No More Learning

92 2. 16 4. 34 3. 05 4. 09 3. 48 2. 50 3. 08 3. 15 2. 34 3. 97 2. 66 2. 60 2. 86 2. 24 2. 69
3. 29 2. 90
Rank D. P. (14)
TABLE 10 (IV)
MEANS AND DISCRIMINATORY POWERS OF THE E-SCALE ITEMS (FORM 78) a
. . . . . N 0
aThe four groups on which these data are based are: Group Group 8, U,C, Public Speaking Class Men (N =52). Group Group D, Professional Women (N = 63).
Public Speaking Class Women (N =140). Extension Psychology Class Women (N =40).
3. 99 3. 73 3. 36 4. 89 3. 69 2. 66 2. 06 4. 11 4. 26 2. 49 3. 59 4. 44 2. 51 2. 52
3. 45
0. 17 3. 94 3. 01 2. 76 4. 54 1. 89 2. 85 2. 38 3. 76 2. 27 2. 70 2. 44 2. 52 2. 96
2. 73
3. 44 3. 29 3. 69 3. 90 3. 12 4. 42 1. 75 3. 69 3. 60 2. 56 3. 37 4. 33 2. 75 2. 46
3. 31
2. 29 2. 93 2. 15 3. 35 3. 29 2. 36 1. 86 3. 14 3. 58 2. 00 1. 50 1. 86 2. 21 2. 93
2. 53
4. 30 3. 58 3. 32 5. 60 4. 52 2. 00 2. 35 5. 22 5. 28 2. 90 3. 22 3. 95 2. 98 2. 25
3. 68
A, u. c. C, u. c.
1. 45 4. 07 3. 95 2. 29 4. 04 1. 91 2. 34 2. 69 3. 54 4. 72 1. 95 3. 11 3. 65 1. 87
2. 97
4. 06 2. 90 2. 30 4. 02 2. 30 1. 89 1. 50 4. 33 3. 21 2. 03 2. 43 3. 17 2. 14 1. 75
2. 72
2. 12 5. 12 3. 92 4. 74 4. 12 2. 57 1. 60 3. 98 3. 06 3. 33 3. 21 3. 22
3. 05 3. 00
3. 36
. . . ,
bin obtaining the over-all means, the individual group means were not weighted by M.
(1) ~ 1:"1
(5)
(4)
> Q
(2) s
(12)
(13)
(7)
(3)
(6)
';<j
~
> z
"C
1:"1
';<j
( 11) Cll
0
(10)
(8) t"
(9)
::J
. . . ::
~
? THE STUDY OF ETHNOCENTRIC IDEOLOGY I 2 I
Item 34, stating that worrien should be restricted to "feminine positions" such as nursing, ranked I2 out of 14. It is interesting that the women's groups (A, C, Din Table w(IV)) tended predominantly to disagree with this item, obtaining means of only 1. 9 to 2. 7, while the group of college men showed a slight tendency to agree, having a mean of 4-4- Despite the similar Dis- criminatory Powers for men and women, the D. P. for women is probably more significant statistically, since their mean is so much lower. It would appear that the ethnocentric women are more bound, at least on the surface, to the traditional imagery of femininity, while the nonethnocentrists wish to emancipate women, occupationally and otherwise, from their traditionally imposed limitations. While the relationship is far from perfect, it suggests that different patterns of ideology regarding masculinity and femininity may exist in the two groups. This general problem is investigated more fully in later chapters.
The attempted improvements in Items 37 and 54 were also relatively un- successful. Item 37, which makes the Negroes entirely responsible for their own poverty, was apparently too strongly worded to receive much agree- ment (mean= 1. 92). The low mean indicates that the D. P. of 2. 16 is very significant; there is very little overlapping between low and high scorers, the former tending almost uniformly to disagree strongly (-3), while the latter disagree on the average only slightly ( - 1). Similarly, the relatively low D. P. of 2. 7 and mean of 2. 2 on Item 64 (No Japs in California) might have been higher had the item been given a pseudodemocratic coloring, thus allowing the ethnocentrists more moral justification for agreeing with it. Item 54, rejecting the bulk of the people as "innately deficient and in- capable," has a more ambiguous relation to ethnocentrism. The subjects were evenly divided on this issue, and the D. P. of 2. 7, while indicating a sig- nificant difference between the high and low quartiles, nevertheless permits considerable agreement by low scorers, disagreement by high scorers.
Of the four entirely new items, two were among the least discriminating. Item 51, which suggests that our social problems could be solved by eliminat- ing "bad" people (rather than by changing the underlying social forces and institutions), had a D. P . of 2. 3, rank 11, indicating a clear-cut difference be- tween the high and the low quartiles, but numerous exceptions as well.
