(odd) (even) include only the three groups taking the total form since the (A) and (B) halves of the split form do not
correspond
to the odd and even halves of the total form.
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
.
.
5.
30 2.
00 (10) >-l
3. 32 2. 76 (4. 5) 4. 74 1. 99 (11)
><
>-l
? POLITICO-ECONOMIC IDEOLOGY AND GROUP MEMBERSHIPS rGr
average D. P. (and thus the low reliability) is not due to the counterbalanc- ing of several very good items by several very poor ones; the best item has a D. P. of 3. o-not extremely high by previous standards-and the values di- minish very gradually. It is noteworthy that the best items deal with a variety of trends: acceptance of depressions as natural (Item 5); values for the "middle of the road" and slow social change (Item r5); and for con- formity to existing authority (Item 27); and "liberal" items supporting eco- nomic security, increased government functioning, and unions (Items 36, 6I, 68, 76).
Of the seven items with the lowest D. P. 's, six had means greater than 5? 3 or less than 2. 4; that is, these items tended to evoke almost uniform responses of agreement or of disagreement. (None of the 9 best items had such extreme means. ) In view of the greatly reduced variability of response to these items, the D. P. 's are more significant than they at first appear. Only the lowest D. P. of 0. 32 (for Item 22) can be considered clearly insignificant. For the other low D. P. items the difference between the high and low quartiles is statistically significant. For a given item the difference is not that one quartile consistently agrees while the other disagrees; it lies rather in the fact that one quartile consistently obtains an extreme score while the other tends to be more neutral in its stand. Thus, with regard to the importance of teaching a child the value of a dollar (Item I), the extreme conservatives most fre- quently responded with a +3, while the extreme liberals tended to respond +I, a difference in emphasis rather than an actual opposition. There were four such items (I, 8, 22, 71), all dealing with conventional values, on which very few subjects made extremely low scores, and two (13, 44) on which there were few high scores. Some of these items were reformulated in the succeeding form of the scale, with the intention of eliminating possible am- biguities and thus increasing the D. P. 's. The possibility remains, of course, that the relative uniformity of response to these items reflects an actual uni- formity of belief on the part of these groups of subjects.
Since most of the 78 items in this series are agreed with by the high scorers on the various scales (A-S, E, PEC, F), disagreed with by the low scorers, the question of a mechanical "set" to agree or to disagree may be raised. For instance, once an individual gets set consistently to disagree, is he not likely to continue disagreeing regardless of the content of the items? The "set" factor was considered, and found to play a negligible role, in the previous scales. By way of further evidence, we may consider the five "liberal" items, that is, those which tend to be agreed with by individuals who usually dis- agree with the other items. The 5 rank D. P. 's range from r to q, and aver- age 7. I, or slightly better than the scale average of 8. 5. Furthermore, the extreme liberals tended, as noted above, to agree even with some of the "conservative" items. The great variability of the item means is also a sign of selective response to each item. It seems safe to conclude that set is not a
? THE AUTHORITARIAN PERSONALITY
major determinant of response, although it may enter as a minor complicat- ing factor.
The item means in Table 3(V) are also worthy of note. The highest means are on items expressing conventional values (I, 8, 22, 7I). The very high mean (6. 38), as well as the negligible D. P. , on Item 2 2 is probably due in part to inadequate formulation; but also to the actual tendency of most Americans to regard the good politician as a fighter against vice rather than as one who understands the political and economic problems of democratic government. In contrast to this, the two "conventional values" items which discriminated very well and whose means were near the neutral point of 4. 0 are particularly important. These items, IS (Middle of the road) and 27 (Re- bellious ideas), seem to reflect a primary personality trend underlying ideo- logical conservatism, namely the surface acceptance of authority and the overcoming of rebellious tendencies. It seems possible that the rebellious tend- encies have not actually been outgrown but have rather been inhibited, so that the emphasis on conformity now serves as a defense against underlying hostility toward accepted authority. This hypothesis, which arose previously in the case of the ethnocentrists' uncritical submission to ingroup authority, will be considered in detail in the chapters which follow.
Among the more directly ideological items, the highest mean, 4? 58, was made on Item 68 (Unions stronger). This result may indicate a fear of union strength, and perhaps a sense of alienation from the working class, among numerous middle-class individuals who are "liberal" with respect to the other political trends expressed in the scale. For example, the means on the several items (36, 44? sz, 6I, 76) dealing with social security and extension of government functions are all well below 4. 0, indicating considerable sup- port for the liberal viewpoint.
