‘ I {v.13}: ‘9‘... a 3",; y"; .9. 333 33' -., purge“ e u»; 3., at: ' - .’ v ‘ '5' V ‘ a . ,_ ‘ 3-, - ,( tw“.\\f a t'oh.‘ Jflé‘sl‘? I‘m 0 l. 6 V .\‘ 6 A“:KQC ‘0‘ =1 .‘N4w' ”Vb? "1; ’4"-171<'3§31K" .3" b.- . Ur“! . 1.‘~. .quud.‘ .411 5 & “-! o O ' v r .. o ‘fl“fi I...“ 8' 0.2-; fifirr'g‘. xi t‘ “ o 0 Gina é '0 050‘- Jkk. U. Act. u"! y. . (I Q n. a .‘ A If I " W ‘ 49“ ‘J‘ fl‘ f.‘ .I o ‘1‘ 5.- H ’ ‘ 3 7: '.‘ \ .1 J ' " :— '3‘ “ o:\1'§ 00.";‘11 5-”.15-3- vvud— \u‘u / :. IL: 3 ‘ ‘ “I g. 5 f : .-:~:g.".-«;--'* ."‘~ ”1‘1; - \ on.-. \3 ~51 IL; ~91... Ab~¢o¥ooo - c-.. J . 3L \ L , I I I . . 4 . N '_ . I _ ~ | K ‘ u. '--J._.L_L.' f U‘ - ‘vv ‘R’ 1”, 4““ l - V‘ a “a ‘ ' . I ( n ‘. U . I I — \ ' t . . A3” _.’ I; h "Anemia and Author-1 tariani am : 3 .. ‘~“ «- a a r 1‘ _ . th r . F]. a; xv , .. ‘ § -;1 1 ; ~. : This is to certify that the ' 7 ..-. thesis entitled p -_ A Study in Reliability" 'f-‘H' } presented by F .1 T fxii .~ 14+ ‘11,: S ' .' ‘1' .43; “r “:2 I ; if?“ I’ i P ' Keith Gawain McKitrick I I .r ' . ‘-‘ ".&.:‘ ' . ’. 7" f]. ’9. ’-' .‘5'3. has been accepted towards fulfillment .‘E‘Z-‘WT "1.1.: “f i" (a . If L. ,-'.;. of the requirements for gt 1.- o 5, ‘ .- H l" -4 °g.;$.- . . c» ,v .3 h . .' "a \ . ~. n3: ‘. 1" I I “t I | -, ‘l .‘u .., I” § .- , u ‘v I: ei‘ii.ufi ' ' d‘ Z!!- M" 'f ’7‘ .1. l _ ' 9 Major professor Ii. 9; , WU; ‘ Date m £6L 1952 '1‘», a, 7‘ .1 7.3.? T . . Q ,.. ..‘ ' ii I;’ -—& \I' I "E‘I‘l’l --_.-- 3...--- v‘. If] I II-.- ANOMIE AND AUTHORITARIANISM: A STUDY IN RELIABILITY By KEITH GAWAIN MCKITRICK 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 THESiS ACKNOWLEDGMENT Grateful acknowledgment is made to Dr. Milton Rokeach for his inspiration and patient assistance throughout the course of this research. ‘4‘}.4‘ \ " ’ 4- II. III. IV. V. VI. VII. TABLE OF CON‘ENTS Acknowledgment List of tables INTRODUCTION STATEMENT OF THE PROBLEM HYPOTHESES AND METHODOLOGY PROCEDURE PRESENTATION OF THE DATA DISCUSSION OF THE RESULTS SUMMARX AND CONCLUSIONS Bibliography Appendix iii 13 17 26 32 33 ii II III IV VI LIST OF TABLES Comparison of Means and Percentages Between Sub-Samples AF and A—F— Percents of all "Agree" Responses to Both Forms of A and F Items Comparisons of Percents of "Agree" Responses to Both Forms of the A and F Items Between the Education Groups Comparisons of Percents of "Agree" Responses to Both Forms of the A and F Items Between the Income Groups Comparison of Within-Sample Total "Agree" Responses on all Scales; Education and Income Groups Correlations of each item with total score; A11 Scales iii Page 16 18 19 20 21 22 ANOMIE AND AUTHORITARIANISM: A STUDY IN RELIABILITY INTRODUCTION The term "anomie" was first used in connection with social theory by the French sociologist Emile Durkheim in his book, De la division du_t§§vail social (3). Simpson, in his trans- lation of this volume, has reinstated the old English word "anomy," obsolete since 1755 and first used in 1591. It was employed in the earlier period to mean "disregard of (divine) ' and later "lawlessness." Its derivation is from the law,‘ Greek. When Durkheim wrote about this concept, he had in mind primarily the disintegrated and demoralized state of a society in which there are no common values or morals that effectively govern the conduct of individuals. Others, according to their points of view have described this situation in different ways. Theologians would see a breakdown of faith in divine guidance; and political scientists would lay the blame on an ineffective governmental system. Thus the concept has definite collective connotations---those of an abnormal state of society. Merton (6), characterizes the instability of a society leading to anomie as "a symptom of dissociation between culturally pre— scribed aspirations and socially structured avenues for real— izing these aspirations." He emphasizes the intense preoccu- pation with attainment of wealth to the disregard of legitimate avenues for achieving this end. This would appear to be closely related to the "lawlessness" of which Simpson spoke. Where psychology becomes concerned with the concept of anomie more directly is upon the latter's manifestation in the behavior of the individual. What are the symptoms of v anomie? Durkheim spoke of the hopelessness brought on by the economic factors of the division of labor and of fluct- uations in the business cycle. In Lg Suicide he described the intolerable sense of futility culminating in "anomie suicide." DeGrazia, in his very readable book (2), combines a social—political orientation with a psychoanalytic descrip— tion of anomie. In his protest against the harsh competitive nature of our society, he brings in pertinent