Make your own free website on

Sociological Analysis 1972, 32, 170-179.

(With three post-publication addenda following the original article)

RELIGIOCENTRISM AND ETHNOCENTRISM: Catholic and Protestant in Australian Schools

John J. Ray

University of New South Wales


Dianne Doratis

University of Sydney

The assumption that Roman Catholics are more religiocentric is challenged. New attitude scales to measure religiocentrism and ethnocentrism were constructed. In a study of Australian fifth form students in two Catholic and two public schools, religiocentrism and ethnocentrism were shown to be uncorrelated with religious background. The implications of this for teachers and for social science are briefly explored.


"Ethnocentrism" is the social scientist's value-neutral term for ethnic or racial prejudice. It refers to ethnically-based sentiments of exclusiveness without any implication of their moral worth or justifiability. The term was first popularized by Adorno, Frenkel-Brunswik, Levinson and Sanford -- in their work The Authoritarian Personality -- although it was in fact originated by W. G. Sumner almost fifty years earlier.

By analogy, the term religiocentrism is derived here to mean religiously based sentiments of exclusiveness -- beliefs that one should marry within one's own religion, work with members of one's own religion, and in general prefer members of one's own religion above others. This will also entail ipso facto devaluative judgments of other religions. Ulster currently provides some extreme examples of religiocentrism.

In Australia it is a popular belief that Roman Catholics are more religiocentric than people of Protestant background. There are many tales of the discriminatory hiring and fining practices of Catholic-owned businesses, and the private Catholic school system certainly would seem to promote a religion-based exclusiveness. Universal Catholic doctrinal prohibitions on interfaith marriages also would seem to be all the evidence one needs of religiocentrism.

Using similar reasoning, it has been argued that Catholics are also more ethnocentric. The most notable example of ethnocentrism in recent times arose in Catholic Bavaria and was led by the Catholic Hitler. The notable Fascist dictatorships (as distinct from purely military dictatorships) have all arisen in Catholic countries -- Italy's Mussolini, Hungary's Admiral Horthy, Argentina's Peron, and Spain's Generalissimo Franco. In Australia, too, the historically strongest bastion of support for the white Australia policy was the Catholic based Australian Labour Party (Lipset, 1960).

All the considerations above give no place to the forces of liberalism in the Catholic church that have steadily wrought great changes in not only the practices but also in the teachings of the Church in recent times. The recent relaxation of the doctrinal prohibitions on interfaith marriages was brought about largely by pressures from within the Church. On the modern Australian political scene, too, the almost entirely Catholic Democratic Labour Party has been the most forthright of all parties in its support for Asian immigration. (See also Peak, 1945).

This liberalism is certainly no accident and is certainly not mere window-dressing. Because Catholics generally would seem to take their religion more seriously, the idealism of Catholics tends to be expressed in a Christian form -- and there is no element of Christian doctrine more prominent and less disputable than the adjuration to brotherly, love and Christian charity. It is understandable then that recent findings (Knopfelmacher & Armstrong, 1963; Beswick & Hills, 1969) have shown levels of ethnocentrism among Catholics that are no higher than the community average. As a logical extension, this paper asks whether religiocentrism, too, might be no higher among Catholics than among Protestants.

A second important purpose of this paper is to recheck the above findings. Knopfelmacher's ethnocentrism scale was of a most rudimentary nature and reliability statistics are not reported. His further claim that Catholics are less ethnocentric on the Bogardus social distance scale may also be explicable by there being a higher proportion of the key minority groups (e.g. Italians) among the Catholic sample. The Beswick and Hills study is also not without some major defects, stemming from the nature of the measure they used. Their scale is greatly deficient in content validity. On the one hand, it omits almost entirely mention of the one outgroup Australians are most likely to feel strongly about "New Australians" (meaning usually Italians and Greeks). On the other, it contains many items having no obvious reference to racial issues whatever, such as items about the desirability of "National Service" (conscription).

