. :3; : E, ,, 1 l t b t . ‘j‘ a! R "'!¢fl§ . 3m, .0. 51's TH E311: ~. This is to certify that the thesis entitled Anomie and authoritarianism: A Study in validity presented by Alan Harvey Roberts has been accepted towards fulfillment of the requirements for M. A. degree inmogy Major professor Date—Nlay 20, 1952 0-169 ‘ ct A" ‘1“ ..\tr.. ”J. ; v. .a . ”(A . J‘a....4\\... .0“ l. 1...): ~ . ‘ f u“. v“!n3ul.fhnl".. . LWr iv 33. . . s . . v...‘ t .3. ‘ J'fljsJ (.51Hv.. llW‘ 71 (9th“) 4 . V n.-.. ,. ANOMIE BED AUTHORITARIANISM: A.STUDY IN VALIDITY By ALAN HARVEY ROBERTS A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1952 ACKNOWLEDGKELTS The author wishes to express his indebtedness to Dr. Milton Rokeach.under whose inspiration, guidance, and continued interest this study was undertaken and to whom the results are herewith dedicated. Grateful acknowledgment is also due to Dr. Dorothy/ Zietz and Mr. Bernard Ross of the Department of“ Social Service for their aid in securing interviewers. The writer extends his sincere thanks to the many unnamed students who willingly volunteered to administer the questionnaires. TABLE OF CONTE {TS PAGE INTRODUCTION . . . . . . . . . . . . . . . . . . . PROBLEM o . o . o . o o . o . o . o o . . . o o o ASSUMPTIONS AND HYPOTHESES . . . . . . . . . . o o NATURE MID I‘LEISUPEII’IEHT OF THE VARIABLES . . . . . CDCDCDFPH Ethnocentrism................. Anomie . . . . . . . . . . . . . . . . . . . . . 10 Authoritarianism . . . . . . . . . . . . . . . . 13 Education and Income . . . . . . . . . . . . . . 15 METHOD . . . . . . . . . . . . . . . . . . . . . . 18 Subjects . . . . . . . . . . . . . . . . . . . . 18 Procedure . . . . . . . . . . . . . . . . . . . 30 RESULTS . . . . . . . . . . . . . . . . . . . . . 27 DISCUSSION . . . . . . . . . . . . . . . . . . . . 40 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . 45 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . 48 APPENDIX A. Lansing Opinion Survey - (P) 50 APPENDIX B. Lansing Opinion Survey - (M) 55 TABLE I. II. III. IV. V. VI. VII. VIII. X. XI. LIST OF TABLES PAGE COmparison of Means and Standard Deviations for Samples P and M . . . . . . . . . . . 19 Age of Subjects, Samples P and M . . . . . 21 Education of Subjects, Samples P and M . . 22 Income of Subjects, Samples P and M . . . . 25 Means, Standard Deviations, and Ranges for Forms P and M of the A and F Scales . . . 28 Distribution of Scores on Forms P and M of the A and F Scales . . . . . . . . . . . 29 Intercorrelations among the Relevant Variables, Sample P . . . . . . . . . . . 50 Intercorrelations among the Relevant Variables, Sample M . . . . . . . . . . . 31 First-order Partial Correlations between Anomie and the Relevant Variables . . . . 54 First-order Partial Correlations between Authoritarianism and the Relevant Variables . . . . . . . . . . . . . . . . 55 Comparison of Correlations between E and A with Correlations between E and F Holding Relevant Variables Constant . . . . . . . 57 INTRODUCTION The socio-psychological factors which are related to prejudice have been eXplored a great deal in recent years with the implied or stated purpose of eradicating authoritarianism through an understanding of the factors which favor its acceptance. ‘ Outstanding among these studies is the research reported by Adorno and others in The Authoritarian Person- ality (l). The major hypothesis upon which this study was based was stated as follows: . . . that the political, economic, and social convictions of an individual often form a broad and coherent pattern, as if bound together by a ”mentality“ or “spirit,“ and this pattern is an ex- pression of deep-lying trends in his personality (l,p.l). The major finding of this study was that there is a close correspondence in the type of approach and outlook that a subject is likely to have in a great many areas such as family and sex adjustment, religion, and social and political philosophy. More specifically, it was found that the personality variables purportedly measured by the F scale (authoritarianism or fascism scale) were highly correlated with measures of etnnocentrism. A more recent study by Srole (9) has cast some doubt on the validity of these findings, however. Srole hypothesized that the phenomena of social dysfunction or disorganization, group alienation, or demoralization, all of which Srole subsumed under the concept of 'anomie”, comprise one of the prime forces contributing to ethno— centrism. He studies a sample of #01 Springfield, Mas- sachusetts bus riders using a five-item scale of social distance attitudes, a five-item F scale, and a five-item anomie (A) scale which he constructed himself. He concluded after an analysis of the manner in wnich these tnree Avariables were inter-related that it was the variable of anomie which was primarily related to prejudice and that authoritarianism, as measured by the P scale, “is no longer highly correlated with social distance attitudes toward minority out-groups, at least not independently of the psycho-sociological factors presumably measured by (the) Anomie Scale.” (9. p.7) A further finding of the Srole study was that the .relationships among anomie, authoritarianism and social. distance are different within different education—status ‘groups in the pepulation. His data showed that as the education of his subjects increased, the independent relationship between anomie and social distance decreased. For his subjects with greater than high School education, anomie showed no independent correlation with social distance. In the case of authoritarianism, the Opposite was true; as education increased, the independent relationship between authoritarianism and social distance increased. The term 'anomie” or “anomy” was first used in its present form by Durkheim in his works on Suicide (3) and the Division of Labor (2) during the latter part of the nineteenth century. In its original meaning, it referred to a disregard of law, particularly a disregard of divine law (7). As it is used today by sociologists, it refers to societal normlessness, cultural chaos and social dysfunction. Merton, in his analysis of social structure (6), states that anomie is likely to deveIOp in a society which emphasizes specific goals without a parallel emphasis upon institutional procedures. This differential emphasis upon goals and procedures leads to a preference for