material of Horney in which she speaks of the "unendurable emotional isolation stemming from the hostility of competitiveness." Jung is quoted with regard to the loss of meaning of life in general; the reader is referred to Fromm and Alexander where they speak of the "lack of purposefulness and the uncertain— ties of goals." DeGrazia aptly describes the "bewilderment and dulling anxiety" tormenting the Jobhunter on the trek in search of a Job which he knows is not there. While similar syndromes have been noted regularly in clinical and social psychology, the first time that "anomie" had been put under systematic scrutiny and measurement was in 1951. At that time sociologist Leo Srole (8), attempted, by the use of attitude scales, to relate anomie to certain socio-psychological variables, specifically prejudice and authoritarianism. Before examining this latter study and its implications for the present research, this theoretical introduction will be rounded out with a brief account of the concept of author- itarianism and the scale which has been designed to measure it. In their search for a coherent pattern of deep-lying personality trends entering the makeup of the potentially fascistic individual, Adorno et al. (I), documented results from their California Study which are considered classic in present-day social psychology. Among the findings that emerged from this research was the demonstration that an individual is likely to maintain a basic similarity of approach and out- look in a great many areas of life. A distinct personality syndrome was isolated which exemplified this generality of approach, and was described as the Authoritarian Personality. Persons of this type were characterized as rigid, conventional, power-oriented and punitive, with a repressive denial of weakness in themselves and a disdain of it in others. Their conceptions of sex roles and moral values are dichotomized, outgroups are stereotyped, and minority-group prejudice with self-righteousness is evident. It was found that harsh, ex- ploitive parent-child relationships were closely associated with this prejudice in later life. Earlier in the research, after selecting prejudiced in- dividuals by means of attitude questionnaires, clinical interviews, and projective tests, the investigators became concerned with.the idea of constructing a scale that would measure prejudice without appearing to have this aim and without mentioning Specific minority groups. The second and more central purpose was that of constructing a scale which "might yield a valid estimate of antidemocratic tendencies at the personality level." After extensive revision and statist- ical refinement, a highly reliable scale was formed (split— half reliability averaged .90) which contained 34 items and was called the F (fascist) Scale. Srole, in contrast to the California researchers, believed that the formations of patterns of prejudice could better be accounted for through the impact of socio-situational forces upon the individual resulting in a state of anomie. Here is seen a sharp divergence from the psychoanalytically oriented view which holds, rather, that the potential for prejudice lies deeply imbedded in the personality structure, conditioned by early family relationships. Srole's first step in gathering support for his hypothesis was in the construction of a scale for measuring anomie. This scale consisted of five items, each "on theoretical grounds" representing a component of this state of social dysfunction. Into a questionnaire form he included this anomie scale, a five-item prejudice inventory, and a five- item version of the California F Scale. Following the admin- istration of this combined questionnaire to 401 bus riders in Springfield Massachusetts, some remarkable results were ob- tained. By means of these short scales to measure each of the three variables involved, his findings apparently showed a closer relationship existing between anomie and prejudice than between authoritarianism and prejudice. While author- itarianism was still related to prejudice, it apparently was not independent of the sociological factors presumed to have been measured by the anomie scale. These relationships seemed to be in evidence most markedly within his low-education group. STATEMENT OF THE PROBLEM The urgency of reaching a solution to the problem of reducing intergroup prejudice is clearly recognized. An adequate understanding of the problem of prejudice demands a sound theoretical representation in terms of the dynamics of its origin. The California finding with respect to the relationship between authoritarianism and prejudice is gen- erally regarded as a