The low means on the "government" items raise another question: Why do many individuals who are otherwise conservative support an increase in government activity? In some cases this inconsistency probably reflects ideological confusion or the beginnings of change from right to left or vice versa. However, this apparent contradiction may reflect something much more basic, namely a shift from traditional laissez-faire conservatism, whose economic unit was the individual competitive businessman, to a new type of conservatism whose economic unit is organized big business. As was pointed out earlier in this chapter, the assumption of liberalism-conservatism as a simple quantitative dimension holds only in the most general sense. It was for this reason, among others, that the theory guiding scale construction was presented in some detail. It is possible, then, for an individual to make a moderately high rather than a very high score, not because of any true liberal tendency, but because of a change in the nature of his conservatism. He is now willing to extend the functions of government for reasons that are the opposite of liberal, for he conceives of government as a tool of busi-
? POLITICO-ECONOMIC IDEOLOGY AND GROUP MEMBERSHIPS 163
ness rather than as a means of controlling corporate capital and of pre- venting concentrations of economic power. We shall return again to this question after considering the relation between the PEC and E scales.
3. THE SECOND PEC SCALE (FORM 60)
In Form 6o the PEC Scale (see Table 4(V)) was shortened to 14 items, and numerous changes were made in content and wording. Items 2 7 and 6o,
TABLE 4 (V)
THE SEcoND FoRM oF THE PoLITico-EcoNoMic CoNSERVATISM (PEC) SCALE (FoRM 6o)
4? " Labor unions should become stronger and have more influence generally.
9. " Most government controls over business should continue after the war.
I 3? America may not be perfect, but the American Way has brought us about
as close as human beings can get to a perfect society.
I 5. " If America had more men like Henry Wallace in office, we would get along
much better.
zo. " The artist and the professor are of just as much value to society as the
businessman and the manufacturer.
26. It would be dangerous for the U. S. to cooperate too closely with Russia. 27. The best political candidate to vote for is the one whose greatest interest is
in fighting vice and graft.
p. " No one should be allowed to earn more than $25,ooo a year.
37? " It is up to the government to make sure that everyone has a secure job and
a good standard of living.
43. " The government should own and operate all public utilities (transportation,
gas and electric, railroads, etc. ).
48. " Depressions can be prevented by proper government planning.
54. " Poverty could be almost entirely done away with if we made certain basic
changes in our social and economic system.
56. Men like Henry Ford or J. P. Morgan, who overcome all competition on
the road to success, are models for all young people to admire and imitate. 6o. In general, the best way of aiding our fellow men is to give time or money
to some worthy charity.
a These nine items are "liberal," the other five are "conservative. " A high score is given for agreement with the conservative items, disagreement with the liberal items.
referring to political candidates and the importance of charity, respectively, are reformulations of items that were unsuccessful in Form 78; the present formulations are, presumably, more clear-cut expressions of the trends initially hypothesized. Two items which worked relatively well in Form 78, numbers 27 (Rebellious ideas) and 61 (Security is bad), were placed in the Form 6o F scale (see Chapter VII) because they seemed on theoretical grounds to fit better there.
Several totally new items have been added. Item 13 (The American Way) was taken from the Form 78 E scale (see Chapter IV); it is transitional be- tween conservatism and ethnocentrism in that it expresses both conservative
? THE AUTHORIT ARIAN PERSONALITY
support of the politico-economic status quo and ethnocentric idealization of the ingroup. Taken literally, however, it seemed to fall more within the sphere of political ideology. Correlational analysis is required before this item can properly be placed within one scale or the other.
Three of the new items provide vivid reminders of the speed of historical change. Item 15 (Wallace) was formulated when the confirmation of Henry Wallace as Secretary of Commerce was the issue of the day. Item 26 (Russia) reflected the atmosphere of the initial postwar period, when cooperation rather than containment was the prevailing attitude toward Russia. Item 3I referred to President Roosevelt's wartime suggestion of a $25,ooo limit on yearly incomes. It will be noted that the Form 6o scale contains fewer gen- eralizations and more concrete references to specific issues than did Form 78.
