroes, a drunken Negro shoots and kills an officer who is trying to arrest him. The white pop- ulation should immediately drive all the Negroes out of town.
European refugees may be in need, but it would be a big mistake to lower our immigration quotas and allow them to flood the country.
The many faults, and the general inability to get along, of the Okla- homans ("Okies"), who have recently flooded California, prove that we ought to send them back where they came from as soon as condi- tions permit.
The people who raise all the talk about putting Negroes on the same level as whites and giving them the same privileges are mostly radical agitators trying to stir up conflicts.
It has become clear that the Germans and Japanese are racially war- minded and power-seeking, and the only guarantee of fumre peace is to wipe out most of them and to keep the rest under careful control.
A large-scale system of sterilization would be one good way of breed- ing out criminals and other undesirable elements in our society and so raise its general standards and living conditions.
THE STUDY OF ETHNOCENTRIC IDEOLOGY I I I
_ _ 25. An occasional lynching in the South is a good thing because there
_ _ _ 26.
_ _ 27. _ _ _ 28.
_ _ _ 29.
_ _ 30. - - 3 ! . - - 3 2 ? - - 3 3 ? - - 3 4 ?
is a large percentage of Negroes in many communities and they need a scare once in a while to prevent them from starting riots and dismrb- ances.
Mexico can never advance to the standards of living and civilization of the U. S. , due mainly to the innate dirtiness, laziness, and general backwardness of Mexicans.
Filipinos are all right in their place, but they carry it too far when they dress lavishly, buy good cars, and go around with white girls.
It would be best to limit Negroes to grammar and trade school educa- tion since more schooling just gives them ambitions and desires
which they are unable to fulfill in white competition.
Zootsuiters demonstrate that inferior groups, when they are given too much freedom and money, just misuse their privileges and create disturbances.
The most vicious, irresponsible, and racketeering unions are, in most cases, those having largely foreigners for leaders.
There is something inherently primitive and uncivilized in the Negro, as shown in his music and his extreme aggressiveness.
W e are spending too much money for the pampering of criminals and the insane, and for the education of inherently incapable people. There will always be wars because, for one thing, there will always be races who ruthlessly try to grab more than their share.
Most Negroes would become officious, overbearing, and disagreeable if not kept in their place.
? II2 THE AUTHORITARIAN PERSONALITY
1. RELIABILITY
Data indicating the reliability and related statistical properties of the E scale and its subscales are given in Table 5(IV).
TABLE 5 (IV)
RELIABILITY OF THE E. "I'HNOCENTRIS'If (E) SCALE AND ITS SUBSCALES
Property E Scale Reliabilitya . 91
Negroes
? 91 12
2. 72 2. 65 2. 78
1. 25 1. 25 1. 42
1. 0-5. 6
Minorities
? 82 12
3. 32 3. 23 3. 40
1. 21 1. 37 1. 28
1. 0-6. 0
Patriotism
. 80 10
3. 53 3. 88 3. 18
1. 26 1. 26 1. 46
1. 0-6. 1
Number? of iterns Mean (total) b Mean (odd half) Mean (even half) S. D. (total)b S. D. (odd half) S. D. (even half)
Rangeb
34
3. 17
3. 02 3. 32
t. 15 1. 17 1. 21
1. 2-5. 6
Total Subscales
s. . rhe split-half reliability of each scale was obtained by correlating the sum of the scores on the odd items with the sum of the even items. and correcting this value by the Spearman-Brown formula,
brhe means, s. D. ? s, and ranges are given in terms of mean score per
item on the scale or subscale in question, If this value is multiplied by the number of items in the scale o~ subscale, it is converted into mean total score,
The split-half reliability of the total E scale is . 91, a value which meets accepted statistical standards. 4 The odd and even halves were roughly equiv- alent in the sense that they contained about equal numbers of items from the three subscales. The lower mean of the odd half seems due to the slight over- weighting with low-mean Negro items. The obtained range covered most of the possible range, with the exception of the extremely high end. The absence of very high scores (averages of over 6 points per item) is also reflected in the relatively low group mean of 3? I7, as compared with the neutral point of 4. 0 per item. The distribution of scores is very symmetrical-the mean divides the range in half, and the median is 3. 2-but platykurtic, with very little clustering of scores around the mean.