significant theoretical contribution to this understanding. Divergent evidence such as the findings of Srole must, therefore, be examined critically in the int- erest of: (a) The eXposition of faults in, and consequent revision of, the existing theory; or, (b) exposing faults in the methods of acquiring the divergent evidence. HYPOTHESES AND METHODOLOGY Within the context of the above general problem, it is the purpose of this thesis to investigate the reliabilities of Srole's two measures, his five-item anomie (A) scale and the five-item authoritarianism (F) scale. In addition, two further measures of A and F will be examined with respect to their relative reliabilities. These two different scales, used for the purpose of this present study, will be described below. Working in conjunction with Roberts (7), an experiment was designed in order to test the following general hypothesis regarding the high relationship found by Srole between anomie and prejudice in the low-education group: The low-status respondents will tend to agree with the anomie statements (which are worded in such a manner that agreement indicates a state of anomie) in order to avoid the threat which might occur upon resistance or disagreement with the interviewer. In view of the relatively insecure position of these respondents, the interview situation may thereby generate a state of subordinate—conformity. If such conformity were operative in Srole's sample, it can be seen what a gross distortion of results could accrue in the low—education group if their reaponses were influenced by this factor. It was necessary of course, to hypothesize that a similar effect would occur with the F statements, which are also worded in such a manner that agreement in their case indicates authoritarianism. Accordingly, alternate A and F scales were devised where- by each of the statements were reversed in meaning so that agreement with them would then logically reflect a lack of anomie or authoritarianism, respectively. (Agreement former- ly assumed the presence of the attribute in question.) These were called the A- and F- scales. Thus if the statements were still responded to in the affirmative, it can be assumed that such "agrees" were given for the sake of acquiescence or conformity, and not in the interest of the actual ideational content of the items. This rationale must rest on the assump- tion that each of the changed statements carries a connotation which is exactly, or at least approximately, opposite from that of its original counterpart. The scales in each case would then be negatively equivalent. (A list of each item next to its changed version is found on Page 11.) There are two aspects of this experiment which merit de- tailed consideration. One has to do with the validity of the two measures, and this forms the main subject of Roberts' (7) report. The second aspect, that of the reliability of these measures, is contained in this present thesis. In an effort to determine the presence of conformity, and in turn the effect that it might have upon the reliabilities of the scores obtained from the various scales, the following four hypotheses are offered, along with the means for testing each: 1. Conformity in the face-to-face interview situation will bias the responses toward excess agreement on both forms of the A and F scales, original and reversed. This will appear throughout the total sample, effecting a reduction in the reliabilities of all A and F scores. (a) This rests upon the assumption of equivalent forms, (negative equivalence) and identity of sub-samples. (b) The test: A determination will be made of the per- cents of the AF group responding to each A and F item with "agree." Also, the percents of the A-F- group res- ponding to each A- and F- item with "agree" will be found. Excess agreement on both forms of each item will be noted. Conformity will Operate in the low-education groups to a greater extent than in the high-education groups. (a) The test: Similar comparisons (as in Hypothesis 1) will be made between the AF and the A-F- low-eduggtign groups. Likewise, the conformity responses of the high- education groups will be noted. The relative degrees of conformity will then be compared. Closely related to (2) above, an analysis of conformity reaponses of the low and high income groups should paral- lel closely that of the education groups. 