the most effective procedure whether culturally acceptable or not. As this process continues, the society becomes un- stable and anomie deveIOps. PROBLEM The findings and conclusions of Srole are of extreme theoretical importance. His approach to the problem of prejudice contrasts sharply with the theory outlined in Thegguthoritarian Personality (l). Srole would contend that prejudice is a function of anomie which is, in turn, a function of socioeconomic status. The California Study, on the other hand, would contend that prejudice is a part ' of a broad and coherant pattern of beliefs and attitudes which are eXpressions of deep personality trends, and thus it is these personality trends, as tapped by the F scale, which account for prejudice. Srole's contention that anomie is related to social status was tested in his study by using the educational level of the subject as a “rough indicator of class status" (9,p.7). He found that anomie was the prime variable re— lated to social distance only for subjects of high school education or less, and that the relationship decreased relative to the level of education. The Opposite was true for authoritarianism; the closer the relationship between authoritarianism, the higher the educational level of the subject. These relationships, plus the fact that both the A and F scales were phrased so that agreement would indicate the presence of anomie or authoritarianism, suggest that the results of the Srole study may have been due to a tendency for subjects of low education to agree with the anomie statements regardless of content in order to avoid the threat which would arise when tney disagreed with statements which they felt were endorsed by the interviewer. One purpose of this paper is to determine to what extent this conformity factor might have been operating in Srole's study. A second purpose is to test certain derivations of Srole's theory on the one hand, and the alternative derivations of the California-Group theory on the other. ASSUMPTIONS AND HYPOTHESES Two assumptions were made in constructing the hypotheses to be tested in this study: 1. It was assumed that the anomie and author- itarianism scale items could be semantically reversed without altering their meaning or content. 2. It was assumed that a replication of the Srole study would yield results which duplicated those of the original study; namely, that anomie would be significantly and independently correlated with prejudice, and that when anomie is held constant, authoritarianism would not be independently related to prejudice. On the basis of these assumptions, it was now possible to test the following hypotheses: l. (a) If conformity was a significant factor in determining agreement with the items on Srole'e anomie scale, then a semantic reversal of the anomie scale should lead to a negative instead of a positive relationship between anomie and prejudice, and (b) if conformity was a significant factor in determining agreement with the items on the authoritarianism scale, then a semantic reversal of these items should also lead to a negative instead of a positive relation- ship between authoritarianism and prejudice. ' 2. If Srole's theory that anomie is.a function of socioeconomic status is correct, we should expect to find that (a) anomie is related to a more direct measure of socioeconomic status in addition to his education measure, (b) anomie is related to this more direct measure of status even when education is held constant, and (c) anomie is not related to prejudice when status, as measured by education and income, is held constant. 3. If, orlthe other hand, the contention of the California Study that prejudice is to a large extent a function of deep-lying personality trends is correct, then we should expect to find that these trends, as measured by the authoritarianism scale, are related to prejudice independently of measures of education and status. NATURE AND MEASUREMENT OF VARIABLES Five major variables were considered in this study. They are: ethnocentrism, authoritarianism, anomie, education and income. Ethnocentrism. The prime purpose of this study is to determine the relative eXplanatory power of anomie and authoritarianism as determdners of prejudice. The term “ethnocentrism“ has been defined as "prejudice, broadly conceived.“ (l,p.102). In.this study, ethnocentrism refers to a relatively consistent hostile feeling toward any group, even though the hostility or discrimination is not built upon racial or religious grounds. 'The California Study points out that the term ethnocentrism shifts the emphasis from “race“ to “ethnic group." (l,p.103). In this paper the term will be used interchangeably with the terms ”social distance“ and "prejudice.' the California Study summarizes its findings by stating: . . .in general, ethnocentrism is based on a pervasive and rigid ingroup-outgroup distinction; it involves stereo- typed negative imagery and hostile attitudes regarding outgroups, stereotyped positive imagery and submissive attitudes regarding ingroups, and a hierarchical, author- itarian view of group interaction in.which ingroups are rightly dominant, outgroups subordinate (1,p.150). 9 To measure this variable, a revision of the ten—item Ethnocentrism (E) Scale, Form #5, from the California Study was used (l,p.l28). This scale contains items relating to Negroes, Jews, zootsuiters, foreigners, and natibnalism. The California Study found that this scale had an over-all reliability of .79 with an N of 3&0. Broken down, these reliabilities ranged from .65 for a sample of 110 San Quentin Men prisoners, to .91 for a sample of fifty working class men and women. The ten-item scale as it appeared in our study was as follows: 8. Zootsuiters prove that when people of their type have too much money and freedom, they just take advantage and cause trouble. lo. 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. 11. The worst danger to real Americanism during the last fifty years has come from foreign ideas and agitators. 13. One trouble with Jewish businessmen is that they stick together and prevent other people from having a fair chance in competition. lb. It would be a mistake ever to have Negroes for foremen and leaders over whites. 18. I can.hard1y imagine myself marrying a Jew. 19. If Negroes live poorly, it's mainly because they are naturally lazy, ignorant and without self control. 