There are many different questions which may be asked concerning the number and nature of socio-political and social attitude dimensions. Some of the major parameters of these questions concern such things as whether interest is directed towards narrow and specific dimensions of ideology such as anti-semitism or whether one is interested in broad, generalized and higher-order dimensions such as 'Liberalism vs. Conservatism'.

Another major parameter of these questions concerns the specific domain of attitudes. Sidanius, Ekehammar & Lukowsky (1983) have delineated two major and somewhat overlapping domains of socio-political ideology. One is referred to as 'issue-oriented' political attitudes and primarily concerns political 'issues of the day' that are debated in parliament and editorial pages such as whether or not to increase or decrease a particular sales- tax or whether a hydro-electric plant should be built at a particular location. The second major domain of socio-political attitudes is basically concerned with those attitudes which interact with personality needs and cognitive functioning. Examples of socio-political attitudes within this second domain would be dimensions like ethnic and racial prejudice, authoritarianism, punitiveness, etc. For the sake of convenience, we will refer to dimensions in this second domain as 'social attitudes'.

Resumé

Differences in social attitudes between youthful supporters of different Swedish political parties were studied within a cluster sample of 783 Swedish high school students. Analyses of variance disclosed that, overall, respondents with different political party preferences differed with respect to social attitude dimensions: Political-Economic Conservatism, Social Inequality, Punitiveness, Militarism, Pro-western Sentiments, Racism, Xenophobia and Religion in that order. Multivariate, logistic regression analysis disclosed, among other things, that (1) the issue of political-economic conservatism (i.e. 'left-right') was a highly salient one for all party supporters with the exception of fascists, and (2) with the exception of Political-Economic Conservatism, different combinations of social attitudes were shown to discriminate any given party from all other parties. The theoretical implications of the findings are discussed.

Whereas 'issue-oriented' or strictly 'political' attitudes have been

of

primary concern to political scientists, 'social attitudes' have been the major
focus of political psychologists for some time (see e.g. Ferguson 1939).

While a number of studies have been directed at the manner in which supporters of different Swedish political parties differ with respect to 'issue-oriented' political attitudes (e.g. Linden 1975; Holmberg 1978,1981; Ekehammar & Sidanius 1977) very little has been done to investigate the differences between supporters of different Swedish political parties with respect to psychologically more relevant 'social attitudes'. One of the few studies with this second direction was conducted by Sidanius, Ekehammar & Lukowsky (1983) who investigated the nature of differences between supporters of different ideological blocs within a higher-order, social attitude space. The results uncovered two major higher-order dimensions optimally differentiating between different ideological blocs (i.e. liberals, socialist, etc.). The more important of these higher-order dimensions was labelled 'Left-Right' and the second and less important dimension was labelled 'Humanism-Xenophobia'. By and large, these results confirm the distinction between 'Economic' vs. 'Social' conservatism made by Lipset (1959) and 'Radicalism-Conservatism' vs. Toughmindedness-Tendermindedness' made by Eysenck (1.951, 1954, 1978). Although the work of Sidanius et al. (1983) had addressed the issue of the higher-order discriminating social attitude space for political blocs, one can still ask questions concerning the manner in which supporters of specific political parties differ from each other along narrower and more specific so-called 'first-order' dimensions of social attitudes.

Because the overwhelming number of studies done within political psychology are conducted in the United States with American samples, it seems especially important to examine the relationship between political party preference and socio-political attitudes within an entirely different context - Sweden. The use of Swedish data has an additional advantage in that the range of genuine and viable political choice is much greater in Sweden, and in many other Western European countries, than it is in the USA. That is to say, there is a larger number of political parties, each representing a different and relatively coherent political philosophy. These advantages are not slight if we are interested in being able to derive generalizable principles concerning the nature of interaction between belief systems and political choice.