The high reliabilities of the subscales are noteworthy, especially in view of the small number of items in each.
In terms of reliability, equivalence of halves, and form of distribution,
4On the chance that the 19 minority group members might be atypical in some way, a separate reliability was computed for the 125 remaining subjects. The obtained value was .
91, identical with that for the total group.
? Negroes Minorities Patriotism
Negroes
. 74 . 76
Minorities . 74
. 83
Patriotism
? 76 ? 83
Total E
. 90 . 91 . 92
THE STUDY OF ETHNOCENTRIC IDEOLOGY IIJ
then, it seems safe to conclude that the E scale and its subscales provide ade- quate measuring instruments. To the extent that the scale is valid, it provides a measure of ethnocentrism, in most of its generality and complexity. It may be claimed that the higher an individual's score, the greater his acceptance of ethnocentric propaganda and the greater his disposition to engage m ethnocentric accusatiops and programs of one form or another.
2. INTERCORRELATIONS AMONG THE SUBSCALES
The reliability data support the hypothesis that there is such a thing as general ethnocentric ideology and that people can be roughly ranked ac- cording to the strength of their acceptance or rejection of it. Support for this hypothesis is also given by the high intercorrelations among the sub- scales, as shown in Table 6(1V).
TABLE 6 (IV)
CORRELATIONS OF THE E SUBSCALES WITH EACH 01'H. rn AND WITH THE TOTAL E SCALEa
aThese are the raw correlation coefficients. If they were corrected for attenuation to give the maximal value theoretically obtainable (with perfectly reliable instruments), they would all be . 9 or over.
The subscale intercorrelations, which range from ? 74 to . 83, are of con- siderable significance. The fact that they involve items dealing with so great a variety of groups and ideas suggests again that ethnocentrism is a general frame of mind, that an individual's stand with regard to one group such as Negroes tends to be similar in direction and degree to his stand with regard to most issues of group relations.
The intercorrelations of . 90 to . 92 between each subscale and the total E scale make the same point; an individual's score on any one subscale per- mits one to predict very closely his score on the entire E scale. Or, to put it in another way: While almost every subject shows some variability in his responses to the individual items (as will be shown below), almost every one demonstrates a general degree of pro- or anti-ethnocentrism which is relatively consistent from one group or type of group to another. And ethno- centric hostility toward outgroups is highly correlated with ethnocentric idealization of ingroups.
? THE AUTHORITARIAN PERSONALITY
3. INTERNAL CONSISTENCY: STATISTICAL ANALYSIS OF THE INDIVIDUAL ITEMS
The functions of item analysis, and the procedures involved, have been discussed in the previous chapter. The data on the item analysis of the E scale are presented in Table 7(IV). Each item is identified by a key word or phrase, and the letters N, M, and P refer to the subscales Negroes, Minori- ties, and Patriotism respectively. It will be recalled that the Discriminatory Power (D. P. ) equals the mean for the high quartile minus the mean for the low quartile. The total group mean is, of course, based on all four quartiles.