10 22. There may be exceptions, but in general, Jews are pretty much alike. 24. The trouble with letting Jews into a nice neigh— borhood is that they gradually give it a typical Jewish atmosphere. 25. Even though the United States is a member of the United Nations, the United States must be sure that she loses none of her independence and complete powers in matters that affect this country. These items are numbered as they appear on the questionnaires used in this study. (See Appendix A and B.) The E scale was scored by assigning a value of one if the subject disagreed with the statement, two if he was undecided, and three if he agreed. This method allows for a range of from ten to thirty, the higher scores indicating higher ethnocentrism. As originally designed by the California Study (1), the subject was presented with six choices of response for each item: slight, moderate, or strong agreement, and the same degrees of disagreement. The number of choices was reduced from six to three in order that this study conform as closely as possible with the design of the Srole study.‘ Anomie. Anomie, as discussed previously, has been defined by Srole (9,p.2) as social dysfunction or disorgani— zation, group alienation, and demoralization. To measure 11 this variable, Srole constructed thirty agree-disagree type items which he pretested by means of fifty interviews. From these items he selected five items, each one representing a hypothetical component of anomie. The first component was the individual's feeling that community leadership was not meeting his needs. The item which purportedly measures this component reads: 6. There's little use writing to public officials because often they aren't really interested in the problems of the average man. The second component concerned the individual's perception of the social order as unstable, changing, unpredictable and orderless, leading to the feeling that he could do nothing to direct his future or plan ahead: 9. Nowadays a person has to live pretty much for today and let tomorrow take care of itself. The next component involved a feeling of lack of progress or inability to progress with respect to the self: 15. In spite of what sose pe0ple say, the lot of the average man is getting worse, not better. The fourth component which, as Srole points out, is perhaps closest to Durkheim's conception of anomie, has to do with a loss of meaning of norms, values, and goals leading to a feeling that life itself is futile and with- 12 out meaning: 23. It's hardly fair to bring children into the world with the way things look for the future. The final anomie component is defined as the individ- ual's feeling that he could no longer depend upon immediate personal relationships for support: 20. These days a person doesn't really know whom he can count on. The numbering for the above statements corresponds to the item numbers in Form P (see Appendix A). This five-item scale was also scored on a three- point scale of disagree, undecided, and agree, allowing for a range in scores of from five to fifteen. Again, the higher scores indicate higher anomie. Srole reports that a latent structure analysis on these items shows that they apparently measure one coherent attribute or continuum. He also finds that there is a high relationship between this scale and the five-item F scale, but that the two scales apparently measure different aspects of attitude phenomena. In order to test our first hypotheses, it was necessary to reverse each of the five anomie items so that disagree- ment with the item in its revised form would indicate anomie. 13 It was assumed that the items, when reversed, were seman— tically and effectively equivalent to the items listed above. Numbered as they appeared in Form M of our ques- tionnaire (Appendix B), the five reversed anomie items read as follows: 6. writing to public officials often does some good because many public officials are really interested in the problems of the average man. 9. It pays for a person to sacrifice today in order to provide for tomorrow. 15. In spite of what some people say, the lot of the average man is getting better, not worse. 20. In times of trouble, a person can usually find one or more people he can count on. 23. In spite of the way things look for the future, it's hardly fair to stop having children. In order for a high score to indicate anomie, it was necessary to score this scale by assigning a value of one for agree answers, two for undecided, and three for disagree. The range is the same as that for the positively worded A scale. Authoritarianism. Two purposes led to the construction of the original scale of authoritarianism (1,p.222). The first was to construct a scale which would correlate highly with prejudice without appearing to have this aim and with. 14 out mentioning the name of any minority group. The second aim was to construct a scale which would yield a valid estimate of antidemocratic tendencies at the personality level. . The final twenty—nine item scale had an average re- liability of .90. The correlations of this scale with the E scale ranged from .56 to .87 with .75 as an over-all correlation. They state (1,p.264) that if these coefficients were corrected for attenuation their magnitude would be about .90. In the five—item version of the F scale used by Srole, three of the items appear in the original F scale (items 7, 12 and 16), but the source of the other two items could not be ascertained. The scale contains the following items, numbered as they appear in Form P of our questionnaire (see Appendix A): 7. There are two kinds of people in the world: the weak and the strong. 12. The most important thing to teach children is absolute obedience to their parents. 16. Prison is too good for sex criminals. They should be publicly whipped or worse. 17. Any good leader should be strict with people under him in order to gain their respect. 15 21. No decent man can respect a woman who has had sex relations before marriage. This scale was scored in the same manner as the A scale, and has the same possible range. As was done with the A scale, a reversed form of this scale was constructed so that disagreement with the alter- native form.indicated authoritarianism. These items were assumed to be equivalent to the original items. As they appeared in Form M of the questionnaire, these items read as follows: N 7. It is a mistake to think that there are two kinds of people in the world: the weak and the strong. 