Form 6o was administered in the summer of 1945 to several groups which were combined for statistical purposes as follows. (a) University of Oregon Student Women (N = 47), undergraduate students attending summer session courses in psychology. (b) Combined University of Oregon and University of California Student Women (N = 54), the Oregon group being obtained too late to be included in (a); the California group was a summer session education class containing mostly teachers and others of above college age. (c) University of Oregon and University of California Student Men
(N = 57), from the same classes as the (b) women. (d) Oregon Service Club Men (N = 68), obtained at luncheon meetings of service clubs (Ki- wanis, Lions, Rotary). 4
The last three groupings received the total Form 6o questionnaire in one sitting. However, the first group of Oregon Student Women received the questionnaire in two parts, A and B. Part A included the F scale and half of PEC, while Part B contained the E scale and the remaining half of PEC. The purpose of this division was to help determine whether the pres- ence of the E items had any effect on the responses to the F items; the results will be discussed in Chapter VII.
The reliability data for the Form 6o PEC Scale are presented in Table 5(V). The average reliability of . 70 is substantially the same as that of ? 73 for the initial form, and indicates that the changes in wording and content did not improve this property of the scale. The fact that the reliabilities, S. D. 's, and ranges vary so little among these four groups, as well as among those taking the first form (Form 78), suggests that a scale of this length and this degree of inclusiveness can hardly be expected to obtain an average re- liability greater than . 7-. 8. The main reason for the relatively low reliability of PEC, as compared with E, appears to lie in the fact that political ideology is intrinsically less organized and less consistent in the individual today than is ideology concerning group relations.
Once again the group means on PEC are significantly higher than those on 4 For a discussion of the sampling problems involved, see Chapter IV.
? Property
Re 1 i a b i l i t y
Mean (total) Mean (odd half) Mean (even half)
S. D. (total)
RELIABILITY OF THE PEC SCALE (FORM 60)
Grou1t
I II III IV
over-all
. 7 0
3. 92d 3. 74d 4. 23
1. 0-6. 1
S. D. S. D.
Range
(odd (even
half) half)
. 73
3. 72 3. 86(A)c 3. 58(B)c
. 90
. 97(A}c 1. 02(B}c
1. 2-5. 6
. 69
3. 82 3. 60 4. 03
1. 0-5. 5
. 69
3. 77 3. 55 3. 99
. 80
. 78 1. 05
1. 2-5. 0
. 70
4. 40 4. 06 4. 68
. 92
. 95 1. 14
1. 6-6. 1
POLITICO-ECONOMIC IDEOLOGY AND GROUP MEMBERSHIPS I6 ) TABlE 5 (V)
aThe four groups are:
Group 1: Group II:
Group Ill:
Group IV:
University of Oregon Student Women (N = 47)
University of Oregon and University of California Student Women (N = 54)
University of Oregon and University of California Student Men (N = 57)
Oregon Service Club Men (N? = 68)
brn obtaining the over-all means, the individual group means were not weighted by N.
cThe signs (A) and (B) refer to the two parts of the questionnaire given to Group I; half of PEC was in part (A), the other half in part (B).
The reliability for this group is based on the correlation between the A and B halves.
dThe Over-all mean (odd) (even), and S. D.
(odd) (even) include only the three groups taking the total form since the (A) and (B) halves of the split form do not correspond to the odd and even halves of the total form.
E (Chapter IV), suggesting that the level of conservatism is higher than the level of ethnocentrism. Again, the rank order of group means on PEC tends to follow that on E, with the Service Club Men being significantly more con- servative (beyond the 1 per cent level) than the combined university groups. These facts would lead us to expect a significant correlation between PEC and E (see Section C). While the Service Club Men are quite conservative on the average (mean of 4. 4), the lowest score being 1. 6, this group can by no means be considered ideologically homogeneous; indeed, it shows about the same degree of internal variability (range and S. D. ) as do the other groups. We are given another warning against stereotypy in thinking about groups and about group memberships as determinants of ideology. This is not to
say that service clubs are not "conservative groups" in terms of actual policy. Rather, it would appear that group policy and leadership, in this case
? 166 THE AUTHORIT ARIAN PERSONALITY
at least, reflect the average degree of conservatism, the conservative tradition, and frequently the immediate business interests of the group. However, to say that such-and-such is a conservative group, in terms of actual policy, is not necessarily to imply that all members are strongly conservative. Simi- larly, not all individuals who call themselves "New Deal Democrats" are thoroughly liberal in their personal ideologies; not all Catholics support the political program of the policy-makers of the Catholic Church; and so on. This is one of the main problems in bringing together the psychological and the sociological approaches; it is an especially great problem for that theory of social psychology which regards the individual adult as merely
a product or sum of his various group memberships.