The poorest item, with a D. P. of 1. 5, was number 4 (urging that dis- crimination be made illegal). The subjects were apparently evenly divided on this issue, and relatively few were willing to take an extreme stand either way. The fact that the Professional Women had a mean of 4. 1 on this item, as compared with their scale mean of 2. 7, was perhaps a straw in the wind to indicate that the attempted California Fair Employment Practices Law (referendum) of 1946 would receive far less than majority support. In their interview discussions many strongly anti-ethnocentric subjects-who clearly recognized the crucial role of discrimination in maintaining ingroup-out-
?
122 THE AUTHORITARIAN PERSO~ALITY
group conflicts-were nevertheless political pacifists in the sense of being un- willing militantly to oppose discrimination. Some of the psychological
sources of this point of view will be considered in later chapters.
The E scale (Form 78), while adequate for its intended uses, was revised in the light of the results just discussed. The revision also took account of
the correlations now to be considered, between the E and A-S scales.
D. THE INCLUSION OF ANTI-SEMITISM WITHIN GENERAL ETHNOCENTRISM
It will be recalled that the E scale contained no items referring to Jews; rather, the independent A-S scale was included within the total question- naire. We may now consider the correlations between these scales.
The initial form of the questionnaire, administered in 1944 to the Uni- versity of California Psychology Class Women, contained the sz-item A-S scale and the 34-item E scale. Correlations of the A-S scale with the E scale and its subscales are presented in Table I I (IV).
TABLE 11 (IV)
CX>RRELATIONS BETWEFN THE A-S AND E SCALES (INITIAL FORMS)a
E Subscale
Total E Seale "Negroes" "Minorities? "Patriotism"
A-S . 80 . 74 . 76 . 69
~he reliabilities of these scales, as presented previously, are as follows: A? S = . 92; E = . 91; "Negroes"= . 91; "Minorities"= . 82; "Patriotism"= . 80.
The correlation of . 8o between E and A-S permits a further broadening in the conception of ethnocentrism. The correlations of . 69-. 76 between A-S and the E subscales are only slightly lower than the correlations of . 74-. 83 among theE subscales (see Section C, above). These values indicate once again the generality of the ethnocentric approach to group relations. Anti-Semitism is best regarded, it would seem, as one aspect of this broader frame of mind; and it is the total ethnocentric ideology, rather than prejudice against any single group, which requires explanation. The fact that A-S correlates slightly less with the E subscales than the latter correlate among themselves may be due in part to the shortened range of A-S scores (absence of extreme highs); however, it appears likely that there are certain specific determinants of anti-Semitism apart from those which hold for general ethno- centrism.
The correlations between the A-S and E scales in Form 78, presented in
? THE STUDY OF ETHNOCENTRIC IDEOLOGY 123
Table I2 (IV), provide a further indication of the generality of ethnocen- trism. The average correlation, . 68, is lower than that of . So obtained in the initial form. This difference is due partially to the decreased reliability of the shortened E scale. However, it suggests once again that anti-Semitism,
TABLE 12 (IV)
roR. RELATIONS BETWEEN THE A-S AND E SCALES (FORM 78)
Group
Public Speaking women Public Speaking men Extension women Professional women
Mean ra
&. . rhe mean r is unweighted for N and not based on Zr?
while it is essentially a facet of a broader ethnocentric pattern, may neverthe- less have certain independent determinants of its own.
1. THE THIRD FORM OF THE E SCALE (FORM 6o)
The close, though imperfect, relationship between the A-S and E scales in Form 78 led, in the construction of Questionnaire Form 6o, to their con- traction into an inclusive I 2-item E scale. This scale, presented in Table I 3
(IV), contains 4 A-S items (numbers 8, 21, 38, 47), 3 Negro items (14, 28, 42), and 5 items dealing with other minorities and with patriotism. Once again slight revisions in the wording have frequently been made, particularly in the direction of simplicity and brevity. Again items were selected on the basis both of statistical adequacy and theoretical significance. Two items (42, 55) were included, despite their previous low means and D. P. 's, because they seemed to express significant ideas and because the means, and thus the D. P. 's, were expected to go up in some of the groups to be tested. Item 59 ("Okies") was recovered from the initial E scale.
a. THE GRouPs STUDIED. Questionnaire Form 6o was administered in the summer of 1945 to the following groups:
r. Summer session students in psychology classes at the University of Oregon (one of us taking advantage of his summer teaching position there). 2. Summer session students in an education class at the University of Cali- fornia. 8 Whereas the Oregon students were almost entirely of college age, the members of this particular education class at the University of California were somewhat older (age range about 25-45), the most common occupation
8 vVe wish to express our thanks to Dr. Harold Carter for the administration of the questionnaires to his class.
N Correlation
140 . 71 52 . 75 40 ? 63 63 . 63
295 . 68
? 124
3?
8.
14.
21. 28.
33?
38. 42.
47?