10 (a) The same considerations as are listed in (2) with respect to methods will apply. As a consequence of conformity-determined reaponses, the A and F scales are not internally consistent to a significant degree. (a) Separate items will be correlated with total score. A... ll Anomie items: Original (A), and reversed (A-). There's little use writing to public officials, because often they aren't really interested in the problems of the average man. Writing to public officials often does some good, because many public officials are really interested in the problems of the average man. Nowadays a person has to live pretty much for today and let tomorrow take care of itself. It pays for a person to sacrifice today in order to provide for tomorrow. In spite of what some people say, the lot of the average man is getting worse, not better. In spite of what some people say, the lot of the average man is getting better, not worse. These days a person doesn't really know whom he can count on. In times of trouble, a person can usually find one or more people he can count on. It's hardly fair to bring children into the world with the way things look for the future. In spite of the way things look for the future, it's hardly fair to stop having children. 12 Authoritarianism items: Original (F), and reversed (F-). .11 There are two kinds of people in the world: the weak and the strong. It is a mistake to think that there are two kinds of people in the world, the weak and the strong. The most important thing to teach children is absolute obedience to their parents. To teach children absolute obedience to their parents is not the most important thing. Prison is too good for sex criminals; they should be publicly whipped or worse. Although some peeple 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 strict with people under him in order to gain their reapect. Any good leader should be lenient with peOple under him in order to gain their respect. No decent man can respect a woman who has had sex relations before marriage. A man can respect a woman even though she has had sex relations before marriage. 13 PROCEDURE After constructing the alternate items, the four scales (A,A-,F,F-) were arranged in questionnaires as follows: In the first form which was called the AF Form, was included Srole's five—item A Scale, his five-item F scale, and one of the versions of the E Scale used in the California research. This latter was Form 45, employed to measure prejudice or eth- nocentrism, hence the letter E. (1, p.128) The second form of the questionnaire contained the semantically reversed counterparts of the A and F items, plus the same E scale. This form was called the A-F- Form. (One of the complete quest- ionnaire schedules is included in appendix A.) The rationale for inclusion of the prejudice scale is in line with the attempt to duplicate Srole's study with respect to the questioned inter- relationships among the three variables of anomie, authoritar- ianism, and prejudice. This aspect is described in detail by Roberts, where his work is concerned directly with correlations among these three factors. Each of the two forms of the 20-item questionnaire was orally administered to a portion of a stratified sample of 158 residents of Lansing, Michigan. The stratification was accomplished by choosing approx- imately half of the respondents from modest-appearing dwellings; the other half was chosen from objectively more prosperous sections of the city. In this way a preponderance of high 14 and low—status individuals was sought for, rather than an approximation to the general population. The interviewing was conducted on a door-to-door basis, and the residents were told in effect that an opinion survey was being performed by Michigan State College, of which the interviewer was a representative. It was stressed to each resident that his own opinions with respect to certain present- day social issues were considered important to the survey, but that strict anonymity would be preserved. Surprisingly few refusals occurred. After initial rapport was established, the questioner and respondent seated themselves in the latter's living room where the interview was begun. The respondent's age, marital status, and years of schooling were obtained first. The statements of the questionnaire were then read aloud to him with three possibilities of response availible, "agree," "disagree," and "can't decide." After this was completed, each individual was questioned with respect to his general outlook or mood of the day, thus, "Would you say that you are feeling better than usual today, worse than usual, or about the same as usual?" (It is possible that the pessimistic connotations of the anomie statements could be related to momentary mood.) Finally, each interviewee was asked to check the salary bracket located at the end of the schedule which represented his own income. The sub-sample to which the AF forms were given totalled 86; the