The data on item analysis are presented in Table 6(V). The over-all D.
P. of 2. 08 is almost identical with that of 2. I4 on the initial PEC scale, as might be expected from the similar reliabilities. The best items deal for the most part with government functioning; ownership of utilities, controls over business, limitations on income. Item 4 (Unions) worked out relatively well (rank order 6) despite its having the highest over-all mean, 5? 35; even in the university groups the mean did not fall below 5. o. Item I5 (Wallace) came out similarly; it had the third best D. P. despite a mean of 5. 00. Other items with D. P. 's of over 2. 0 include 13 (American Way), 54 (Poverty), and 56 (Ford and Morgan).
The five poorest items are also of some interest. Three of these, 20 (Artists, businessmen), 27 (Political candidate), and 6o (Charity), are reformulations of poor items in Form 78. Almost none of these subjects disagreed with the idea that the artist and professor are as important as the businessmen, al- though the liberals agreed more emphatically than the conservatives (the difference being statistically significant only in the Service Club Men). The D. P. of 1. 06 on Item 27, while statistically significant, indicates considerable overlap between the high and low quartiles. Further exploratory research is required in order to determine possible differences between liberals and conservatives with respect to underlying imagery of "the good political candidate. " Item 48 (Depressions) is an example of not leaving well enough alone. In the initial form this item had a D. P. rank of 4. 5; in this form, after drastic revision, its rank was I2. Both the mean and D. P. on Item 26 (Dan- gerous to cooperate with Russia) are somewhat surprising. The mean of 2. 57 indicates that very few individuals agreed with this item. The D. P. of 1. 60 is more significant than it at first appears, because of the low mean, but it shows that even conservatives were divided on the Russian issue at the close of the war. This is shown most dramatically by the Service Club Men who,
although strongly conservative on most domestic issues, obtained on the Russian item a mean of 2. 5 I and a D. P. of ? 93? How and why slight support has, within less than two years, changed to bitter antagonism, is a problem beyond the scope of the present study.
? No.
I t e m
(Unions) 5. 43
Mean
D. P .
Mean
D. P . Mean
D. P . . . . . . .
4.
2. 65 4. 43 3. 54 4. 15 1. 97
5. 06 2. 64 4. 81 2. 26 4. 91 2. 22 4. 72 2. 64 1. 22 0. 38 2. 54 1. 47 3. 30 0. 34 4. 74 1. 98 4. 72 3. 72 4. 93 2. 77 2. 76 0. 85 3. 63 2. 02 3. 22 2. 12 2. 87 1. 73
1. 60 2. 94 3. 08 2. 33 0. 14 1. 52 1. 19 3. 42
5. 65 5. 87 4. 68 5. 84 1. 87 2. 51 3. 21 5. 56
2. 76 5? . 35 2. 82 4. 88 2. 16 4. 38 2. 51 5. 00 1. 15 1. 44 0. 93 2. 57 1. 99 3. 49 2. 12 5. 00 2. 99 4. 74
9. 13. 15. 20. 26. 27? .
(Gov't. controls)
(American Way)
(Wallace) 4. 79 (Artists and professors)
(Russia)
(Political candidate)
($25,000/yr. )
(Gov? t. and jobs)
2. 89 (2) 2. 36 (7) 2. 62 (3)
0 z 0
31. 37. 43. 48. 54. 56. 60.
2. 44 (5)
C)
(Utilities) (Depress ions) (Poverty) (Ford, Morgan) (Charity)
4. 61 3. 48 3. 09 1. 30 4. 21 1. 86 3. 00 1. 75 2. 47 0. 91
5. 82 3. 23 4. 41 1. 94 4. 31 2. 38 3. 34 3. 13 2. 88 2. 22
2. 61 (4) 4. 96 3. 36 (1)
~
Mean per item
3. 72 2. 16
3. 82 1. 94
3. 77 1. 91
4. 40 2. 31
~
aGroup I: Group II: Group III: Group IV:
University of Oregon Student Women (N =47)
University of Oregon and University of California Student Women (N =~) University of Oregon and University of California Student Men (N =57) Oregon Service Club Men (N =68)
l""::
TABLE 6 (V)
MEAI\5 AND DISCRIMINATCRY PORERS <F THE PEC SCAlE ITEM3 (FCRM 60~a
'"d
Group I Mean D .