Consequently, this study will ask two questions: (1) Along which primary social attitude dimensions do supporters of different Swedish political parties differ? (2) What is the maximum number and nature of those primary social attitude dimensions necessary in order to discriminate supporters from any given political party from supporters from all other

parties? In other words, which particular ideological dimensions distinguish
a particular party from other parties?

Method

Respondents

Seven hundred and eighty-three Swedish high school students were cluster sampled from five neighborhoods within metropolitan Stockholm during the spring of 1979. The clusters were chosen in such a way as to represent all major types of academic and vocational lines of study. The median age was approximately 18 years.

Procedure

Each respondent was given a questionnaire booklet which, among other things, contained the S6 Conservatism Scale (see Sidanius 1976) and questions concerning the respondents' political party preferences. The present version of the S6 Scale consists of 36 items to which the respondents are to respond along a seven-step scale from 1 'very negative' to 7 'very positive'. The items are balanced against acquiescence response set and are similar in content to the C-Scale (Wilson & Patterson 1968), Eysenck's Social Attitude Scale (1975) and Kerlinger's Referents Scale (1967). The S6 Scales have proven to be highly valid and reliable and have been used and validated in both Sweden and Australia (see Sidanius 1976; Sidanius & Ekehammar 1976, 1979; Sidanius, Ekehammar & Ross 1979). (See S6 Conservatism Scale in Table 1.)

The respondents were asked: 'Which political party do you think is best today?' The respondents with party preferences on the 'extreme right' and the 'extreme left' were grouped into 'extreme right' and 'extreme left' categories. For example, respondents preferring the New Swedish Movement (Nysvenska Rorelsen), the Swedish National Union (Sveriges Nationella Forbund), the Nordic National Party (Nordiska Rikspartiet) and the Democratic Alliance (Demokratisk Allians) were grouped into a single category to be referred to as 'fascist'. Respondents preferring the Left Communist Party (Vansterpartiet Kommunisterna), the Swedish Communist Party (Sveriges Kommunistiska Parti), the Communist Union of Marxist-Leninists (KFML-R) and the Worker's Communist Party (Arbetarpartiet Kommunisterna) were grouped together and will be referred to as 'communists'. This grouping procedure was necessary because of the small number of respondents with these party preferences. The distributions of the resulting parties and party categories were: fascists, N = 13; conservatives (Moderata Samlingspartiet), N= 196; liberals (a) People's Party (Folkpartiet), N = 96, (b) Center Party, N = 67; social democrats,

N = 195; and communists, N = 45 (171 respondents did not indicate any
party preference and were subsequently not included in the analyses).1

Results

Primary Social Attitude Dimensions

The primary dimensions of social attitudes were derived by the use of principal factors analysis of the S6 Scale. The number of factors to be extracted was determined by use of the number of interpretable factors satisfying Kaiser's criterion (i.e. eigenvalues 2=1.00). Eight interpretable factors accounting for approximately 55 per cent of the total variance were selected and rotated into oblimin biquartimin simple structure using a delta value of 0.00. The factors and the items primarily defining them were (1) Social Inequality (items 21, 23, 30, 34), (2) Religion (items 13, 25, 33), (3) Political-Economic Conservatism (PEC) items 4, 8, 22), (4) Punitiveness (items 1, 19, 29, 32), (5) Xenophobia (items 5, 10, 15, 16, 20, 27, 32), (6) Pro-West (items 2, 11, 31, 35), (7) Militarism (items 6, 26) and (8) Racism (items 9, 12, 16). The dimensions found here are fairly well-known and relatively stable having been found before in a comparison of Swedish and Australian young adults (see Sidanius, Ekehammar & Ross 1979).

Approximate factor scores were generated for each respondent on

each

factor by multiplying the items primarily defining each factor above by their loadings on the factor score matrix (see Gorsuch 1974, 236-239; see Table 1). (Note that the primary dimensions were scored so that a high score indicates 'much of the dimension in question.) The resulting correlations among these factor score defined primary dimensions are of some interest and are found in Table 2.