This then leads to what is believed to be the third contribution of the present study -- the construction of a new set of attitude scales for measuring the concepts being discussed. There appears, in fact, to be no scale previously existing which is designed to measure religiocentrism. The present paper then is something of a pioneering effort in this field -- with perhaps some of the defects that this must lead to. Even in respect to ethnocentrism there appears to be no scale available suitable for immediate Australian use. The one alternative to the Beswick and Hills scale is that by Hughes (1968). In some respects this scale is the opposite of that used by Beswick and Hills. Far from ignoring Southern European immigrants, it consists of five questions about each of three groups, Italians, Greeks and Jews. The difficulty is that these questions are by and large identical for the three groups. The only change is to insert "Greeks" where previously "Italians' appeared, for example. This is at once a conceptual and a measurement defect. It is a conceptual defect because the things objected to in Italians are not likely to be the same things that are objected to in Jews (and vice versa). It is a measurement defect because it creates spurious consistency in the answers given by respondents. They may answer two items the same way not as an outcome of real beliefs but as an outcome of the similarity of wording. Every test constructor knows how intrusive wording effects can be. For example, one factor analysis of Beswick and Hill's protoscale done by those authors yielded a third factor identified at first by them as "authoritarianism," but which in fact seemed to load highly any item which included the word "Australian," in whatever context that word was used. To get only "true" consistency and to avoid wording effects in general it was therefore deemed desirable to have a scale in which each item was clearly different from any other,


Constructing the Scales

For the norming study, the subjects were 154 first-year students at the University of Sydney -- 46 of Catholic origin and 108 Of Protestant background. Filling out the questionnaire was an option for gaining course marks and some respondents of agnostic or non-Christian background did appear. These were discarded in analysis.

Two protoscales to measure ethnocentrism and religiocentrism respectively were included in the questionnaire, together with Ray's religious conservatism scale. Final versions of the two prejudice scales were constructed by item analysis (item-to-total correlations: Guilford, 1954) and all subsequent results reported are based on the reduced scales.

The religiocentrism scale was constructed to meet the particular needs of this study wherein the differentiation attempted was between student Catholics and Protestants. This meant that items specifically referring to Catholics and Protestants had to be excluded if any opportunity for similarity of response (the null hypothesis) was to be given. Less obviously, reference to atheists and agnostics also had to be avoided. This was because atheism and agnosticism were described as being the present beliefs as opposed to belief backgrounds of some of the sample. The items finally tested, then, fell into two categories: (1) items referring non-specifically to "religions other than my own," and, (2) items referring to clear religious outgroups such as Hindus. During administration, subjects were read a preamble explaining that items referring to "my religion" should be taken generally, i.e., as referring to Catholic vs. Protestant vs. Atheist, vs. Buddhist, rather than as Baptist vs. Methodist, etc.

In this study it seems most important to control for strength of belief. If Catholics are by and large more accepting of church teachings, they might show religiocentric attitudes which stern from their belief system rather than from personality needs. It is of course no easy problem to sort out the two possible grounds for assent to an ethnocentrism item, but of the available possibilities it seemed most promising to get an independent estimate of belief for each subject by using Ray's religious conservatism scale (see Table 1). This scale contains items covering a range of belief from a most fundamentalist, "hellfire" kind, to beliefs that there is not much truth in any religious teachings. The "theological" content of the items was very low so that strong believers of either faith might be expected to get high scores (e.g., items about hellfire were used rather than items about the Pope). The reliability of this scale on its norming sample of students in a public boys high school had been .83, using Cronbach's (1951) coefficient "alpha." Note that the "alpha" statistic is also evidence of an important sort of validity (see Werts & Linn, 1970).

Using the Scales

The final forms of the scales were then applied to a sample of fifth-form students in four Sydney metropolitan schools. The classes were all of around fifty students each. There was one class of Catholic boys, one class. of Catholic girls, one public school class of boys and one public school class of girls. The Catholic schools were deliberately selected for their low proportion of students having migrant backgrounds. There were five such students in the public school sample and eight in the Catholic school sample.