3. 32 2. 76 (4. 5) 4. 74 1. 99 (11)
><
>-l
? POLITICO-ECONOMIC IDEOLOGY AND GROUP MEMBERSHIPS rGr
average D. P. (and thus the low reliability) is not due to the counterbalanc- ing of several very good items by several very poor ones; the best item has a D. P. of 3. o-not extremely high by previous standards-and the values di- minish very gradually. It is noteworthy that the best items deal with a variety of trends: acceptance of depressions as natural (Item 5); values for the "middle of the road" and slow social change (Item r5); and for con- formity to existing authority (Item 27); and "liberal" items supporting eco- nomic security, increased government functioning, and unions (Items 36, 6I, 68, 76).
Of the seven items with the lowest D. P. 's, six had means greater than 5? 3 or less than 2. 4; that is, these items tended to evoke almost uniform responses of agreement or of disagreement. (None of the 9 best items had such extreme means. ) In view of the greatly reduced variability of response to these items, the D. P. 's are more significant than they at first appear. Only the lowest D. P. of 0. 32 (for Item 22) can be considered clearly insignificant. For the other low D. P. items the difference between the high and low quartiles is statistically significant. For a given item the difference is not that one quartile consistently agrees while the other disagrees; it lies rather in the fact that one quartile consistently obtains an extreme score while the other tends to be more neutral in its stand. Thus, with regard to the importance of teaching a child the value of a dollar (Item I), the extreme conservatives most fre- quently responded with a +3, while the extreme liberals tended to respond +I, a difference in emphasis rather than an actual opposition. There were four such items (I, 8, 22, 71), all dealing with conventional values, on which very few subjects made extremely low scores, and two (13, 44) on which there were few high scores. Some of these items were reformulated in the succeeding form of the scale, with the intention of eliminating possible am- biguities and thus increasing the D. P. 's. The possibility remains, of course, that the relative uniformity of response to these items reflects an actual uni- formity of belief on the part of these groups of subjects.
Since most of the 78 items in this series are agreed with by the high scorers on the various scales (A-S, E, PEC, F), disagreed with by the low scorers, the question of a mechanical "set" to agree or to disagree may be raised. For instance, once an individual gets set consistently to disagree, is he not likely to continue disagreeing regardless of the content of the items? The "set" factor was considered, and found to play a negligible role, in the previous scales. By way of further evidence, we may consider the five "liberal" items, that is, those which tend to be agreed with by individuals who usually dis- agree with the other items. The 5 rank D. P. 's range from r to q, and aver- age 7. I, or slightly better than the scale average of 8. 5. Furthermore, the extreme liberals tended, as noted above, to agree even with some of the "conservative" items. The great variability of the item means is also a sign of selective response to each item. It seems safe to conclude that set is not a
? THE AUTHORITARIAN PERSONALITY
major determinant of response, although it may enter as a minor complicat- ing factor.
The item means in Table 3(V) are also worthy of note. The highest means are on items expressing conventional values (I, 8, 22, 7I). The very high mean (6. 38), as well as the negligible D. P. , on Item 2 2 is probably due in part to inadequate formulation; but also to the actual tendency of most Americans to regard the good politician as a fighter against vice rather than as one who understands the political and economic problems of democratic government. In contrast to this, the two "conventional values" items which discriminated very well and whose means were near the neutral point of 4. 0 are particularly important. These items, IS (Middle of the road) and 27 (Re- bellious ideas), seem to reflect a primary personality trend underlying ideo- logical conservatism, namely the surface acceptance of authority and the overcoming of rebellious tendencies. It seems possible that the rebellious tend- encies have not actually been outgrown but have rather been inhibited, so that the emphasis on conformity now serves as a defense against underlying hostility toward accepted authority. This hypothesis, which arose previously in the case of the ethnocentrists' uncritical submission to ingroup authority, will be considered in detail in the chapters which follow.
Among the more directly ideological items, the highest mean, 4? 58, was made on Item 68 (Unions stronger). This result may indicate a fear of union strength, and perhaps a sense of alienation from the working class, among numerous middle-class individuals who are "liberal" with respect to the other political trends expressed in the scale. For example, the means on the several items (36, 44? sz, 6I, 76) dealing with social security and extension of government functions are all well below 4. 0, indicating considerable sup- port for the liberal viewpoint.