12. To teach children absolute obedience to their parents is not the most important thing. 16. Although some peeple say that sex criminals should be publicly whipped or worse, it is better to treat them just like other criminals. 17. Any good leader should be lenient with people under him in order to gain their respect. 21. .A man can respect a woman even though she has had sex relations before marriage. This scale is scored in the same manner as the reversed A scale, and has a comparable range. Education and Income. The measurement of these two variables presented some problems. Parten (8,p.hl6) 16 reports that these two variables constitute the data most difficult to obtain in surveys. In one of our samples, it was found that 100 per cent of the subjects reported their education and 90 per cent reported their income. For the other sample, 99 per cent reported their education and 96 per cent reported their income. (see Tables III and IV.) This rather remarkable success, particularly with education, can perhaps be accounted for by the way in wnich the questions were asked. The interviewer, in the case of education, was instructed to ask the subject, ”Do you remember the name of the last school you went to?‘l The answer to this question was not recorded, but it was felt that after the respondent had answered it, he would be more likely to tell how many years he had completed in school and less likely to exaggerate his answer. There was somewhat more difficulty with the income item. At first, the respondent was handed a card with the various income brackets listed on it and he was asked, ”would you tell me which letter on this card best represents your yearly income?“ However, it was found that this approach was not only clumsy, but too direct. The procedure 1? was then changed so that the schedule was handed to the respondent and he was asked to mark his yearly income on the sheet himself. The number of refusals decreased notice— ably after this procedure was initiated, although some were still obtained, particularly in the upper income neighbor- hoods. In some instances, this study was confused with a widely publicized survey which was being conducted simul- taneously with this one, and which was primarily interested in obtaining financial information from the reapondents. Both education and income may be considered as measures of socio-economic status according to Parten (8). In some instances, educational level may serve as an indicator, but in most instances, economic level is the better indicator. Income has the advantage of being a more direct measurement of status, particularly when it is being related to factors such as prejudice, since education may be considered as being related to these factors in ways other than as a measure of status. However, income has the disadVantage of being more difficult to measure accurately. ME TE 0 D Subjects. The subjects for this.study were all adult, white, non-Jewish, native American residents of Lansing, Michigan. Two samples were used, each randomly selected by a procedure which will be described in a later part of this paper. The original ethnocentrism, anomie and authoritar- ianism scales were admirnstered to one sample, which we will hereafter refer to as the P or “Plus“ sample. To our second sample, hereafter referred to as the h or "Minus" sample, we administered the ethnocentrism scale, and the reversed forms of the anomie and authoritarianism scales. Sample P contained eighty—six subjects of which thirty- eight were males and forty—eight were females. The M sample contained twenty—five males and forty-seven females for a total of seventy-two subjects. These two samples were compared on the basis of income, education, age, and ethnocentrism by McKitrick (5) with the results as shown in Table 1. None of the means or standard deviations for the two samples are significantly different from each other except for education, and McKitrick has shown that the significance of the difference for this variable could be 19 TABLE I COMPARISON OF MEANS AND STANDARD DEVIATIONS FOR SAMPLES P AND M Standard Mean deviation Variable P M P M Education 10.47 11.47 2.76 5.25 Ethnocentrism 22.6 21.1 5.1 4.8 20 reduced to below the 5 per cent level by removing one extreme case from each.sample. McKitrick also reports that sample P contains 56 per cent females as compared with 65 per cent females in sample M, and that sample P contains 90 per cent married respondents as compared with 83 per cent married respondents in sample M. Again, there is no significant difference between these percentages for the two samples. The age, education and income for the two samples have been broken down in Tables II, III and IV respectively. It may be concluded from.this data that there were no essential differences between the two samples, P and M. Procedure. In order to test the hypostheses, two interview schedules were constructed: Form P and Form M. Each schedule has essentially the same format, and was con- structed throughout with the consideration that, in many instances, the subject would insist upon looking it over before agreeing to answer the questions. There were also printed on the forms explicit directions for administration of each item since several interviewers were used. At the top left of each schedule was printed the name of Michigan State College for prestige purposes, and TABLE II AGE OF SUBJECTS, SAMPLES P AND M Male 3 Age 1 B m a m 21-24_ 25-28 29~32 53-56 57-40 41-44 45-48 49-52 55-56 57-60 61-64 . 