Who is shalom schwartz

We do not have specific expectations as to how these predictions will unfold, since our work seeks to explore new methodological venues of the theory and is not meant to further confirm its validity. These represent our country-units of analysis: there were 30 country-units of analysis, out of which 20 were available from both ESS rounds, 9 from ESS round 6, and 1 from ESS round 7.

Thus, the data set comprised 60 country-units from two ESS rounds. The European Social Survey utilizes national random probability samples. To further augment country representativeness, the provided design and population weights were applied for the current analyses.

All the study materials guide included are deposited online and are open access Witte et al. Each item presents a brief description of some person and pertains to one specific value type. Of the 10 values, nine are assessed by two items each, Universalism is assessed by three items. Respondents were asked to indicate to what degree a fictitious person was like themselves on a response scale ranging from 1 very much like me to 6 not like me at all.

She [He] wants people to admire what she [he] does. We, thus, work with scale scores as manifest variables, for which reliability and validity has been confirmed in a voluminous body of prior work. The world countries rank was used; countries are ranked from highest to lowest in terms of goods and services.

Higher scores indicate more inequality. Life expectancy at birth is a nation-level indicator of the number of years a newborn infant could expect to live if the age-specific mortality rate at the time of birth remains unchanged UNDP, Higher values on the index signify higher life expectancy.

Education was assessed as completed years of full-time education. Ethnic fractionalization is a nation-level indicator of the probability that two members of the same country do not belong to the same ethno-linguistic group Alesina et al. Higher values on the indicator signify higher ethnic diversity. Higher scores on the index signify fewer civil liberties Freedom House, Proportion of religious people was assessed as the proportion of individuals who declare belonging to a religious denomination.

The ideal value types were assessed on the ordinal level and were taken from Table 1. Alternatively, one could first have transformed the value score profile into a value rank order. However, we decided against this procedure because of the mathematical reduction that emerges regarding ties e. For each individual case in the dataset, 10 correlation coefficients resulted. To illustrate: The raw value profile of a case was first correlated with the ideal value profile of Conformity highest score for Conformity , second with the ideal profile of Tradition highest score for Tradition , third with the ideal profile of Security highest score for Security , and so on.

For cases with more than one correlation coefficient above the threshold, the strongest coefficient defined the value class of the case. Cases with no correlation coefficient above the threshold were not classified into one of the 10 value categories but into a new category—the non-classified.

The more frequent the classification of people into a value type in a country-unit, the higher the priority of the respective value class in the culture. Conversely, the less frequent the classification of people into a value type in a country-unit, the lower the priority of the respective value in the culture.

Based on standardized frequency scores percentages , value classifications in a culture were then rank-ordered from highest priority to lowest priority. Subsequently, we examined the dimensionality of cultural values.

We analyzed the country-units in the data set according to their similarities and differences on the ranked percentages of value classes. Formally, the 11 value classifications were treated as cases rows in the dataset and value rank orders per country-unit were treated as variables columns in the dataset in a Principal Component Analysis PCA.

We correlated the country-units over the ranks of the value types due to their percentages. We chose PCA as a variant of exploratory factor analysis although traditional rules of thumb pertaining to the ratio of cases n rankings of value preferences according to the distribution approach, here 11 to variables p country-units, here 60 were not satisfied.

In a simulation study, Preacher and MacCallum showed that this rarely has decisive consequences for the results if only few components are extracted from the covariance matrix Preacher and MacCallum, Important to note, the aim of this analysis is to find latent variables that allow us to group countries. The emergent factor scores not loadings! In other words, the emergent factor scores are to be interpreted as the structure of the latent cultural value profiles in Europe.

If more than one factor exists, the observed empirical value profile of a country-unit is a mixture of more than one latent profile that weights the extracted factor scores pertinent to all country-units by the unique loadings of the respective country-unit.

In more formal terms, factor loadings indicate the similarities of each country-unit to the latent value profile carried by the profile of each set of factor scores. Similarity in cultural values of two country-units is indicated by similar factor loadings on the common latent profile the factor scores. For illustration purposes, the observed cultural value profile in a country-unit is given by the following formula:. In Formula 1, Y is the observed rank-ordered value profile V in Culture C , l is the country-unit loading on Factor 1 indicated by subscript F 1 and on Factor 2 indicated by subscript F 2 , and f is the hidden value profile on Factor 1 indicated by subscript F 1 and on Factor Score 2 indicated by subscript F 2.

In other words, the emergent factors with their scores of values can be interpreted as the latent profile of cultural values in Europe that can be used to predict the observed profile of each country.

Similarity in cultural values of two countries is indicated by similar factor loadings because the common latent profile the factor scores is used with equal weights in both countries.