The Religious Conservatism scale*

1. If God listens to prayers, He certainly doesn't do much about them. R
2. You cannot be psychologically complete without believing in God.
3. One's chances of life after death are not likely to be affected by what religious denomination one belongs to. R
4. The form of worship (music, communion, ritual, etc.) is unimportant.
5. Those people to whom "God has revealed Himself" have been subject to delusions. R
6. Every word of the Bible has been inspired by God.
7. I believe there is a physical Hell where men are punished after death for the sins of their lives.
8. I believe there is a supernatural being, the Devil who continually tries to lead men into sin.
9. The church has a duty to guide its believers in what they should read.
10. There are other religions in the world as good as Christianity. R
11. Fasting on certain days of the year is absurd. R
12. No church property should be exempted from taxation. R
13. Even in this century, miracles like those in Christ's day have happened.
14. No country should have an established church. R
15. A marriage that is not solemnized in a church is not complete.
16. To me the most important work of the church is the saving of souls.
17. I believe that there is a life after death.
18. I believe that there is a Divine plan and purpose for every living person and thing.
19. The only benefit one receives from prayer is psychological. R
20. The churches are out of date and could easily be done away with. R

Reliability coefficient = .83 on 154 first-year university students.

* Response options are from Strongly Agree (Scored 5) to Strongly Disagree (1). A mid-point of "not sure" (3) is allowed. Items marked R are reverse scored.


The Ethnocentrism scale

(scored as in Table 1).

1. A major fault of the Jews is their conceit, overbearing pride, arid their idea that they are a chosen race.
2. The white Australia Policy is a good policy because it helps to keep Australia white.
3. Greeks just don't know bow to reciprocate the friendliness of the Australian people.
4. It is wrong, to say New Australians, in general, are not to be trusted. R
5. Italians are no better or no worse than any other people. R
6. Asians should be allowed to migrate to Australia. R
7. If only New Australians weren't so sloppy and dirty, Australia would be a much nicer place in which to live.
8. Italians and Greeks care about no one but themselves.
9. Jews are as valuable, honest and public-spirited as any other group. R
10. Australians should not be too keen on friendship with the Japanese after the lessons of the last war.
11. I think the Chinese should be kept out of Australia.
12. Migrants tend to make the best of the opportunities available to
them. R
13. We must be very careful not to let too many Asians into this country or they'll take over the place.
14. Italians seem to have an aversion to plain hard work; They tend to be a parasitic element in society by finding easy, non-productive jobs.
15. Jews may have moral standards that they apply in their dealing with each other, but with Gentiles they are unscrupulous, ruthless and undependable.
16. I can hardly imagine myself marrying a Greek.
17. Migrants are as friendly as Australians. R
18. Italians should be more concerned with their personal appearance, and not be so smelly and dirty.
19. Southern European migrants have lowered the standard of living in Australia because they live like animals.
20. Migrants don't even try to speak English properly, and then blame others when they can't get by.
21. Chinese people have a respect for family and tradition which can be admired. R
22. Greek and Italian migrants have made no contributions whatsoever to the Australian way of life. .
23. If we are not careful, the New Australians will make our living conditions unbearable.
24. There is something different and strange about Jews; one never knows what they are thinking or planning, or what makes them tick.
25. Migrants try hard to make the best of life in Australia. R
26. Greeks are really no different to anyone else. R
27. The Australian culture and way of life is enriched by the many different people immigrating and living here. R
30. When he is given a fair chance, the Aborigine can live as decently as any white man. R
31. Visits to Australia by Japanese politicians and businessmen should be welcomed. R
32. We should be more selective in allowing people to settle in Australia, since most of the present migrants are just plain ignorant and irresponsible.
31. There is no reason to believe that innately the Jews are less honest and good than anyone else. R
32. I wouldn't like any member of my family to marry an Aborigine.

Reliability coefficient = .86 on 154 first-year university students.

In Australia, Catholics are roughly one quarter of the population and there is a well-developed and long-established Catholic school system which receives a certain amount of Federal government financial support. The Catholic schools are hence in competition with the state-run public school system and provide a thoroughly legitimate alternative to it at all levels from first grade to sixth form, Education in Australia is hence institutionally religiocentric. The minority of Catholics who send their children to state-run schools are usually regarded as being Catholic "in name only"

In the questionnaire used with the above-described subjects, there were also "face-sheet" questions asked concerning political preference, father's occupation, religious background and present religious belief.