65 and over Not reported Total oupmmmepmamuaw owommmewammwwfi ()1 CO [0 U! Hmwuuqumwmwfiwg a lqoqamumewumawfii p O) EDUCATION OF SUBJECTS, SAMPLES P AND M TABLE III Years Males Females completed ‘P M P 1% 1-2 'U 15 O ‘0 5-4 0 0 1 Or 5-6 2 l 1 1 7-8 10 4 15 8 9-10 5 2 7 8 11-12 16 8 21 18 15-14 4 5 2 5 15-16 2 2 2 3 17-18 1 1 l 5 19-20 0 2 O 0 Not reported 0 O O 1 Total 58 25 48 47 22 TABLE IV INCOME OF SUBJECTS, SAMPLES P AND M $13 a CO L Females Income ' IBSS than $1000 1000-1999 2000-2999 5000-5999 4000-4999 5000-5999 6000-6999 7000-7999 8000-8999 9000-9999 10,000 and over Not rgported Total mmmpmamoowoww% ommooaommmows |-’ meowmemwommdz 3 CfiCflOONFU‘ICDCOOMbNJM (A (D N) 01 .p. q 24 space was provided for the date of administration. The top right-hand corner contained a place for the Case Number and Block Number. The name of the survey, "Lansing Opinion Survey,” then followed. It was felt that this name was innocuous enough so as not to bias the subject in any manner. Following this was the main body of the question— naire including the introductory statements of the inter- viewer and the instructions to the subject which were standard for all interviews. The introductory statement read as follows: Q I am . . . . . .from Michigan State College. We are making a survey of what the peOple of Lansing think and feel about a number of important social questions. We are not interested in names; there's nothing personal in this research. YOur Opinion is important for us to understand how people are thinking nowadays. We will try to cover many points of view with these questions. You may find yourself agreeing strongly with some of the statements, disagreeing just as strongly with others, and perhaps.uncertain about others. Whether you agree or disagree with any statement, you can be sure that many other people feel the same way you do. The interviewers were instructed to deviate from this introduction only if it was necessary to do so in order to secure the interview. The words "survey'I and ”research,“ and the phrases I'not interested in names" and 'nothing personal“ were used as suggested by Parten (8,p.350) in 25 ‘order to convey the impression of a harmless statistical survey. All background information concerning the subject except for income was obtained before asking the attitude statements in order to I‘warm up” the subject. It was felt that income should be obtained last, after the greatest rapport with the subject had been obtained, and also so that refusal to answer would not bias responses to the attitude statements. The two schedules differed only with respect to the attitude_items within them. Form P contained the ten ethnocentrism items, the five-item anomie scale, and the five-item authoritarianism scale. Form M contained the same ethnocentrism items, the reversed form of the A scale, and the reversed form of the F scale. The interviewers, in addition to the author and McKitrick, were all advanced undergraduate and graduate Social Service majors at Michigan State College, each of whom had taken at least one course in interviewing tech- niques and procedures. They were given a one-hour orienta- tion on the particular methodology required for this survey before being allowed to proceed. 26 Each interviewer was assigned a block number and given an equal number of Form P and Form M schedules. The interviewer was then instructed to alternate the two forms for each interview, thus insuring a nearly equal distribution for each form on every block. The discrepancy of ten cases between the two samples may be accounted for by an accumulation of chance errors during the total interviewing time, and it may be assumed that this discrepancy does not affect the results of this study materially. RESULTS The attitude scales for both samples were scored in the manner described earlier in this paper. The means and standard deviations for the four subscales are shown in Table V. It may be noted from this table that the range and variability of both the AIM) and F(M) scales are greatly reduced as compared to the A and F scales. The distributions for all four scales are shown in Table VI. The distributions for the A, F, and F(M) scales are all approximately normal, although there is a tendency toward bimodality in all three distributions, and all are skewed toward either the high or low end. The distribution for the A(M) scale, on the other hand, is J-shaped. This extreme deviation from normality suggests that our reverSal of the A scale produced conformity—type statements with which the majority of reapondents tend to agree. The Pearson.product-moment coefficient of correlation was then computed among all the relative variables, separ- ately for both samples. The results are shown in Tables VII and VIII. For Table VII, with an N of 86, the r must be at least .21 to be significantly different from zero at the 5 per cent level of confidence, and .28 to be signifi- TABLE V MEANS, STANDARD DEVIATIONS, AND RANGES FOR FORMS P AND M OF THE A AND F SCALES Scale Mean S.D. Range N II 9300 ‘3;04 5-1 '86 A(N) 6.40 1.56 5-11 72 F 11.01 2.79 5-15 86 F(Ml, 9.26 2.16 5-14 72 28 TABLE VI DISTRIBUTION OF SCORES ON FORMS P AND M OF THE A AND F SCALES . Scale Score A ATM? F FUN) 15- 4 0 _‘ 12—’ II 15-14 15 0 25 6 11-12 16 2 16 17 9-10 17 8 22 24 7-8 21 22 7 17 5-6 15 46 4 8 Total 86 72 86 72 TABLE VII INTERCORRELATIONS AMONG THE REIEVANT VARIABLES, SAMPLE P (Nu-86) A E ‘F Ed Inc A .55’ .47’ -.51 -.41 .64 -.50 -.25 F -.45 -.24 Ed- .47 Inc .21 significant at the 5% level of confidence .28 significant at the 1% level of confidence TABLE VIII INTERCORRELATIONS AMONG THE RELEVANT VARIAEIES, SAMPLE M (N='72) A(N)- E F(M) Ed Inc NM) «19 ~30 -.07 -728 E .07 -.51 «.20 F(M) .09 -.01 Ed .56 Inc .25 significant at the 5% level of confidence .50 significant at the 1% level of confidence 3]. 32 cantly different from zero at the 1 per cent lever of con- fidence. For Table VIII, the minimum significant r's are .23 and .30 at the 5 per cent and l per cent levels of confidence, reapectively. In both samples, significant correlations were found between E and education, and income and education. The relationship between E and income was significant in the 5 per cent level for sample P, but was not significant for sample M. The correlation between E and education is unexpectedly high in both instances. This may be accounted for by the fact that the scale was administered orally rather than as a paper-and-pencil test. For sample P, as shown in Table VII, both A.and F correlate highly with E, the correlation between F and E (.Oh) being somewhat greater than the correlation between A and E (.55). A shows higher correlations with both measures of social status than does F. We find that anomie is correlated with income {-.41), this correlation being significantly different from zero at the l per cent level of confidence. This would tend to support Srole's contention that anomie is a function of socioeconomic status. 