The low means on the "government" items raise another question: Why do many individuals who are otherwise conservative support an increase in government activity? In some cases this inconsistency probably reflects ideological confusion or the beginnings of change from right to left or vice versa. However, this apparent contradiction may reflect something much more basic, namely a shift from traditional laissez-faire conservatism, whose economic unit was the individual competitive businessman, to a new type of conservatism whose economic unit is organized big business. As was pointed out earlier in this chapter, the assumption of liberalism-conservatism as a simple quantitative dimension holds only in the most general sense. It was for this reason, among others, that the theory guiding scale construction was presented in some detail. It is possible, then, for an individual to make a moderately high rather than a very high score, not because of any true liberal tendency, but because of a change in the nature of his conservatism. He is now willing to extend the functions of government for reasons that are the opposite of liberal, for he conceives of government as a tool of busi-
? POLITICO-ECONOMIC IDEOLOGY AND GROUP MEMBERSHIPS 163
ness rather than as a means of controlling corporate capital and of pre- venting concentrations of economic power. We shall return again to this question after considering the relation between the PEC and E scales.
3. THE SECOND PEC SCALE (FORM 60)
In Form 6o the PEC Scale (see Table 4(V)) was shortened to 14 items, and numerous changes were made in content and wording. Items 2 7 and 6o,
TABLE 4 (V)
THE SEcoND FoRM oF THE PoLITico-EcoNoMic CoNSERVATISM (PEC) SCALE (FoRM 6o)
4? " Labor unions should become stronger and have more influence generally.
9. " Most government controls over business should continue after the war.
I 3? America may not be perfect, but the American Way has brought us about
as close as human beings can get to a perfect society.
I 5. " If America had more men like Henry Wallace in office, we would get along
much better.
zo. " The artist and the professor are of just as much value to society as the
businessman and the manufacturer.
26. It would be dangerous for the U. S. to cooperate too closely with Russia. 27. The best political candidate to vote for is the one whose greatest interest is
in fighting vice and graft.
p. " No one should be allowed to earn more than $25,ooo a year.
37? " It is up to the government to make sure that everyone has a secure job and
a good standard of living.
43. " The government should own and operate all public utilities (transportation,
gas and electric, railroads, etc. ).
48. " Depressions can be prevented by proper government planning.
54. " Poverty could be almost entirely done away with if we made certain basic
changes in our social and economic system.
56. Men like Henry Ford or J. P. Morgan, who overcome all competition on
the road to success, are models for all young people to admire and imitate. 6o. In general, the best way of aiding our fellow men is to give time or money
to some worthy charity.
a These nine items are "liberal," the other five are "conservative. " A high score is given for agreement with the conservative items, disagreement with the liberal items.
referring to political candidates and the importance of charity, respectively, are reformulations of items that were unsuccessful in Form 78; the present formulations are, presumably, more clear-cut expressions of the trends initially hypothesized. Two items which worked relatively well in Form 78, numbers 27 (Rebellious ideas) and 61 (Security is bad), were placed in the Form 6o F scale (see Chapter VII) because they seemed on theoretical grounds to fit better there.
Several totally new items have been added. Item 13 (The American Way) was taken from the Form 78 E scale (see Chapter IV); it is transitional be- tween conservatism and ethnocentrism in that it expresses both conservative
? THE AUTHORIT ARIAN PERSONALITY
support of the politico-economic status quo and ethnocentric idealization of the ingroup. Taken literally, however, it seemed to fall more within the sphere of political ideology. Correlational analysis is required before this item can properly be placed within one scale or the other.
Three of the new items provide vivid reminders of the speed of historical change. Item 15 (Wallace) was formulated when the confirmation of Henry Wallace as Secretary of Commerce was the issue of the day. Item 26 (Russia) reflected the atmosphere of the initial postwar period, when cooperation rather than containment was the prevailing attitude toward Russia. Item 3I referred to President Roosevelt's wartime suggestion of a $25,ooo limit on yearly incomes. It will be noted that the Form 6o scale contains fewer gen- eralizations and more concrete references to specific issues than did Form 78.