Data Analysis

It was decided to present the relationships in the form of correlations rather than as mean scores. This is because of the ease of statistical control offered by partial correlation. A dummy "religion" variable was therefore created by scoring Catholic as 1 and Protestant as 2. The correlations between prejudice and religion could therefore, if necessary, be calculated with the influence of belief-strength partialled out. A positive correlation between religion and religiocentrism is then equivalent to saying that Protestants are more religiocentric. Students of atheist or non-Christian beliefs were removed from the sample. The final n was 134.


Tables 1 to 3 present the final forms of the three scales and their reliabilities on the norming sample. In view of the furore over acquiescent response set (Rorer, 1965) Some balancing was attempted but, in view of Block's finding that the characteristics of partially balanced and fully balanced scales are essentially indistinguishable, a perfect balance of positive and negative items was not seen as important.

In Table 4, the correlations of all variables measured in the school student sample are presented. Neither ethnocentrism nor religiocentrism show any relation to religious background. The perfect correlation between religious background and present belief indicates that no student had changed from Protestant to Catholic or vice versa. Note that atheists or agnostics had already been excluded on a priori grounds.

Since religious conservatism too was not related to religious background, it would not be a valid procedure to partial it out. In the university student sample also (used to norm the scales), the relation of religion to religiocentrism and ethnocentrism was insignificant (r = .442. and .097).


The religiocentrism scale

1. It is good for people of different religions to intermarry. R
2. I would prefer to go out with members of the opposite sex who are of my own religion.
3. I would not marry a person of another religion.
4. I often feel attracted to religions other than my own. R
5. The Buddhist religion is nothing more than a lot of superstition.
6. It is important to be tolerant of other people's religious beliefs. R
7. It makes no difference to me what religion my friends are. R
8. If I were an employer I would give people of my own religion first preference.
9. I think my religion is nearer to the truth than any other.
10. It would be good if everyone were of the same religion, preferably my own.
11. The Buddhists have very unusual religious ceremonies which have a strange perversity.
12. Having many different religions is good since it generates thought about and evaluation of your own religious beliefs. R
13. I wouldn't mind if someone in my family married a Buddhist
14. My religion is the one and only true religion.
15. I think that it's better if you stick to friends of the same religion as your own.
16. Seventh Day Adventists and Jehovah's Witnesses must be quite mad to believe some of the things they do believe.
17. There is not much point in trying to make friends with people whose religious beliefs differ from your own
18. The Moslems are so stupid and ignorant that you couldn't expect them to realize how wrong their religion is.
19. I could go out with a person of a different religion to my own, but I wouldn't like to marry them.
20. Hindus and Moslems are hyperaggressive and sexually overindulgent.
21. There aren't too many Moslems who really understand their religion.
22. I think a person's religion should not have any bearing on whether he gets a job or not R
23. I generally get on best with people of my own religion.
24. If the Hindus and Moslems did not spend so much time procreating, they could both ease the population explosion, and spend more time in educating themselves.
25. I would prefer to work for an employer of my own religion.
26. You can tell a person's religion just by the way he acts.
27. Most Moslems, Buddhists and Hindus are very stupid and ignorant.
28. The Moslems have so many followers because, in places like Turkey, the people are very simple and tend to believe anything.
29. One religion or philosophy is as good as any other. R
30. It is sometimes good to question your own beliefs and not to follow a religion blindly R
31. Most Moslems wouldn't lift a finger to help a person of a different religion to their own.
32. What religion a person is doesn't matter to me at all. R
33. Christianity is one of a number of equally interesting religions. R

Reliability coefficient = .88 on 154 first-year university students.


Intercorrelations from 154 Fifth form students aged 15 - 17*


Age....................................1.00...-.06...-.00......11.....11.....-.16......12...... -.02
Political preference.....................1.00.....16....-.06.....06........20...-.06.........03
Religious Origin.............................................1.00...1.00.....-.08......05........04
Religious Belief..............,.........................................1.00.....-.08......05........04
Religious conservatism.......................................................1.00....-.16........45

* Father's occupation is scored 1 for manual, 2 for non-manual (See Ray, I971).
Sex is scored 1 for male, 2 for female.
Religion is scored 1 for Catholic, 2 for Protestant.
Political preference is scored 5 ( D.L.P. ), 4 ( L.C P. ), 3 (no preference), 2 ( A.L.P. ) and 1 (Communist).
The significance level of r is .14 (p = < .05, two-tailed).