33 The data for sample M, as shown in Table VIII, shows that there are no significant relationships among the scores on the reversed scales and the other variables except be— tween A(M) and income which has an r significant at the 5 per cent level of confidence. In order to further test our hypotheses, a series of first-order partial correlations was computed for the variables in sample P. By this methOd, we are able to nullify the effect of a third variable upon two variables which are being correlated. (4,p.3u5). The first-order partials relative to this study are shown in Tables IX and X. In these tables, the minimum significant r at the 5 per cent level of confidence is .22 except for correlations between A and income and F and income where the minimum significant r is .24. At the 1 per cent level of confidence, the null hypothesis may be discarded if r is .28 or greater, again with the exception of correlations between A and in- come, and F and income where r must be .31 or greater. Our final treatment of the data consisted of determinp ing the relative relationships between.A and E, and F and E, holding both education and income constant. This was done according to the method outlined by Guilford (4,p.346) TABLE DI FIRST-ORDER PARTIAL CORRELATIONS BETWEEN AN OMIE AND THE RELEVANT VARIABLES CorreIat ion Var iEble he Id be tween constant anomie and: E _F 4 Ed Inc Ethnocentfism .57 .59 . 51 Authoritarianism( F) .18 .51 .45 Education -.52 -.57 -.59 Income -.55 -.55 -.22 35 TABLE X FIRST-ORDER PARTIAL CORRELATIONS BETWEEN AUTHORITARIANISM AND THE RELEVANT VARIABLES CorrElatIOn it between Variables held authoritarianism. constant and: E A Ed A Inc thnocentrism .55 .54” .62 Anemia .18 .51.. .45 Education -.20 -.27 -.59 111001716 “'0 12 -006 “.04 36 for computing second—order partial correlations. The results are as follows: rAE.EDInc ~“0 rFE.EDInc = ’5” These r's are not significantly different from each other, the standard error of the difference being .10, and the t being 1.45. In computing this t, the standard error of the difference was corrected for the correlation be— tween samples as suggested by Guilford (4,p.224). A summary of the data shown in the previous talbes may be found in Table XI in which we compare the various relationships between A and E with the relationships be— tween F and E. It may be noted that the correlation between E and F is, in each instance, greater than the correlation between E and A. Although none of these correlations are significantly different frouleach other, the trend suggests that F is more closely related to ethnocentrism than anomie. The hypothesis that anonde is related to measures of status even when education is held constant is not supported by the data shown in Table IX. The correlation between anomie and income when education is held constant (—.22) is TABLEIXI COMPARISON OF CORRELATIONS BETWEEN E AND A WITH CORRELATIONS BETWEEN E AND F HOLDING RELEVANT VARIABLES CONSTANT : 5 .- I'EA '. 5 rEF - '54 TEA.Inc : '51 REF.Inc Z 962 I'EA.Ed 3 '59 rEF.Ed Z '54 I'EA.F 3 '57 I‘EF.A - '55 I’EA.Ec11nc 3 '40 rEF.EdInc 3 '54 37 38 not significantly different from zero. This finding does not support Srole's contention that anomie is a function of socioeconomic status. It shows that most of the cor- relation between anomie and income can be accounted for by the fact that education and income are highly related. We may note also in Table IX that the correlation between anomie and education when income is held constant (—.39) is significantly different from.sero beyond the l per cent level of confidence. The data does not support the hypothesis that, when education and income are held constant, anomie and prejudice would be unrelated. Instead, it was found that the cor- relation between A and E when income and education are held constant is .40 which is, again, significant beyond the l per cent level of confidence (Table XI). This finding is inconsistant with Srole's theory. The hypothesis that authoritarianism and prejudice would be correlated independently of income and status is supported by the data shown in Tables X and XI. The cor- relation between F and E when income is held constant (.62), when education is held constant (.54), and when both are held constant (.54) are not significantly different from 39 the correlation between E and F when none of these variables are held constant (.64). DISCUSSION The first thing that is apparent from the results of this study is that neither one of the basic assumptions of this study was substantiated. Our analysis of the data shows that the A(M) and F(M) scales do not correlate sig— nificantly with any of the variables (except income in one instance), in either a positive or negative direction. From this it may be concluded that the two M scales do not measure the same thing as the two P scales. No Objective analysis of this discrepancy will be attempted, but it may perhaps be fruitful to criticize these items in terms of the criteria set down by Wang (10) for writing attitude statements. Wang states that ”an attitude statement must be debatable.I When the statement represents an Opinion which is generally accepted, it cannot adequately be used to measure attitudes. Ifwould appear that statements 6, 9, 15, 20, and 23 of Form M all fail to meet this criteria. For example, statement 20, “In times of trouble, a person can usually find one or more peOple he can count on,‘ is hardly debatable. The phrasing of the statement almost demands agreement. 41 Wang also feels that an attitude statement should be susceptible to only one interpretation. It appears that statements 7 (It is a mistake to think that there are two kinds of people in the world: the weak and the strong) and 16 (Although some peOple say that sex criminals should be publicly whipped or worse, it is better to treat them just like other criminals) may possibly be interpreted in more than one way. Statement 23 (In spite of the way things look for the future, it's hardly fair to stop having children) can be criticized on grounds that it is “double-barreled.' Subjects might be agreeing or disagreeing with the implica- tion of either the first or second half of the statement. Any or all of the above criticisms of the items in the reversed scales may be enough to account for the fact that these scales do not appear to measure the same thing as the original A and F scales- Our second assumption, that a replication of the Srole study would duplicate his results, is also not substantiated. It will be recalled that