The factor scores are independent elements that predict the observed empirical value profile of a country. The factor scores are, due to the orthogonal rotation, uncorrelated and have nothing to do with the factor loadings. Factor loadings can be different for two countries because they can have unique weights on the common factor scores.

Correlations between loadings imply that the unique weights across countries are similar. Factor scores, however, are always independent. There is no redundancy if loadings are correlated because the factor scores the latent cultural value profiles in Europe cannot be predicted from factor loadings the latent cultural value profile of a specific country.

Table 4 documents the classification of participants into value types and the rank order of these value classes across all countries country-units in Round 6 and 7 of the ESS. Finland was the country with the highest percentage of people preferring Universalism Vice-versa, Lithuania was the country with the lowest percentage of people preferring Universalism 2.

The percentage of unclassifiable participants was highest in Israel Table 4. Proportions of classified individuals into each value-type and their rank order across countries in rounds 6 and 7 of the ESS. In our analyses, we took this quantity into account and used it as a variable. For Round 6, the median value classification was Benevolence Data of countries with measurements on both rounds allowed us to cross-validate the cross-cultural structure of values. Value rank orders of the 10 countries in only one round of the ESS were assumed as robust for the round without observed data.

Effectively each country thus had two columns in the analyzed data set, regardless of whether it had seen one or two rounds of ESS surveying see Table 3. This was done to avoid having results biased toward the countries with original data in the two rounds. This assumption was justified by the high correlations of countries in the two rounds.

The PCA extracted three factors with eigenvalues above 1. However, the scree plot indicated that two factors explained almost all the overall variance The procedure recommends the retention of factors with eigenvalues greater than 1 that are also greater than the eigenvalues of factors from the randomly generated data set. The results of the parallel analysis showed that a two-factor solution was only slightly worse than a single factor solution [see repository output Witte et al.

Based on all these criteria, we decided that two factors were most reliable in summarizing the data at hand. As shown in Table 5 , Factor 1 was represented by the following rank order of value typologies: Universalism, Self-Direction, Benevolence, and Stimulation vs. Factor 2, on the other hand, was represented by this rank order of value typologies: Conformity, Tradition, Benevolence, Security, and Hedonism vs.

Table 5. Factor score coefficients and the relative importance given to individual level values across countries in Europe. Factor loadings on the two factors were treated as country coefficients of two distinct dimensions of the latent cultural value profile in Europe. The value profile of each individual country was reproduced by its unique loadings on the European latent value profile. A country that had a very high loading on the first factor score had a straightforward value profile that was reliably represented by this factor e.

However, for countries with smaller loadings on the first factor e. In such cases, the empirical value profile must be adjusted based on the latent value profile of the second factor.

To summarize, the very high negative correlation between factor loadings indicates that the value profiles of all countries in Europe available in the dataset can be reproduced by two latent cultural value dimensions in a specific way: The dominant value profile in Europe is the profile of the first factor and the deviation from this latent profile is best described by the increasing influence of the modification by the second latent profile, the greater the deviation the higher the loadings of the second factor.

Figure 2 shows how countries in the data set were positioned in the two-dimensional space of the latent cultural value profile country loadings on the two factors in Europe. Exempting Lithuania, all countries were positioned in the positive sector of Factor 1 and all countries, no exception, were positioned in the negative sector of Factor 2, which was numerically inverted to ease interpretation. There was a clear linear arrangement of countries in the European context.

Whereas countries from the former Communist bloc were grouped at the lower ends of the bi-modal space, countries from Central Europe and Western Europe were clustered at the upper ends of this space. Iceland, Switzerland and all Scandinavian countries were situated in the highest echelons. Figure 2. The position of European countries along two dimensions of cultural values as informed by the distribution approach.

Due to the high negative correlation of the loadings we know that the deviation from the dominant profile is also an increasing similarity with the profile of the second factor. Nevertheless, we will give the results from both analyses. We first correlated the factor loadings with the proposed objective indicators see the Supplementary Table 2.

Then, two multiple hierarchical regressions were conducted one each for the economic and the socio-political indicators. The per-country factor loadings of the two dimensions were treated as dependent variables and all the objective indicators were treated as predictors. Table 6 summarizes the results. The latent value profile of the first factor was associated with all positive conditions and the latent value profile of the second factor was associated with the negative conditions in a country.

We z -transformed indicators with a significant regression coefficient see above and we created indices by summing them up 1. Reverse coding was applied to align the objective indicators in terms of low or high coefficients.

Two indicators of societal challenges were therefore constructed. Small societal challenges were defined as high GDPpc, low GINI, high life expectancy, high civil liberties and low proportions of religiosity. Finally, we correlated these indices with the country-loadings on the two dimensions for a similar procedure see Wainer, The complexity of value profiles of countries increases under moderate levels of SWB because the influence of the latent value profile of the second factor becomes more important.