It should be noted that of the twenty-eight correlations among the primary factors, all but four were statistically significant at the .001 level or beyond and all of these four exceptions involved correlations with the dimension Religion with the other social attitude dimensions. Furthermore, with the exception of Religion's correlation with the other dimensions, all but two of the correlations were positive. This means that the greater one's degree of conservatism in one dimension, the greater one's degree of conservatism in the other dimensions as well. For example, the correlations between Racism and Social Inequality (r = .52) indicated that the greater one's racial prejudice, the more one tends to embrace social elitist ideas.

Primary Social Attitudes vs. Political Party Preference

The first major question posed by this study concerns differences between youthful supporters of various Swedish political parties along certain major social attitude dimensions and was attacked by use of One-way ANOVAs. Eight One-way ANOVAs were computed, one for each attitude dimension, using the supporters' political party preferences as the independent variable and the specific attitude dimensions as the dependent variables. The results revealed that respondents with different political party preferences showed statistically significant differences with respect to all eight social attitude dimensions (see Table 3).

The correlations between the respondents' political party preferences and the various social attitude dimensions in order of strength were: (1) Political-Economic Conservatism (Eta = .80), (2) Social Inequality (Eta = .53), (3) Punitiveness (Eta = .44), (4) Militarism (Eta = .40), (5) Pro-West (Eta = .38), (6) Racism (Eta = .33), (7) Xenophobia (Eta = .29) and (8) Religion (Eta = .18).

Inspection of Table 3 reveals, among other things, that with the exception of the dimension Religion, the two 'right-wing' categories 'fascists' and 'conservatives' were more conservative on all dimensions of social attitude than liberal and left-wing party supporters. Furthermore, the differences in social attitudes between the two liberal party supporters (the People's

Party and the Center Party) were quite negligible for practically all of the attitude dimensions considered here. However, there were a number of dimensions which clearly discriminated between the two 'right-wing' parties. These differences showed that the fascists were substantially more racist, social elitist, militarist and xenophobic than the conservative party supporters. However, the 'fascists' were the least religious of all the parly supporters, even less religious than the communists!

The second question asked concerned the maximum number and nature of those social attitude dimensions which were needed to discriminate supporters of a given party from all other parties. This question was attacked by use of a series of Back-out Logistic Regression analyses. In each analysis, a distinction was made between supporters of a given parly (which were coded '1') and supporters of all other parties (which were coded '0'). The back-out analyses first entered all eight of the sociopolitical dimensions into the logistic equation and thereafter deleted those dimensions from the logistic equation which did not incur a statistically significant loss of information in the ability to distinguish between the given party P; and all other parties. The results of these analyses are found in Table 4.

Inspection of Table 4 reveals that different combinations of the primary social attitude dimensions optimally discriminated particular parties from all other parties. For example, only two social attitude dimensions were necessary to make a distinction between fascist party supporters and 'others'. The optimally discriminating dimensions were Racism and Social Inequality. Examination of the logistic regression coefficients in Table 4 reveals that of the primary social attitudes, Racism was distinctly more important than Social Inequality in the contrast between fascists and 'others'. Inspection of Table 4 reveals further that Political-Economic Conservatism (PEC) was the social attitude dimension most able to discriminate given political parties from all other parties. This indicates that PEC is a highly relevant and salient dimension for all the political parties represented in the Swedish Parliament.