Srole found that while both anomie and authoritarianism were related to social distance, authoritarianism did not correlate with prejudice when anomie 42 was held constant. The data from this study shows that A and E correlate .37 when F is held constant, and that F and E correlate .53 when A is held constant. This contradicts the results obtained by Srole. Our failure to construct reversed A and F scale items which were equivilant to the original items did not allow us to test our first hypothesis to the effect that if conformity was a factor which influenced the results of the Srole study, then we would find that the reversed A scale correlated neggtively with prejudice, and the reversed F scale correlated positively with prejudice. Further research along these lines may prove fruitful if it is possible to construct equivilant reversed items. Our second hypothesis stated that if the Srole theory were correct and anomie were a function of socioeconomic status, then we would eXpect anomie (a) to be related to income, (b) to be related to income when education was held constant, and (c) to be unrelated to prejudice when education and income were held constant. It was found that anomie and income correlate significantly (-.41), but that this relationship drops to -.22 (not 43 significant) when education is held constant. This would suggest that anomie.is not related to status as Srole contends, but to the education of the subject. It was further found that when income and education were held constant, anomie was still significantly correlated with prejudice (.40). This would suggest that, although anomie and prejudice are significantly and independently re- lated, the relationship cannot be accounted for by a correlation between anomie and socioeconomic status as Srole would contend. And finally, our third hypothesis stated that if the theory underlying the California Study is essentially correct and the F scale is a measure Of personality trends, than we should expect to find authoritarianism and prej- udice-correlated independently of measures of education and status. This hypothesis is supported by the data which shows that F and E are significantly related when income is held constant (.62), when education is held constant (.54), and when both are held constant (.54). It should be emphasized, however, that anomie and ethnocentrism are also significantly related when income is held constant (.51), when education is held constant 44 (.39), and when both are held constant (.40). These data would suggest that further research upon this variable’is warranted. However, no evidence was found in this study to suggest that authoritarianism is not related to prejudice independ- ently of anomie. Rather, the data show that authoritarianism bears a closer relationship to prejudice than does anomie. SUMMARY.AND CONCLUSIONS It was the purpose of this paper to examine the re— 1ative validities of anomie and authoritarianism as deter- miners of prejudice. The F scale, according to Th3 Apthoritarian Personality (1), reflects deep—lying person— ality trends whereas the A scale, according to Srole (9), reflects socioeconomic status. Srole's finding that the relationships among anomie, authoritarianism, and ethnocentrism were different for different levels of education suggested that the results of his study may have been caused by a tendency for subjects of low education to agree with anomie statements in order to avoid the threat which would arise when they disagreed with statements which they felt were endorsed by the interviewer. Specifically, this study was designed to test three hypotheses: 1. (a) If conformity was a significant factor in determining agreement with the items on Srole's anomie scale, then a semantic reversal of the anomie scale should lead to a negative instead Of a positive re- lationship between anomie and prejudice, and (b) if conformity was a significant factor in determining agreement with the items on the authoritarianism scale, then a semantic reversal of these items should also as lead to a negative instead of a positive relationship between authoritarianism and prejudice. 2. If Srole's theory that anomie is a function of socioeconomic status is correct, we should expect to find that (a) anomie is related to a more direct measure of socioeconomic status in addition to his education measure, (b) anomie is related to this more direct measure of status even when education is held constant, and (c) anomie is not related to prejudice when status, as measured by education and income, is held constant. 3. If, on the other hand, the contention of the California Study that prejudice is to a large extent a function of deep-lying personality trends is correct, then‘we should expect to find that these trends, as measured by the authoritarianism scale, are related to prejudice independently of measures of education and status. To test these hypotheses, the same anomie and auth- oritarian items employed by Srole were administered to a sample of eighty-six adult residents of Lansing, Michigan, together with a ten-item ethnocentrism scale. Reversed forms of the anomie and authoritarian scales were admin- istered to an equivalent sample Of seventy-two Lansing residents together with the same ethnocentrism scale. The near zero correlations of the reversed scale items with the relevant variables indicated that these scales were not equivalent to the original scales; hence, we were not able to test the first hypothesis. The data confirmed only a part of our'second hypothesis. A7 It was found that anomie is significantly related to income, but that this relationship is not significant when education is held constant. It was further found that anomie and ethnocentrism are significantly related even when education and income are held constant. These findings fail to confirm Srole's theory that the relation- ship between anomie and ethnocentrism is determined by status. Our third hypothesis to the effect that authoritarianism and ethnocentrism are related independently of income and education was confirmed by the data. This tends to support the point of view of the California Study which states that ethnocentrism is determined by personality trends which are measured by the F scale. In general, it may be concluded that the data from this experiment do not support Srole's contention that authoritarianism is not related to ethnocentrism inde— pendently of anomie. On the contrary, the present data suggest that authoritarianism bears a closer relationship to prejudice than does anomie. However, the small but significant relationship between anomie and ethnocentrism indicates that further research is merited. 