A-F- forms were administered to 72 persons. Between 15 the two latter groups, there were no significant differences in means, standard deviations, or percentages with regard to the following variables: income, education, age, sex, marital status, and score on the E Scale. These data are shown in Table I; they provide the basis for assuming homogeneity be- tween the two sub-samples. Vibe-r. TABLE I COMPARISON OF MEANS AND PERCENTAGES BETWEEN SUBhSAMPLES AF AND A-F- AF A-F- Mean t Income 4900 4790 .25 Education 11.47 10.47 below 1.96 Age 44.45 42.15 1.00 Ethnocentrism 21.16 22.31 .43 Percent Sex M 35% M 44% 1.15 F 65% F 56% Marital Status M 83% M 90% 1.56 16 l7 PRESENTATION OF THE DATA It will be recalled that an affirmative response to the original A item is assumed to denote the presence of anomie. Disagreement with the same statement in its reversed form should likewise imply the presence of anomie; the same assump- tions hold for responses on the F and F- scales. To repeat, if it can be shown that the makeup of the AF sample is homogeneous with that of the A-F- sample, and if the contents of the original and reversed items prove to be equivalent, ("agree" on AF = "disagree" on A-F-) is it not evidence for conformity if there occurs an excess of "agree" on bgth items? In Table II, immediately following, percents of "agree" responses to each item are matched with those of "agree" on the same item which was altered so as to have the cpposite meaning. The percentages were in each case derived from the appropriate sub-samples, AF, 85; A-F-, 71. (One subject was removed from each sample so as to effect a comparable distri- bution in terms of education.) Any given p does not imply an exact q; (proportion of disagrees) "undecided" reSponses accounted for an additional percentage, the average of all of which was 7.45. The tabular data relevant to the three additional hypoth- eses are found immediately following Table II. 18 TABLE II PERCENTS OF ALL "AGREE" RESPONSES TO BOTH FORMS OF A AND F ITEMS A A- F F— lgem No- 1. 32.1 88.6 79.8 37.1 2. 47.6 81.4 65.5 47.1 3. 27.4 72.9 44.0 57.1 4. 70.2** 97.1** 57.1** 61.4** 5. 15.5 78.6 41.7 65.5 **Excess agreement, both forms of same item COMPARISONS OF PERCENTS OF "AG TABLE III BOTH FORMS OF THE A AND F Low education .H “a" 19 u RESPONSES TO THE EDUCATION GROUPS ITEMS BETWEEN High education Item No. A A‘ A A‘ 1. 44.4 85.7 17.9 91.4 2. 55.5* 82.9* 38.5 80.0 3. 31.1 62.9 3.1 82.9 4. 75.5* 97.1* 64.1* 97.1* 5. 22.2 71.4 7.7 85.7 Low education High education Item No. F F‘ F F- _ 1. 84.4 40.0 74.4 34.3 2. 80.0 31.4 48.7 62.9 3. 53.3* 54.3* 33.3 50-0 4. 57.8* 68.6* 56.4* ' 54,3. 5. 46.7 68.6 35.9 62.9 *Excess agreement, both forms of the same item 20 TABLE IV COMPARISONS OF PERCENTS OF "AGREE" RESPONSES TO BOTH FORMS OF THE A AND F ITEMS BETWEEN THE INCOME GROUPS Low income High income Item No. A A' A A- 1. 35.1 94.3 25.0 82.4 2. 56.8* 80.0* 37.5 85.3 3. 32.4 65.7 25.0 70.6 4. 75.7% 97.1* 57.5“ 97.1* 5. 18.9 77.1 7.5 79.4 Low income High income _I_t_e_m NoL F F- F F- 1. 86.5 42.9 72.5 35.3 2. 83.8 34.3 47.5 55.9 3. 51.4 45.7 40.0 67.6 4. 54.1* 71.4 55.0* 58.8* 5. 45.9 62.9 32.5 67.6 *Excess agreement, both forms of the same item TABLE V 21 COMPARISON OF WITHIN-SAMPLE TOTAL "AGREE" RESPONSES, ALL SCALES; EDUCATION AND INCOME GROUPS FORM AF Low education High education t A scale 45.8% 30.3% 3.3 F scale 64.4 49.7 2.3 Low income .High income A scale 43.8 30.5 2.7 F scale 64.3 49.5 2.9 FORM A-F- Low education High education t A- scale 80.0% 87.4% not signif F- scale 52.6 54.9 not signif Low income High income A— scale 82.8 82.9 not signif F- scale 51.4 57.1 not signif Subgroup n's AF Low education 45 High education 39 Low income 37 High income 40 A-F- 35 35 35 34 TABLE VI CORRELATIONS or EACH ITEM WITH TOTAL scone; ALL SCALES Item No. A scale Phi F scale Phi 1. .467 .356 2 .709 .573 3 .379 .657 4 .468 .250 5 . 408 . 500 N = 60; minimum significance level: .333 Item No. A- scale Phi F- scale .185 .250 .147 .102 .102 .185 .102 .102 .218 .185 minimum significance .364 23 It can be seen from Table II that, with the exception of specific items No. 4 A,A-; 4 F,F-, there is no evidence of excess agreement on both forms of the anomie and authoritarian- ism items. It can be concluded then, that conformity has not been demonstrated throughout the total sample. (Hypothesis 1) Hypothesis (2), that conformity would be more pronounced in the low-education groups, called for a delineation of these groups within each sub-sample. Low education was defined as eleventh grade schooling and below; the high-education group consisted of those who had completed high school or above. Table III shows that, as