Of all the other things which could be said concerning the results presented in Table 4, I should like to comment upon four points: (1) The supporters of the fascist parties were unique in that they were the only group for whom the issue of Political-Economic Conservatism (i.e. Capitalism vs. Socialism) was not distinguishing. (2) The fascists were the only respondents for whom Racism was of primary discriminating importance. (3) Examination of the absolute relative sizes of the logistic regression coefficients tells us something about the relative importance of the distinguishing dimensions for the contrasts between a given party and other parties. For example, inspection of Table 4 shows that for supporters of the Conservative and Social Democratic parties, PEC was far more important than all

other distinguishing dimensions. The inequality of the saliences of the discriminating dimensions for each party were examined by computing the standard deviations of the logistic regression coefficients for each party vs. 'others' contrast. In degree of 'salience heterogeneity', the results were: (a) social democrats vs. 'others' (sd = .439), (b) conservatives vs. 'others' (sd = .336), (c) People's Party vs. 'others' (sd = .192), (d) Center Party vs. 'others' (sd = .009), (e) communists vs. 'others' (.008) and (f) fascists vs. 'others' (sd = .004). These results indicate substantial differences in the relative importance of the social attitudes for the contrasts between given

parties and all 'others' for the parties considered. For example, for the social democrats and the conservatives the issue of capitalism-socialism was the decidedly most important distinguishing issue, while for the supporters of the two 'extremist' parties (fascists and communists) the distinguishing dimensions found were all of relatively equal salience or importance. (4) Contrary to what one might expect, the supporters of the Conservative Party were found to be less racist than everybody else, once the effects of the other significant attitude dimensions of PEC, Social Inequality, Pro and Punitiveness were taken into account. Similarly, the One-way ANOVA indicates that if we only consider one ideological dimension at a time, the social democrats were found to be less xenophobic, punitive and militaristic than average (i.e. M= — .04, M= — .22 and M = —.19 respectively, see Table 3). However, once the effects of all the social attitude dimensions were considered simultaneously, the positive logistic regression weights of .543, .193, and .251 indicate that being a social democrat in Sweden is associated with higher rather than lower levels of xenophobia, punitiveness and militarism compared to 'everybody else. These phenomena are examples of the sometimes amazing and counter intuitive Suppressor Variables Effects (see Cohen & Cohen 1975).

The nature of the data in Table 4 gives reason to believe that the slightly higher racism exhibited by conservatives (M = .23, see Table 3) is driven by the conservatives' concern with social elitism rather than with racism per se. This interpretation was tested by examination of the first-order partial correlation coefficient between the Conservative vs. 'others' distinction and Racism while controlling for Social Inequality.² This first-order partial correlation was found to be —.10. This confirms the moderating role of Social Inequality and indicates that once Social Inequality is held constant, conservatives are found to be slightly less racist than everybody else. Likewise, aid in the interpretation of the suppressor variable effects found for social democrats can be found with the use of the partial correlation coefficient. Here again, a first-order partial correlation coefficient was computed between the social democrat vs. 'others' distinction and Xenophobia controlling for PEC. The resulting first-order partial was found to be .23. This indicates that once the effects of PEC itself are held constant, social democrats are found to be slightly more xenophobic than 'everybody else'. (More will be said about this later.)

Summary and Discussion

This study has been directed at uncovering the manner in which respondents
with different political party preferences in Sweden differ with respect to
certain major social attitude dimensions not often studied by political

scientists. The social attitude dimensions of particular interest in this regard were Xenophobia, Punitiveness, Militarism, Pro-Western sympathy and Racism. Two separate kinds of questions were asked: (a) Along which social attitude dimensions do youthful supporters of different political parties in Sweden show significant differences? This question was answered using a series of One-way ANOVAs. (b) What are the maximum number and nature of social attitude dimensions necessary to distinguish any given party from all other parties? This question was answered by using a series of Back-out Logistic Regression analyses. The bivariate ANOVAs relating political party preferences to the eight social attitude dimensions showed that the youthful supporters of the political parties in Sweden were significantly different with respect to all eight of the major attitude dimensions found. Not surprisingly, the results showed that the issue of capitalism vs. socialism (i.e. PEC) was the singly most important dimension distinguishing between the parties. This result is most consistent with all other studies of this issue in Scandinavia known to this author (see e.g. Ekehammar & Sidanius 1977, 1979; Peterson 1976; Linden 1975, Lund 1974; Sidanius & Ekehammar 1979; Sidanius, Ekehammar & Ross 1979) and is also fairly consistent with the Downsian model of political conflict in Western democracies (see Downs 1957). The only exception to this general trend was the inability of the PEC dimensions to contribute to the contrast between fascists and 'others'. The fascists were distinguished by their relatively high degree of Racism and Social Elitism. This result is of considerable theoretical interest in that (1) it supports the notion that fascism is primarily a 'nationalistic' rather than a 'conservative' ideology, (2) it gives us some indication of the potential political niche some fascist party might occupy in the event of acute ethnic conflict in Sweden at some future point and (3) it partially confirms certain fundamental assumptions of Kerlinger's Criterial Referents Theory (see Kerlinger 1967, 1972).