Form 6o was administered in the summer of 1945 to several groups which were combined for statistical purposes as follows. (a) University of Oregon Student Women (N = 47), undergraduate students attending summer session courses in psychology. (b) Combined University of Oregon and University of California Student Women (N = 54), the Oregon group being obtained too late to be included in (a); the California group was a summer session education class containing mostly teachers and others of above college age. (c) University of Oregon and University of California Student Men
(N = 57), from the same classes as the (b) women. (d) Oregon Service Club Men (N = 68), obtained at luncheon meetings of service clubs (Ki- wanis, Lions, Rotary). 4
The last three groupings received the total Form 6o questionnaire in one sitting. However, the first group of Oregon Student Women received the questionnaire in two parts, A and B. Part A included the F scale and half of PEC, while Part B contained the E scale and the remaining half of PEC. The purpose of this division was to help determine whether the pres- ence of the E items had any effect on the responses to the F items; the results will be discussed in Chapter VII.
The reliability data for the Form 6o PEC Scale are presented in Table 5(V). The average reliability of . 70 is substantially the same as that of ? 73 for the initial form, and indicates that the changes in wording and content did not improve this property of the scale. The fact that the reliabilities, S. D. 's, and ranges vary so little among these four groups, as well as among those taking the first form (Form 78), suggests that a scale of this length and this degree of inclusiveness can hardly be expected to obtain an average re- liability greater than . 7-. 8. The main reason for the relatively low reliability of PEC, as compared with E, appears to lie in the fact that political ideology is intrinsically less organized and less consistent in the individual today than is ideology concerning group relations.
Once again the group means on PEC are significantly higher than those on 4 For a discussion of the sampling problems involved, see Chapter IV.
? Property
Re 1 i a b i l i t y
Mean (total) Mean (odd half) Mean (even half)
S. D. (total)
RELIABILITY OF THE PEC SCALE (FORM 60)
Grou1t
I II III IV
over-all
. 7 0
3. 92d 3. 74d 4. 23
1. 0-6. 1
S. D. S. D.
Range
(odd (even
half) half)
. 73
3. 72 3. 86(A)c 3. 58(B)c
. 90
. 97(A}c 1. 02(B}c
1. 2-5. 6
. 69
3. 82 3. 60 4. 03
1. 0-5. 5
. 69
3. 77 3. 55 3. 99
. 80
. 78 1. 05
1. 2-5. 0
. 70
4. 40 4. 06 4. 68
. 92
. 95 1. 14
1. 6-6. 1
POLITICO-ECONOMIC IDEOLOGY AND GROUP MEMBERSHIPS I6 ) TABlE 5 (V)
aThe four groups are:
Group 1: Group II:
Group Ill:
Group IV:
University of Oregon Student Women (N = 47)
University of Oregon and University of California Student Women (N = 54)
University of Oregon and University of California Student Men (N = 57)
Oregon Service Club Men (N? = 68)
brn obtaining the over-all means, the individual group means were not weighted by N.
cThe signs (A) and (B) refer to the two parts of the questionnaire given to Group I; half of PEC was in part (A), the other half in part (B).
The reliability for this group is based on the correlation between the A and B halves.
dThe Over-all mean (odd) (even), and S. D.
(odd) (even) include only the three groups taking the total form since the (A) and (B) halves of the split form do not correspond to the odd and even halves of the total form.
E (Chapter IV), suggesting that the level of conservatism is higher than the level of ethnocentrism. Again, the rank order of group means on PEC tends to follow that on E, with the Service Club Men being significantly more con- servative (beyond the 1 per cent level) than the combined university groups. These facts would lead us to expect a significant correlation between PEC and E (see Section C). While the Service Club Men are quite conservative on the average (mean of 4. 4), the lowest score being 1. 6, this group can by no means be considered ideologically homogeneous; indeed, it shows about the same degree of internal variability (range and S. D. ) as do the other groups. We are given another warning against stereotypy in thinking about groups and about group memberships as determinants of ideology. This is not to
say that service clubs are not "conservative groups" in terms of actual policy. Rather, it would appear that group policy and leadership, in this case
? 166 THE AUTHORIT ARIAN PERSONALITY
at least, reflect the average degree of conservatism, the conservative tradition, and frequently the immediate business interests of the group. However, to say that such-and-such is a conservative group, in terms of actual policy, is not necessarily to imply that all members are strongly conservative. Simi- larly, not all individuals who call themselves "New Deal Democrats" are thoroughly liberal in their personal ideologies; not all Catholics support the political program of the policy-makers of the Catholic Church; and so on. This is one of the main problems in bringing together the psychological and the sociological approaches; it is an especially great problem for that theory of social psychology which regards the individual adult as merely
a product or sum of his various group memberships.