In spite of the measurement and sampling defects in earlier studies, the basic result has been confirmed. Catholics are no more ethnocentric than others. The finding that, as individuals, their attitudes are not more religiocentric is perhaps surprising in view of traditional church teaching but can be seen as a tendency to accept other people on personal rather than religious grounds. Australia is, after all, far from being another Ulster.

This study has been, of course, of expressed attitudes only. its implications for questions about the prevalence of religiocentric or ethnocentric practices are at best indirect.

The finding that Catholics are not more prejudiced towards other religions will no doubt be pleasing to social scientists of Catholic background or convictions (though neither of the present authors fall into this category). For teachers of religion in Catholic schools, however, the selfsame finding might well be seen as presenting a serious problem. What is happening to the teaching that Catholics should avoid marrying non-Catholics? In such respects the teacher would want the students to be exclusive or "religiocentric" -- but this is here found not to be the case. It is perhaps a slightly startling thought that the influence of the secular culture in which we all live may be so powerful as to override (in its effect on expressed attitudes) ten years of daily religious teaching.

This failure to influence the consciously acknowledged attitudes of the young would at the very least seem to augur ill for the actual obedience of these young people to church teachings when they reach the age of full maturity. The relationship between attitudes and behavior in this field (i.e. the predictive validity of attitude scales) is however certainly not a simple one (See Ray, 1971).

Other points in the results worthy of note are the scale reliabilities and the empirical correlates of the religiocentrism scale in particular.

All three scales were sufficiently reliable to warrant their further use with samples of this sort and they may hence be seen as filling a need for adequate instruments not met elsewhere. Further studies of religiocentrism can now proceed in a more objective and empirical way than was previously possible. It is, of course, easy to find fault with the logical structure and status of many of the items used. The point that should be kept in mind in this connection is that the item-selection was done on empirical grounds. The items that seem nonsensical to the well-informed academic are responded to quite consistently and happily by the subject population. This distinction between logical meaning and psychological impact is one quite well-canvassed by Christie, Havel and Seidenberg.

The significant correlation between religiocentrism and ethnocentrism does tell us that prejudice tends to generalize somewhat. This fits the view of prejudice as on occasions fulfilling deep-seated personality needs. The correlation is low enough however (when r = < .35, less than 15 per cent of the common variance has been explained) to enable us to reject the view that religiocentrism is merely the manifestation of an inadequate and generally prejudiced personality. It is also of some note that religiocentrism is not a phenomenon of the lower classes (r = -.03) nor is it stronger in females (r = --.06).

The correlation of religiocentrism and religious conservatism may be seen as concurrent validation for the religiocentrism scale. People who adhere closely to traditional religious teachings also tend to be religiously more exclusive.


Students in Australian Catholic secondary schools are neither more ethnocentric nor more religiocentric than their peers of Protestant background. Among other subject populations, therefore, any assertion of greater religiocentrism among Catholics should not be accepted unless direct evidence to that effect is provided. Ethnocentrism, religiocentrism and religious conservatism were all shown to be separate and distinct factors of attitudes in their own right. They are not just three aspects of the one thing. Religiocentric people do however tend to be both religiously conservative and ethnocentric. There is also a just-significant tendency for religiously conservative people to be anti-ethnocentric. Catholic students, however, were not more religiously conservative. It is inferred that even for students in Catholic schools, the influence of religious education is overwhelmed by the secular culture in matters of attitude formation.


Adorno,T.W., Frenkel-Brunswik, E., Levinson, D.J. & Sanford, R.N. (1950). The authoritarian personality New York: Harper.

Beswick, D.C. & Hills, M.D. (1969) An Australian ethnocentrism scale. Australian J. Psychology 21, 211-226.

Block, J. (1965) The challenge of response sets N.Y.: Appleton Century.