Populations with extremely high or low SWB, which corresponds to well-being shaped by high vs. For instance, the cultural value profiles of Scandinavian countries are almost perfectly reproduced by the first latent common profile existent in Europe. Conversely, the cultural value profiles of East European countries are almost entirely reproduced by the second latent common profile existent in Europe.

The current paper propagates an alternative and psychometrically sound approach to the assessment of cultural values as suggested by Shalom Schwartz.

We have argued that the dominant averaging approach to empirically infer values at the culture level contradicts the theoretical propositions of the circumplex nature of value priorities at the individual level. The averaging approach accommodates insufficiently the at-times-negative and null correlations between value profiles across the 10 value types for individuals.

We have suggested a different approach to measuring culture-level value preferences, namely one that we call the distribution approach. Unlike the averaging approach, which looks at average scores of the individual preferences, the distribution approach looks at frequencies of individuals who prefer each of the 10 values most in each culture. This article brings new insights to the culture-level value theory proposed by Schwartz. Value priorities are a characteristic of the individual.

Thus, values at the cultural level represent a specific frequency distribution of value priorities of members of the culture. A fictitious middle individual as a prototype for a culture is a problematic approximation both from a conceptual and methodological viewpoint.

The interpretation of culture-level dimensions based on average scores does not give justice to the theory of value priorities at the individual level and therefore should be avoided in future research. Our findings suggest two theoretical ways of how individual values extend to values at the culture level see Table 5.

The first is almost an identical reproduction of the circumplex model of individual-level values at the culture level and has little to nothing to do with the cultural dimensions proposed by Schwartz. The second dimension of intercultural values diverges slightly from the first insofar that it does not follow adequately the proposed circumplex model at the individual level. The present findings show that culture-level values can be structured along two dimensions that describe two widely identified societal tendencies, namely maintaining the status quo and progression from the status quo.

This two-dimensional structure is well supported by empirical evidence Bilsky, ; Strack et al. Moreover, this is the very manner in which Schwartz has proposed his theory of individual-level values—there is one dimension which describes the degree of preservation of the status quo in individual values Conservation vs.

Openness to Change and one dimension which describes self vs. Self-Transcendence Schwartz, Our approach shows how these tendencies, which occur at the individual level, can be captured adequately from a measurement standpoint at the cultural level as well. A visual guide to the culture level value dimensions as it is informed by the distribution approach can be seen in Figure 3. Figure 3. Numbers correspond to value rank order of importance reported in Table 4.

The rank of the non-classified has been eliminated. The stronger emphasized colors correspond to value typologies with a higher weight on the respective dimension. The weaker emphasized colors correspond to value typologies with a lower weight on the respective dimension. To arrive at the value structure of a culture, one requires the weights of the respective culture factor loadings on the two dimensions.

We have argued that the value profile of a country is a weighted combination of the two newly identified latent profiles of cultural values.

In Europe, there is a clear dominance of the value profile known from research on individual values: Universalism, Self-direction, Benevolence, and Stimulation Alteration vs. Personality and social psychology bulletin 29 10 , , Journal of personality and social psychology 4 , , The psychology of values: The Ontario symposium 8 , Journal of cross-cultural psychology 26 1 , , Personality and social psychology bulletin 28 6 , , Journal of cross-cultural psychology 21 2 , , Journal of research in personality 38 3 , , Journal of personality and social psychology 89 6 , , Articles 1—20 Show more.

Help Privacy Terms. Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries SH Schwartz Advances in experimental social psychology 25, , Are there universal aspects in the structure and contents of human values?

SH Schwartz Journal of social issues 50 4 , , Normative influences on altruism SH Schwartz Advances in experimental social psychology 10, , A theory of cultural values and some implications for work SH Schwartz Applied psychology 48 1 , , A theory of cultural value orientations: Explication and applications S Schwartz Comparative sociology 5 , , Values and behavior: Strength and structure of relations A Bardi, SH Schwartz Personality and social psychology bulletin 29 10 , , He is on the editorial boards of five international journals.

He has written or edited 9 books and published over articles in international journals in social, cross-cultural and developmental psychology, sociology, education, management, law and economics. His seminal articles on individual and cultural values have been cited more than 50, publications. His individual-level research includes studies of altruism, intergroup contact, individual values as determinants of political orientations, values of bases of emotions, subjective well-being, and prosocial behavior, the development of values in young children, value transmission in families, values as the motivational bases of everyday behavior, differences among ethnic, gender, and religious groups, and value measurement.

He recently introduced a refinement of his theory of individual values in order to improve understanding of attitudes and behavior. He also studies national differences in cultural value dimensions, their origins, and their consequences for societal functioning and policy.