The data on item analysis are presented in Table 6(V). The over-all D.
P. of 2. 08 is almost identical with that of 2. I4 on the initial PEC scale, as might be expected from the similar reliabilities. The best items deal for the most part with government functioning; ownership of utilities, controls over business, limitations on income. Item 4 (Unions) worked out relatively well (rank order 6) despite its having the highest over-all mean, 5? 35; even in the university groups the mean did not fall below 5. o. Item I5 (Wallace) came out similarly; it had the third best D. P. despite a mean of 5. 00. Other items with D. P. 's of over 2. 0 include 13 (American Way), 54 (Poverty), and 56 (Ford and Morgan).
The five poorest items are also of some interest. Three of these, 20 (Artists, businessmen), 27 (Political candidate), and 6o (Charity), are reformulations of poor items in Form 78. Almost none of these subjects disagreed with the idea that the artist and professor are as important as the businessmen, al- though the liberals agreed more emphatically than the conservatives (the difference being statistically significant only in the Service Club Men). The D. P. of 1. 06 on Item 27, while statistically significant, indicates considerable overlap between the high and low quartiles. Further exploratory research is required in order to determine possible differences between liberals and conservatives with respect to underlying imagery of "the good political candidate. " Item 48 (Depressions) is an example of not leaving well enough alone. In the initial form this item had a D. P. rank of 4. 5; in this form, after drastic revision, its rank was I2. Both the mean and D. P. on Item 26 (Dan- gerous to cooperate with Russia) are somewhat surprising. The mean of 2. 57 indicates that very few individuals agreed with this item. The D. P. of 1. 60 is more significant than it at first appears, because of the low mean, but it shows that even conservatives were divided on the Russian issue at the close of the war. This is shown most dramatically by the Service Club Men who,
although strongly conservative on most domestic issues, obtained on the Russian item a mean of 2. 5 I and a D. P. of ? 93? How and why slight support has, within less than two years, changed to bitter antagonism, is a problem beyond the scope of the present study.
? No.
I t e m
(Unions) 5. 43
Mean
D. P .
Mean
D. P . Mean
D. P . . . . . . .
4.
2. 65 4. 43 3. 54 4. 15 1. 97
5. 06 2. 64 4. 81 2. 26 4. 91 2. 22 4. 72 2. 64 1. 22 0. 38 2. 54 1. 47 3. 30 0. 34 4. 74 1. 98 4. 72 3. 72 4. 93 2. 77 2. 76 0. 85 3. 63 2. 02 3. 22 2. 12 2. 87 1. 73
1. 60 2. 94 3. 08 2. 33 0. 14 1. 52 1. 19 3. 42
5. 65 5. 87 4. 68 5. 84 1. 87 2. 51 3. 21 5. 56
2. 76 5? . 35 2. 82 4. 88 2. 16 4. 38 2. 51 5. 00 1. 15 1. 44 0. 93 2. 57 1. 99 3. 49 2. 12 5. 00 2. 99 4. 74
9. 13. 15. 20. 26. 27? .
(Gov't. controls)
(American Way)
(Wallace) 4. 79 (Artists and professors)
(Russia)
(Political candidate)
($25,000/yr. )
(Gov? t. and jobs)
2. 89 (2) 2. 36 (7) 2. 62 (3)
0 z 0
31. 37. 43. 48. 54. 56. 60.
2. 44 (5)
C)
(Utilities) (Depress ions) (Poverty) (Ford, Morgan) (Charity)
4. 61 3. 48 3. 09 1. 30 4. 21 1. 86 3. 00 1. 75 2. 47 0. 91
5. 82 3. 23 4. 41 1. 94 4. 31 2. 38 3. 34 3. 13 2. 88 2. 22
2. 61 (4) 4. 96 3. 36 (1)
~
Mean per item
3. 72 2. 16
3. 82 1. 94
3. 77 1. 91
4. 40 2. 31
~
aGroup I: Group II: Group III: Group IV:
University of Oregon Student Women (N =47)
University of Oregon and University of California Student Women (N =~) University of Oregon and University of California Student Men (N =57) Oregon Service Club Men (N =68)
l""::
TABLE 6 (V)
MEAI\5 AND DISCRIMINATCRY PORERS <F THE PEC SCAlE ITEM3 (FCRM 60~a
'"d
Group I Mean D .