education is broken down thus, a greater suggestion of conformity appears in the data. There are now 339 items which show excess agreement on both forms in the low-education groups; compared to this, there is but one item showing this trend in the high groups. A within-sample tabulation of total "agree" responses (Table V) shows significantly greater agreement per se in the low-education group as compared with the high group on both A and F scales. This is not found to be true in the A- and F- scales. Hypothesis (2) then, is not confirmed. Hypothesis (3) deals with the same considerations with respect to income. With low income defined as below #4000 per year, and high income as above this figure, the results are shown in Table IV. Here, as with education, there are suggestions of greater 24 conformity in the low as compared with the high—income groups. However, this appears only on the A scales; the F results are merely similar to those of the overall sample. According to Table V, there is again significantly greater within-sample agreement with the original A and F items in the low as compared with the high-income group. As with education, there was not found to be greater agreement in the case of the low-income group on the A-F- Form. Hypothesis (3) is thereby not substantiated. Hypothesis (4), that the A scale and the F scale are not internally consistent to a significant degree, was tested by means of item-test correlation. The A- and F- scales were examined similarly. As such instruments become shorter, each item makes an increasingly large contribution to the total score. With scales containing as few as five items, this variance becomes extremely great. Thus the total score for each individual in this case had to be re-computed for every item in terms of the remaining four. This was done in order that each one would be correlated with a score which is independent of the contribution made by that item. As is customary with this procedure, high and low criterion groups were established; in this case, groups had to be redefined for every item of the scales. The upper and lower 35 percent of scale scores made up these groups. Phi coefficients for each item of the A, F, 25 A-, and F- scales are shown in Table VI. All items of Srole's A and F scales were found to be sig- nificantly correlated with the total score of the remaining four items, with the exception of F No.4. In contrast, all items of the A- and F- scales fell below the minimum signif- icance level for N's Of this size. The data thus show all four hypotheses to be unsubstant- iated. 26 DISCUSSION OF THE RESULTS Before integrating these data into the final picture, it will be recalled that the first three hypotheses had to do with the attempt to show that the influence of conformity may have played a role in determining Srole's results. As his questionnaires were orally-administered, in contrast to those Of the California research where the respondent had no personal contact with the experimenter, a new interaction influence may have arisen. First, the parallel Operation of conformity throughout the total Lansing sample was attempted to be shown. Secondly, its primary locus was hypothesized to be found in the lower status groups' responses, those of the low—education and low- income groups. Such conformity, if operative, would in turn have a direct bearing upon the reliabilities of the anomie and authoritarianism scores. Hypothesis (4) made explicit this possibility by asserting that the A and F measures are not internally consistent to a significant degree. Taking the data again step by step, Table II shows that in the case Of the A Scale, (in conjunction with the A- Scale) there is only one item which shows evidence of excess agree- ment. Similarly, item No.4 in the F and F— scales gives the appearance of the conformity influence having been Operative. But four out of five items in each scale are unaffected; thus the first hypothesis is not substantiated. 