THE AUTHORITARIAN PERSONALITY TABLE 13 (IV)
THE THIRD FoRM OF THE E ScALE (FoRM 6o)
Zootsuiters prove that when people of their type have too much money and freedom, they just take advantage and cause trouble.
One trouble with Jewish businessmen is that they stick together and prevent other people from having a fair chance in competition.
Negroes have their rights, but it is best to keep them in their own districts and schools and to prevent too much contact with whites.
I can hardly imagine myself marrying a Jew.
It would be a mistake ever to have Negroes for foremen and leaders over whites.
If and when a new world organization is set up, America must be sure that she loses none of her independence and complete power in matters that affect this country.
There may be a few exceptions, but in general, Jews are pretty much alike. If Negroes live poorly, it's mainly because they are naturally lazy, ignorant, and without self-control.
The trouble with letting Jews into a nice neighborhood is that they gradually
give it a typical Jewish atmosphere.
51? The worst danger to real Americanism during the last 50 years has come
from foreign ideas and agitators.
55? Citizen or not, no Jap should be allowed to return to California.
59? For the good of all, the Oklahomans ("Okies") who recently flooded Cali-
fornia ought to be sent back home as soon as possible.
being that of teacher, although other, particularly semiprofessional, occupa- tions were well represented.
3? Men's service clubs (Kiwanis, Lions, Rotary) in the vicinity of the Uni- versity of Oregon.
In order that separate data for men and women might be obtained, the above groups were divided and recombined for statistical purposes. The first University of Oregon class obtained included enough women to form a statistical group (Group I, N = 47). However, the second class at Oregon was too small to be divisible into statistically adequate subgroups of men and women, and so was the class at California. Accordingly, statistical Group II contains the combined Oregon and California Student W omen (N = 54), Group III the combined men (N = 57). Finally, Group IV contains the Oregon Service Club Men (N = 68).
b. RELIABILITY. The reliability data for the E Scale (Form 6o) are pre- sented in Table 14(IV). The reliabilities, ranging from . 82 to . 88 and aver- aging . 86 for the four groups, are entirely adequate in terms of currently accepted standards. The obtained scores cover most of the possible range ( r. o-7. o) with the exception of the extremely high end; there are few scores of over 6. o. A slight predominance of low scores is also indicated by the group means, which are well below the neutral point of 4. 0.
The differences among the various groups are of some interest. The highest degree of ethnocentrism was expressed by the Oregon Service Club Men.
? Mean (total) Mean (odd half) Mean (even hal f)
3. 43 "3. 48 3. 38
1. 38 1. 63 1. 30
3. 25 3. 24 3. 26
1. 29 1. 77 1. 53
2. 96 2. 95 2. 97
1. 26 1. 38 1. 23
3. 55 3. 72 3. 43
1. 11 1. 21 1. 17
3. 30 3. 35 3. -26
1. 26 1. 50 1. 31
S. D. S. D. S. D.
N
Range
(total) (odd half) (even half)
THE STUDY OF ETHNOCENTRIC IDEOLOGY TABLE 14 (IV)
RELIABILITY OF 'lHE E SCALE (FORM 60) a
125
Property Group Over-all I II III IV
Reliability . 88 . 88 . 86 . 82 . 86
47 54 1. 0-6. 3 1. 1-5. 9
57 68 226 1. 0-6. 3 1. 3-5. 8 1. 0-6. 3
arhe four groups on which these data are based are:
Group I: University of Oregon Student Women.
Group II: University of Oregon and University of California Student
Women.
Group III: University of Oregon and University of c? alifornia Student
Men.
Group IV: Oregon Service Club Men.
Their mean of 3? 55 is significantly higher (1 per cent level) than the lowest mean, 2. 96, obtained by the University Student Men. We may note that the group of Service Club Men was also the most constricted in its range of scores (1. 3-5-8) and in its internal variability (S. D. = I. ll); that is, its members tended to cluster around the middle position so that there are few extreme high or low scorers. These considerations help to explain why the E scale has the lowest reliability in this group and why the average D. P. is, as will be shown below, also lower for this group than for the others. That this group should exhibit a clustering around a modal "point of con- formity" is perhaps not surprising, since conformity is one of its central values. It may, however, be surprising to some that the mode should be in a middle rather than a more extreme position.
It is also of some interest that the California subjects are slightly less ethnocentric than the Oregonians. Thus, Group I, composed entirely of Oregon students, has a slightly higher mean than Group II (3-43 to 3-25), which is more than half Californian in make-up. The likelihood of a regional difference is given greater weight by the fact that at least two items (55, "Japs," and 59, "Okies," and perhaps also 3, "Zootsuiters") refer specifically to conditions in California. A slight, though also not statistically significant difference is found between comparable sex groups, the University Student Women (Group II) having a higber mean than the University Student Men (Group III) (3. 25-2. 96). No consistent, significant difference between com-
? No.
3.
8. 14. 21. 28. 33. 38. 42. 47.