According to Criterial Referents Theory different political categories of people (e.g. liberals, conservatives, moderates, etc.) differ in the kinds of issues which are important, salient or 'criterial' to them. For instance, for conservatives issues concerning religion, morality and property will be 'criterial' or salient in their belief systems, while for liberals issues concerning economic reform, social and economic equality will serve as the major foci in their belief systems. The results of this study partly support this crucial assumption in Kerlinger's theory. Partial confirmation of this notion is found in the fact that different kinds of attitude dimensions do appear to be differentially salient or 'criterial' for different political categories or blocs. However, for five of the six political categories examined, the attitude dimension of Political-Economic Conservatism was not just salient but the most salient of all the dimensions considered. It would seem safe to conclude that certain ideological dimensions will be criterial for a

few political categories while other ideological dimensions will be criterial
for almost all categories.

Another set of results of some theoretical import are the bi- and multivariate findings concerning supporters of the communist parties. The bivariate analysis showed that the communists were less 'conservative' than all other parties in all eight dimensions considered. What is theoretically important about this is that it contradicts the thesis of 'left-wing authoritarianism' espoused by Eysenck (1951, 1953, 1978), Shils (1954) and others. According to Eysenck, one of the foremost spokesmen for this theory, both fascists and communists are supposed to be 'toughminded' as opposed to liberals and socialists who are supposed to be 'tenderminded'. Eysenck maintains that some of the major characteristics of toughmindedness' are punitiveness, racism and xenophobia (see Eysenck & Wilson 1978, 5). Not only were the communists distinctly less punitive and racist: than the fascists, but they were in fact the least punitive and racist of all of the parties considered. Furthermore, it was the communists' lower level of punitiveness which distinguished them from the 'average other' in the multivariate analyses. These are not the only results which contradict the 'left-wing authoritarianism thesis'. Comparisons of supporters of different political parties with respect to intolerance of ambiguity, a central concept in left-wing authoritarianism theory, also showed that communists had the lowest level of intolerance of ambiguity of all major political categories in Sweden (Sidanius 1978; see also Sidanius & Ekehammar 1976; Stone 1980). One possible reason for the differences in the results in this study and Eysenck's results concerning the relative characteristics of communists is the kinds of communists being studied. It is very possible that the communists involved in Eysenck's studies in the 1950s were primarily 'Stalinist' in orientation with all of the personality characteristics that might be implied by that. However, the kinds of communists represented in this study and who constitute the bulk of contemporary Swedish communists and possibly European communists, are what has been described as Euro-communists. Euro-communists, as opposed to Stalinists, believe in democratic and parliamentary institutions, are represented in a number of European parliaments, and draw a disproportionate number of their constituents from the ranks of the intellectual elite (see Holmberg 1981; Nilsson, Ekehammar & Sidanius 1985).³

Another point worth mentioning in these results is the very high relative correlation between political party preference and socio-political attitudes found in this study. It will be recalled that the correlation between Political Conservatism and political party preference was .80 (Eta). This correlation should be compared with correlations of .39 in 'elite' and 11 in 'cross-sectional' samples of Americans between socio-political 'issues' and political party preference (see Converse 1964). This difference is even

more dramatic when one considers that the Swedish sample here consisted of young people, a group known for its relative low level of political interest (see Jennings & Niemi 1974). Despite the possible differences in the relative validities of the attitude scales used by Converse (1964) and the S6 Scale, these differences lend further support to the notion that there is a much greater correspondence between socio-political ideology and political party preference in Europe than in the United States, even for samples of very young, inexperienced and relatively politically disinterested Europeans.