27 Within the status groups there are suggestions Of greater, conformity among the low-education and low-income groups as compared with the high groups. However, here again there is no reason to suspect a significant trend, especially in the light Of the following: As will be recalled, the assumption upon which the rationale of reversing the statements rested, was that of negative equivalence to the originalsl. The confirmation of this was contingent upon the correlations Of the A- and F- scales with the E scale. If a negative correl- ation was found between A- and E, and between F— and E similar in magnitude to the positive correlations between A and E, F and E, this would be prima-facie evidence for the semantic- ally opposite equivalence of the scales. Roberts (7), reports near-zero correlations between A- and E, and between F- and E. Table V gives further evidence against an hypothesis of conformity. Here it is shown that the low-status groups agreed in significantly greater proportions with the statements of the original A and F scales than did the high-status groups within the AF sub-sample. These findings are in line with 1The question may be raised as to what methodological precautions were undertaken to assure that this requirement was met prior to experimentation. In reply, it must be recalled that the original items have been assumed to measure certain attitudes. If the statements are made the Opposite in meaning, then it should follow logically that the attitudinal Opposite should be forthcoming upon such reversal. Lacking, of course, is a prior empirical con- firmation of oppositeness. 28 the theory of anomie (8, p.9) and with the data of Adorno et al., (1, p.267) and could hardly be said to reflect con- formity. The same trend in reverse is shown with the A- and F- scales, though not to a significant degree. Hypothesis (4) stated that, as a consequence Of conformity— determined responses, the scores on the A Scale and the F Scale are not internally consistent to a significant degree. Sig- nificant Phi coefficients in nine Of the ten items point clearly to the lack of support for this final hypothesis. Inasmuch as these five-item A and F scales have attained such sizeable correlations with a prejudice (E) scale, (.55 and .64, respectively)1 thus showing considerable validity, it would seem that they are relatively reliable enough for this purpose. Indeed, the potentialities of these measures for even greater correlation with prejudice might be realized upon fur- ther refinement, especially in the case of the F Scale where one item made but an insignificant contribution tO total score. As to the relative reliabilities Of the A— and F- measures, Table VI shows that in every case the items making up each scale are correlated positively but insignificantly with total score. This fact, and in view of the slight (statist- ically not significant) reverse trend of agreement shown in 10f. Roberts (7) 29 Table V with both the A— and the F- scales, it is concluded that these measures may be reflecting a trend of Opposite attitude, but to an unreliable extent. They must undergo further refinement as to content in order to function as valid complementary measures of anomie and authoritarianism. 30 SUMMARY AND CONCLUSIONS A 20-item questionnaire including Leo Srole's five-item Anomie (A) Scale, his five-item version Of the California F Scale, and a lO-item prejudice (E) scale was administered to 86 residents Of Lansing, Michigan. A second form Of this questionnaire with each of its A and F items reversed in meaning, was given to 72 additional residents. This item- alteration was done in an effort to test the hypothesis that Srole's finding of a high A-E relationship among low-status respondents was a function of interview-conformity rather than Of the sociological factors thought to have been measured by his A Scale. This latter form of the scales was used in order to provide an assessment of conformity in the attempt to show that respondents might also agree with the opposite of the original statements. The aims of the present thesis centered upon the effect that such conformity might have upon the reliabilities Of Srole's A and F measures. The data gave no indication that conformity had been in Operation, either throughout the sample as a whole, or dispro- portionately within the low-status groups. The internal-con— sistency reliabilities Of all but one item score of both the A and the F scales were beyond the significance level. It was concluded, in the light Of these results, that Srole's A and F measures were adequately reliable for the 31 purposes made of them. In addition, upon the finding of insignificant internal- consistency correlations in the case of the experimental reversed scales, it was necessary to conclude that their use in the measurement Of anomie and authoritarianism is not yet warranted at this stage. 32 BIBLIOGRAPHY References . Adorno, Frenkel-Brunswik, Levinson, and Sanford, The Authoritarian Personality. New York: Harper & Brothers, 1950. DeGrazia, Sebastion, The Political Community: A Stpdy Of Anomie. Chicago: University Of Chicago Press, 1948 Durkheim, Emile, De la division du travail social. Translated by George Simpson. Glencoe: The Free Press, 1933. Durkheim, Emile, Le Suicide. Translated by Spaulding and Simpson. Glencoe: The Free Press, 1951. Guilford, J.P., Fundamental Statistigp in Psycholggy and Education. New York: McGraw-Hill, 1950. Merton, Robert K., Social Theory_and Social Structure. Glencoe: the Free Press, 1949. Roberts, Alan H., "Anomie and authoritarianism: A study in validity." Unpublished M.A, thesis, Michigan State College, 1952. Srole, Leo, "Social dysfunction, personality and social distance attitudes." Paper read before the American Sociological Society, 1951 National Meeting, Chicago. 