4. 5. 6. 8. 9. BIBLIOGRAPHY Adorno, T. W., e£_£l., The Authoritarian Personality. New York: Harper and Brothers, 1950. Durkheim, Emile, The Division of Labor in Society. Translated by_Ceorge Simpson. Glencoe, Illinois: The Free Press, 1955. Durkheim, Emile, Suicide. Translated by John Spaulding and George Simpson. Glencoe, Illinois: The Free Press, 1951. Guilford, J. P., Fundamental Statistics in Psychology and Education. Second Edition; New York: MbGraw-Hill seek Company, Inc., 1950. McKitrick, Keith, "Anomie and Authoritarianism: A Study in Reliability." Unpublished Master's thesis, Michigan State College, East Lansing, Michigan, 1952. Merton, Robert K., Social Theory and Social Structure. Glencoe, Illin0is: The Free Press, 1949. iurray, James A., editor, A New English Dictionary of Historical Principles. Oxford: The Clarendoanress, 1888. Parten, Mildred B., Suryeys, Polls and Samples. New York: Harper and Brothers, 1950. Srole, Leo, "Social Dysfunction, Personality and Social Distance Attitudes." Unpublished paper read before the meeting of the American Sociological Society, Chicago, Illinois, 1951. 10. Wang, Charles K., "Suggested Criteria for Writing Attitude Statements," JOurnal of Social Psychology, 5:567-575, 1952. APPENDIX MICHIGAN STAT—...; O ..... ......O...O.... COLLJGE Case No. Date APPENDIXB ............ Block No. LaNsmg oPIquwmm - (M) DON‘T ASK QUhSTION ONE, BUT RECORD: 1. Is the respondent male or female? Male......... Female....... BEGIN HERE INTERVIJWER: PUT YOUR NAME IN THE SPACE BELOW. I AM .. .......... . ..... .............. FROM MICLIGAN STATE COLLJGE. We are making a survey of what the people of Lansing think and feel about a number of important social questions. We are not interested in names; there's nothing personal in this research. Your opinion is important for us to understand how people are thinking nowadays. AFTER GATTING SETTLED FOR T}J INTQR'IJW, S.Y: We will try to cover many points of view with these ques— tions. You may find yourself agreeing strongly with some of the statements, disagreeing just as strongly with others, and perhaps uncertain about others. Whether you agree or disagree with any statement, you can be sure that many other people feel the same way you do. You remember that I said that we are not interested in your name, but for statistical purposes we'd like to have a little in- formation about yourSelf. For example: 2. What is your date of birth?.................... ......... ........ 3. What is your marital status? Married....... ..... ..... Single........ ... ...... Other .................. to rou remember the name of the last school you went to? DO NOT RECORD. What was the last year you completed in school? CIRChE ONE. 1 2 3 4 5 6 7.8 9 10 11 12 13 14 15 16 l? 18 19 20 5. What church were you brought up in? (IF "PBOTDSTANT", What denomination?) I will read you each of the statements and you tell me whether you agree, disagree or can't decide. Lansing Opinion Survey~(M) .....6. 0000.70 0.0.08. 00.0.9. .....10. .....ll. .....12. .....13. .....14. .....15. .....16. .....17. 0.0.0180 0.0.0.190 .....20. ..,,.21. .....22. .....23. ..OOOZL". Page 2 Writing to public officials often does some good because many public officials are really interesed in the problems of the average man. It is a mistake to think that there are two kinds of people in the world: the weak and the strong. Zootsuiters prove that when people of their type have too much money and freedom, they just take advantage and cause trouble. It pays for a person to sacrifice today in order to provide for tomorrow. 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. The worst danger to real Americanism during the last 50 years has come from foreign ideas and agitators. To teach children absolute obedience to their parents is not the most important thing. One trouble with Jewish businessmen is that they stick together and prevent other people from having a fair chance in competition. It would be a mistake ever to have Negroes fer foremen and leaders over whites. In spite of what some people say, the lot of the average man is get~ ting better, not worse. Although some people say that sex criminals should be publicly whipped or worse it is better to treat them just like other criminals. Any good leader should be lenient with people under him in order to gain their respect. I can hardly imagine myself marrying a Jew. If Negroes live poorly, it's mainly because they are naturally lazy, ignorant and without se1f~control. In times of trouble, a person can usually find one or more people he can count on. A.man can respect a woman even though she has had sex relations be— fore marriage. There may be a few exceptions, but in general, Jews are pretty much alike. In spite of the way things look for the future, it‘s hardly fair to step having children. The trouble with letting Jews into a nice neighborhood is that they gradually give it a typical Jewish atmospheres Lansing Opinion Survey - (M) Page 3 000.025. 26. Even though the United States is a member of the United Nations, The United States must be sure that she loses none of her indepen~ dence and complete power in matters that affect this country. Some days a person feels pretty good; other days he feels pretty bad about things in general. Would you say that you're feeling better than usual, the same as usual or worse than usual today? BetterOOIOOOOOOOOOOOODO seine 00.00.009.00000000 Worse ........ ....... .. HAND RESPONDENT INCOME CARD. 2?. Would you tell me which letter on this card best represents your 28. Race? 315 1/8/52 yearly income? A. Less than 1000 G. 6000~6999 B. 1000~1999 H. 7000—7999 c. 2000~29v9 I. 8000—8999 D. 3000—3999 i: i8?86 9:? er E. 4000—4999 F.'5000—5999 DON'T ASK QUESTION 28, BUT RhCORD: White.................. Negro.................. Other.................o 5' 27 “Pk 8017‘, P May 6 157 DIP 1i jugg‘ u ‘ I), . " 4 ' v o A ‘ ‘ .. . 1 u o - ‘ . ‘ n n . l v .V o ‘- 7 .' I c I , \ v \ l ‘ V: ‘ n - > — I n i . D l I'r- l' — ’ r | l l v . _ ‘ y 1 | . ‘ ' '. o r ”liltfinfifliflllfilflfljfilflifllfllfllfllfl“