However, perhaps the most fascinating aspect of these results has to do with the complexity of some of the relationships between political party preference and socio-political ideology, specifically the nature and interpretation of the suppressor variable effects found in the relationships between adherence to the Conservative and Social Democratic parties, on the one hand, and various socio-political attitude dimensions on the other. For example, if one asks, 'Are conservatives more or less racist than other people?' the answer would have to be, 'Well, it depends'. Conservatives do have a tendency to exhibit more racist kinds of behavior in various situations. However, this behavior might not be 'driven' or motivated by higher levels of racism per se but rather by conservatives' greater emphasis on Social Elitism. In a similar vein, social democrats tend to behave in a relatively open fashion when confronted with people from other nations and cultures. However, this openness seems more primarily to be a function of the ideological consequences of socialist principles and rhetoric rather than genuine xenophilia per se. This interpretation of the findings seems consistent with remarks made by Rokeach (1973) to the effect that people may exhibit any given attitude for entirely different reasons or different attitudes for the same reason:

A given attitude held by different persons need not be in the service of the same value or same subset of values. A favorable attitude toward socialized medicine, for instance, may serve the value of equality in one person and the value of family security in another; an unfavorable attitude toward the church may serve one person's value for independence and another person's value for honesty; a favorable attitude toward Blacks may be in the service of one person's value for equality and another person's value for kindness (Rokeach 1973, 96).

These data seem to reinforce, once again, the idea that one must be very careful in interpreting the results of bivariate analyses. The meaning of the analysis is dependent upon and must be derived from a specific context. For example the meaning of the correlations between a given party preference and socio-political ideology must depend upon the entire set of socio-political attitudes being considered.

Finally, these results are interesting in that they provide some notion of
the manner in which Swedish political parties might align themselves along
future conflict dimensions such as the treatment of criminals and foreign

immigrants, issues which are as yet not terribly salient or 'criterial' in the
Swedish parliament.

---

NOTES

1. There is also a religious party in Sweden (Kristen Demokratisk Samling); however, since there was only one respondent with this preference, this category was dropped.

2. This partial correlation examines the relationship between the conservative vs. other distinction on the one hand and Racism on the other, while holding the effects of Social Inequality constant.

3. Approximately 76 per cent of the communists in this sample were supporters of the VPK party, which is generally considered a Euro-communist party.

REFERENCES

Budner, S. 1962. 'Intolerance of Ambiguity as a Personality Variable', Journal of Personality
30, 29-50.

Cohen, J. & Cohen, P. 1975. Applied Multiple Regression/Correlation Analysis for the
Behavioral Sciences. Hillsdale, N.J.: Lawrence Erlbaum.

Converse, P. E. 1964. 'The Nature of Belief Systems in Mass Publics', in Apter, D., td.,
Ideology and Discontent. New York: Free Press.

Downs, A. 1957. An Economic Theory of Democracy. New York: Harper & Row.

Eysenck, H. J. 1951. 'Primary Social Attitudes as Related to Social Class and Political Party',
British Journal of Sociology 2, 198-209.

Eysenck, H. J. 1975. 'The Structure of Social Attitudes', British Journal of Social and Clinical
Psychology 14, 323-331.

Eysenck, H. J. 1953. 'Primary Social Attitudes: A Comparison of Attitude Patterne in
England, Germany and Sweden', Journal of Abnormal and Social Psychology, 563-568.