Christie, R., Havel, J. & Seidenberg, B.(1956) Is the 'F' scale irreversible? J. Abnorm. Soc. Psychology, 56, 141-158.

Cronbach, L.J. (1951) Coefficient alpha and the internal structure of tests. Psychometrika 16, 297-334.

Guilford, J.P. (1954) Psychometric methods. N.Y.: McGraw Hill.

Hughes, A.H. (1968) Problems and solutions in measuring psychological dispositions. Paper delivered at Australian UNESCO seminar on mathematics in the social sciences.

Knopfelmacher, F. & Armstrong, D.B. (1963) The relation between authoritarianism, ethnocentrism and religious denomination among Australian adolescents. Amer. Catholic Sociological Review 24, 99-114.

Lipset, S.M. (1960) Political man N.Y.: Doubleday.

Peak, H. (1945) Observations on the characteristics and distributions of German Nazis. Psychological Monographs 59, 1-42.

Ray, J.J. (1971) Ethnocentrism: Attitudes and behaviour. Australian Quarterly, 43, 89-97.

Rorer, L.G. (1965) The great response-style myth. Psychological Bulletin, 63, 129-156.

Werts, C.E. & Linn, R.L. (1970) Cautions in applying various procedures for determining the reliability and validity of multiple scales. American Sociological Review, 35, 757-759.


1). Other articles looking at the generalizability of prejudice are as listed below. They all show instances of different types of prejudice being only weakly correlated. The .33 correlation (only 11% shared variance) between two types of prejudice reported above is then fairly typical. The implication is that there is a tendency for some people to be wary of anybody at all who is not like themselves but most prejudice is specific to particular groups. So if you do not like blacks (for instance) it does not mean that you will automatically dislike Jews (for instance).

Ray, J.J. (1974) Are racists ethnocentric? Ch. 46 in Ray, J.J. Conservatism as heresy Sydney: A.N.Z. Book Co.

Ray, J.J. & Lovejoy, F.H. (1986). The generality of racial prejudice. Journal of Social Psychology, 126, 563-564.


2). The article above made the conventional assumption among psychologists to the effect that dislike of outgroups and liking for one's own group would be negatively related and thus form a syndrome of "ethnocentrism". In its usual pesky way, however, reality has proved unco-operative with what theorists thought was obvious. The references given below also show that liking for one's own group (patriotism) is in general totally unrelated to attitudes towards minorities. You can like (or dislike) BOTH your own group AND outgroups -- and many people do. The issue is discussed at greater length here

3). Replication is one of the cornerstones of science. A new research result will normally require replication by later researchers before the truth and accuracy of the observation concerned is generally accepted. If a result is to be replicated, however, careful specification of the original research procedure is important.

In questionnaire research it has been my observation that the results are fairly robust as to questionnaire format. It is the content of the question that matters rather than how the question is presented (But see here and here). It is nonetheless obviously desirable for an attempted replication to follow the original procedure as closely as possible so I have given here samples of how I presented my questionnaires in most of the research I did. On all occasions, respondents were asked to circle a number to indicate their response.

Go to Index page for this site

Go to John Ray's "Tongue Tied" blog (Backup here)
Go to John Ray's "Dissecting Leftism" blog (Backup here)
Go to John Ray's "Australian Politics" blog (Backup here)
Go to John Ray's "Gun Watch" blog (Backup here)
Go to John Ray's "Education Watch" blog (Backup here)
Go to John Ray's "Socialized Medicine" blog (Backup here)
Go to John Ray's "Political Correctness Watch" blog (Backup here)
Go to John Ray's "Greenie Watch" blog (Backup here)
Go to John Ray's "Leftists as Elitists" blog (Not now regularly updated)
Go to John Ray's "Marx & Engels in their own words" blog (Not now regularly updated)
Go to John Ray's "A scripture blog" (Not now regularly updated)
Go to John Ray's recipe blog (Not now regularly updated -- Backup here)

Go to John Ray's Main academic menu
Go to Menu of recent writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page (Backup here)
Go to Selected pictures from John Ray's blogs (Backup here)
Go to Another picture page (Best with broadband)