33 MICL-IG‘A1V oooeeeoeoeeoogeoeocee STATE APPENDIX Case No. COLLEGE " Date (Form AF) Block No. parents OPINION SURVEY - LP) DON‘T ASK QUESTION ONE, BUT RECORD: 1. Is the respondent male or female? Male -.~.- -— Female BEGIN HERE INTERVIEWER: PUT YOUR NAME IN TH; SPACE BELOW. 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 peeple are thinking nowadays. AFTER GETTING SETTLED FOR TLL ILTJRVLJ., SAY: 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. You remember that I said that weaare not interested in your name, but for statistical purposes we'd like to have a little information about yourself. For example: 2. “hat iS yOur date Cf birth? .0000oncoo-000.000.000.0000.0.0.0o 3. What is your marital status? Married .................. Single .oooeeao-QQOOOOOOOO other ..OIOOIOOODOOOOOO.'. DO you remember the name of the last school you went to? DC NOT RECORD ANSWER. 4. What was the last year you completed in school? CIRCLE ON». 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 5. What church were you brought up in? (IF "PROTESTANT", What denomination?) 0-0... eeeee e oooooooooooooooooooooooo can 0 ~eooooooeIOOOI..eeeee , I will read you each of the statements and you tell me whether you agree, disagree, or can't decide. ..... 6. There's little use writing to public officials because often they aren't 34' . . o .’ LanSing Opinion Survey~\P) ~2— .0... 7. .0... 8. IO... 9. 0.00010. ....Olli 0.00012. coo-.13- .....lL". 0.00.15. 0.00.16. 0.00.1'70 O oooooloo 9.0.01"). .....20. .....210 .....22. 000.9230 000.024. coo-025. 030.026. There are two kinds of people in the world: the weak and the strong. Zootsuiters prove that when people of their type have too luCh money and freedom, they just take advantage and cause trouble. Nowadays a person has to live pretty much for today and let tomorrow tnke care of itself. 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. The most impggtapt thing to teach children is absolute obedience to their parents. One trouble with Jewish businessmen is that they stick together and pre~ vent other people from having'a fair chance in competition. It would be a mistake ever to have Negroes for foreman and leaders over WhitCSo In spite of what some people say, the lot of the average man is getting worse, not better. Prison is too good for sex criminals. They should be publicly whipped or worse. Any good leader should be strict with people under him in order to gain their respect. I can hardly imagine m self marrying a Jew. If Negroes live poorly, it's mainly because they are naturally lazy, ignorant and without self control. These days a person doesn't really know whom he can count on. No decent man can respect a woman who has had sex relations before marriage. There may be exceptions, but in general, Jews are pretty much alik . It's hardly fair to bring children into the world with the way things look for the future. The trouble with letting Jews into a nice neighborhood is that they gradu— ally give it a typical JeWish atmosphere. 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. Some days a person feels pretty good; other days he feels pretty bad about things in general. Yould you say that you‘re feeling better than usu. , the same as usual or worse than usual today? . Better........... Same .... ..... ... \ 35 HAND RJSPCNDSJT INCOMJ CARD. 27., Would you tell me which letter on this card best represents your yearly income? A. Less than 1000 G. 6000-6999 B. lOOO~l999 H. 7000—7999 C. 2000~2999 I. 8000-8999 D. 3000-3999 J. 9COC-99? J. 4C00~4999 K. 10,000 and over F. 5000-599) DCN'T ASK QUHSTION 28, BUT R CORD: 28. Race? White................. Negro ....... ......... otherOOOOOIOOOIOOO-OOOC 316 1/8/52 - --- V 1' $7 " ' '\ t f g) ..., I! I \ L a" 4'? 'fi [ ' ‘ O L MN " I..' J \ ‘5 .- - ‘ .. ~ ' .I .Q“~‘l T ' n . : I. > I .\ -. ' . r 0 ‘. L . ,5 L. 7' L: .l I V - I ..'.w- 7'7”"? 4 o-* HT W ‘~ . . 1 . l . 1),. -\r' I . ~.r. . l ' 7 l I . .1 ‘ i I I . ‘ I '. V 1 yo- ‘1‘ . . . I I i , l I . I . 5V \ I‘t . , | I I ,4 . ) . 'l .‘ 4 ‘_."IV .{u‘ ‘a . 5' ..- ——'~ . r * .‘ ... ... v ....\N. u . . . ..’o ..u. u. 4‘... .r flow... A. Mo. a “maker 1.2% IHIIIHHIIHHHIHIH 145 5607 [I II II‘ II 1" I I II I ’i l 0'! I llll'l' ' I ll I 1293 03 rlmnmlml