In general the D. P. 's in Table 7(IV) are very satisfactory,5 averaging 2. 97? For the 34 items, 5 D. P. 's are over 4. 0, 13 are between 3. 0 and 3. 9, and 10 are between 2. 0 and 2. 9; only 3 are between r. o and r. 9, and 3 less than 1. 0. Furthermore, all 6 items with D. P. 's of less than 2. 0 have group means of less than 3. 0, so the D. P. is more significant than it appears. 6
The three least discriminating items are I9, 25, and 28, all in subscale N. They are also the only three items with group means of less than 2. 0. Their low means indicate almost unanimous disagreement by all subjects. This is to be expected, since the items are particularly violent and repressive: Ne- groes should be driven out of town, lynched, kept ignorant and uneducated. But these data show the advantage of permitting three degrees of agreement and of . disagreement, and they also reveal a subtle receptiveness in the high-scoring subjects to openly antidemocratic programs. Of the 36 low scorers only one responded with -2 (on Item 28), all other responses on all three items being a firm -3 (and thus a low mean of r. oo). The high quartile, on the other hand, had a mean of I. 8 on each of the three items; nearly half of them responded with -2 or above. One might ask if this is an indication of potential response during a period in which fascism had become a real power. Not all those who score high onE, certainly, are receptive to violent antidemocracy; the task of determining the deeper psychological forces which make for potential receptiveness or opposition to fascism- the ultimate in ethnocentrism-is one which follows the first task of measur- ing ethnocentrism in its presently existing form.
The item analysis indicates that the N, M, and P subscales contributed about equally to the total differentiation between the high and low quartiles on the total scale, the average D. P. being 3. 0, 2. 9, and 3? I respectively. Apart from items I9, 25, and 28, discussed above, the Negro items were highly dis- criminating. Ethnocentrists and anti-ethnocentrists, as measured by the total scale, are clearly differentiated with respect to most of the ideas contained
5 The D. P. 's would be even higher if the "range of talent" in this group included more extreme ethnocentrists. This is shown by results on subsequently tested groups.
6 While correlations between items or between each item and the total scale have not been computed for this group, later data on similar scales suggest that the average correla- tion between single items is about . 4, while between each item and the sum of the remain- ing items the average correlation is about . 6.
? THE STUDY OF ETHNOCENTRIC IDEOLOGY TABLE 7 (IV)
MEANS AND DISCRIMINATORY PO\VmS OF THE E-SCALE ITEMS
No. Item
1. (M: political parties)
2. (N: dance)
3. (P: patriotism)
4. (M: religious sects)
5. (N: lazy)
6. (M: foreign groups)
7. (P: superior nations)
8. (N: bands)
9. (M: women)
10. (P: military training)
Low Quartile
2. 17 1. 97 3. 86 1. 61 1. 53 1. 69 1. 25 1. 25 3. 75 2. 47 1. 69 1. 22 1. 44 1. 72 2. 33 1. 58 1. 92 1. 08 1. 00 3. 50 1. 81 1. 08 1. 50 2. 03 1. 00 1. 06 1. 22 1. 03 1. 39
4. 08 1. 17 3. 72 1. 17 3. 22 1. 53 5. 89 2. 64 4. 75 '1. 06
4. 70 1. 73 4. 34 1. 34 4. 76 1. 89 5. 07 2. 00
11. (N:
12. (P:
13. (P:
14. (N:
15. (M:
16. (N:
17. (M: Japs in army)
18. (P: native white Americans)
19. (N: drive out)
20. (P: refugees)?
21. (M: Okies)
22. (N: agitators)
23. (P: Germans and Japs)
foremen) 6. 05 foreign ideas) 4. 86 conscientious object~rs) 4. 64 districts) 6. 33 voting) 5. 06 menial jobs)
sterilize)
lynch)
Mexico) 3. 69 Filipinos) 5. 64 granunar schools) 1. 86 motsuiters) 5. 58
24. (M:
25. (N:
26. (P:
27. (M:
28. (N:
29. (M:
30. (M: foreigners, unions)
31. (N: primitive)
32. (M: pamper criminals)
33. (P: always war)
34. (N: overbearing)
Means: Total scale Subscale N Subscale . M
Subscale P
Number: T otal H. Q. L. Q.
group
=
144
Range of scores:
Total group= 1. 2-5. 6 H. Q. =4. 2-5. 6 L. Q. =1. 2-2. 2
= 36 = 36
High Quartile
3. 72 6. 17 6. 48 5. 08 3.