Eysenck, H. J. 1954. The Psychology of Politics. London: Kegan Paul.

Eysenck, H. J. & Wilson, G. D. 1978. The Psychological Basis of Ideology. Lancaster: MTP
Press Limited.

Ferguson, L. W. 1939. 'Primary Social Attitudes', Journal of Psychology 8, 217-223.

Holmberg, S. 1978. 'Kårnkraften och vånster-hogerdimensionen: konfliktlinjer i svensk
politik', Statsvetenskaplig Tidskrift 2, 67-76.

Holmberg, S. 1981. Svenska våljare, Stockholm: LiberForlag.

Jennings, K. M. & Niemi, R. G. 1974. The Political Character of Adolescence: The Influence
of Families and Schools. Princeton, N.J.: Princeton University Press.

Kerlinger, F. N. 1967. 'Social Attitudes and their Criterial Referents: A Structural Theory',
Psychological Review 74, 110-122.

Kerlinger, F. N. 1972. 'The Structure and Content of Social Attitude Referents: A Preliminary
Study', Educational and Psychological Measurement 32, 613-630.

Larsen, S. U., Hagtvet, B. & Myklebust, J. P. 1986. Who Were the Fascists?: Social Roots of
European Fascism. Bergen: Universitetsforlaget.

Linden, M. 1975. 'Some Cognitive Models of Political Party Sympathy Based on Swedish
Attitude Data', Scandinavian Journal of Psychology 16, 285-293.

Lipset, S. M. 1959. Political Man. New York: Doubleday.

Lund, T. 1974. 'Multidimensional Scaling of Political Parties', Scandinavian Journal of
Psychology 15. 108-118.

Neter, J. & Wassennan, W. 1974. Applied Linear Statistical Models. Homewood, Illinois:
Richard D. Irwin, Inc.

Nilsson, 1., Ekehammar, B. & Sidanius, J. 1985. 'Education and Sociopolitical Attitudes',
Scandinavian Journal of Educational Research 29, 1-15.

Peterson, O. 1977. Valundersokningar. Rapport 2. Våljarna och valet 1976, Stockholm:
Liberforlag.

Rokeach, M. 1973. The Nature of Human Values. New York: The Free Press.

Shils, E. 1954. 'Authoritarianism: "Right" and "Left"', in Christie, R. & Jahoda, M. eds.:
Studies in the Scope and Method of The Authoritarian Personality. Glencoe, 111.: Free Press.

Sidanius, J. 1976. 'Further Tests of a Swedish Scale of Conservatism'. Reports from the
Department of Psychology, No. 467, University of Stockholm.

Sidanius, J. 1978. 'Intolerance of Ambiguity and Socio-political Ideology: A Multidimensional
Analysis', European Journal of Social Psychology 8, 215-235.

Sidanius, J. & Ekehammar, B. 1976. 'Cognitive Functioning and Socio-political Ideology: A
Multidimensional and Individualized Analysis', Scandinavian Journal of Psychology 17,
205-216.

Sidanius, J., Ekehammar, B. & Lukowsky, J. 1983. 'Social Status and Socio-political
Ideology', Youth and Society 14, 395-415.

Sidanius, J., Ekehammar, B. & Lukowsky, J. 1982. 'Socio- political Ideology and Dimensions
of Political Party Conflict', The Mankind Quarterly 7, 1-25.

Sidanius, J., Ekehammar, B. & Ross, J. 1979. 'Comparisons of socio-political attitudes
between two democratic societies', International Journal of Psychology 14, 225-240.

Stone, W. F. 1980. "The Myth of Left-wing Authoritarianism', Political Psychology 2, 3-19.

Wilson, G. D. & Patterson, J. R. 1968. 'A new measure of conservatism', British Journal of
Social and Clinical Psychology 7, 264-269.