ETHNOCENTRISM AND CULTURAL STEREOTYPES OF MUSLIMS IN SPAIN

This research studies the relationship between ethnocentrism and Muslim stereotypes in Spain. Stereotypes are defi ned as positive or negative orientations related to attributes associated with an image of Muslims. After considering the application of a latent variable or a cumulative scale, we chose the latter in order to minimize the effect of missing values. The two strategies (that measure positive or negative stereotypes) are consistent with each other and express complementary measurements. By specifying and adjusting a structural model, we established an empirical relationship between ethnocentrism and 18 Paloma Alaminos-Fernández y Antonio Francisco Alaminos-Fernández OBETS. Revista de Ciencias Sociales. Vol. 15, no 1, 2020, pp. 17-42. DOI: 10.14198/OBETS2020.15.1.01 stereotypes, controlling for the effect of gender, age, education, income, political ideology, and habitat size. We conclude that there is a direct effect of age and educational level as explanatory variables of ethnocentric feeling, as well as gender, ideological position and habitat size with respect to the presence of positive stereotypes.


INTRODUCTION
Stereotypes are frequently used in daily life. According to previous studies (Alaminos-Fernández, A. F. & Alaminos-Fernández P., 2012;Martínez-Villar, M., & Alaminos-Fernández, A. F., 2017), stereotypes are categories used in social interactions in order to classify and label individuals. They are used both in terms of others and oneself. Thus, when we say the French are romantic or Italians are artistic, we use stereotype categories to simplify reality. They are applied in a generalized way, referring to, for instance, men or women, a city or place of origin, homosexuality, elderly people, etc. This process of generating and applying stereotypes can obviously be applied to ethnic and religious differences. Ethnic stereotypes offer an opinion and appraisal of other cultures deemed different. Brown (1997) suggested that stereotypes are a representation of reality, often full of aspects of value and emotion that indicate to the person using them whether such aspects are positive, negative or irrelevant 1 . Stereotypes directly refer to personal (collective or motivational), social and cultural processes.
Etymologically, the stereotype as a concept is made up of the Greek words τερεός [stereós], "fi rm, solid", and τύπος [typos], «impression, mark». Walter Lippmann (1922) highlighted the essential role of mental images that individuals have regarding their perceptions of social events and valuation of others. According to Lippmann, human beings fi nd it diffi cult to directly interpret the world due to its complexity, size and state of constant change. In order to process information, individuals construct an abbreviated version of the world in a subjective and biased way, such that, even when the world is shared, they feel and think differently about it. This abbreviated version relies on categories and stereotypes.
Lippmann defi nes stereotypes as stable generalizations that refer to social groups with illogical or irrational content. From this author' s point of view, stereotypes defi ne a biased and incorrect representation of the world, by generalizing and failing to take into account the variability of each individual. Simplifying and subjectivizing leads to knowledge of and perceptions about the world that are negatively conditioned and biased.
A number of studies have been carried out since Lippman' s conceptual proposal, especially in the fi eld of social psychology. The fi rst fundamental work from the fi eld of social psychology regarding the study of stereotypes was that of Katz and Braly (1933). The authors measured stereotyped beliefs and determined the content of a number of ethnic and racial stereotypes. In order to do so, they considered a number of characteristics (such as aggressive, hard-working, etc.) for ten ethnic/nationality groups (Italian, American, Turkish, Black, German, British, Hebrew, Irish, Japanese and Chinese). In 1 The concept of stereotype and prejudice are different, although they are closely linked. One could say that a stereotype shapes the core of prejudice. Prejudices are defi ned as a negative attitude regarding another group or a member of it, and this negative attitude is mainly based on a negative stereotype shaped by social differences with a social group resulting from a set of attributes. this study Katz and Braly defi ned stereotypes as an unrepresentative fi xed and unchangeable feeling about the reality that is claimed to be represented. Stereotypes are produced when of individuals defi ne an individual before observing them. Likewise, stereotyping is a process of creating an indiscriminate concept involving different kinds of experiences and associating these experiences with concept on the basis of a misleading correspondence. In this case, stereotypes are considered to be a way of abnormal and (negatively) biased thinking based on an attitude of rejection toward other groups. The experiment concludes that stereotypes are maintained under the defi nition of being a mental bias of an individual infl uencing his/her perception, leading to incorrect and inadequate inferences of reality.
Subsequently, various psychological approaches are oriented to the study of stereotypes. This is the case of the psychoanalytic explanations proposed by Dollard et al. (1967) regarding frustration, aggressiveness, or Adorno et al. (1950 regarding authoritarian personality. These theories were heavily criticized by Pettigrew (1958), a social psychologist, who claimed that stereotypes are more of a matter of differences between sociocultural norms than personality traits. Other approaches look at stereotypes in terms of a social dimension and suggest that they are used as a weapon in disputes between groups, in an attempt to discriminate on the basis of limited resources. In this sense, authors such as Stonequist (1979), McConahay (1986 or Van Dijk (1993) have worked on these approaches.
In line with the suggestion that stereotypes are a product of the social and cultural contexts, cultural psychology proposes that reality is not defi ned by an individual' s perception but is the result of the system of categories integrated within the individual' s culture. A Kantian approach suggests that reality is interpreted through a set of pre-established categories that related to an individual' s cultural context. From this point of view, stereotypes are rooted in the history and language of a community, and individuals are agents of socialization in the creation and preservation of stereotypes.
In summary, two different approaches are laid out regarding stereotypes. One considers them to be part of a mistaken conception of reality. The other proposes that stereotype production processes are a normal and common phenomenon that is a part of the cognitive activity of individuals. Stereotypes are a mechanism that is inserted into a social and cultural context, as well as in the relationship among groups. As such, stereotypes are fl exible and dependant on these group relationships. In this second approach, stereotypes are not a set of aberrant thoughts, but rather a standard way of dealing with reality. This approach is supported by the work of Allport (1973), which highlights how social categorization processes of stereotypifi ng intertwine, and the work of Sherif et al. (1961), in which stereotypes are a consequence of interactions among groups. Also, Tajfel suggests that stereotypes are not irrational judgements, but are generated in normal processes of human thinking, ie; they are a part of social categorization and differentiation of subjects in different positions.
Current research on stereotypes has also been infl uenced by Tajfel and Turner' s (1979) social identity theory. According to this theory, the image that a person has of him/herself is derived from a feeling of belonging to a specifi c group (working-class, women, socialist, etc.) In general, people try to develop a positive social identity, by belonging to a group that is positively valued in comparison to other groups. Tajfel considers stereotypes to be linked to social identity, which is formed by that part of the image of oneself that is derived from belonging to one or more social groups, alongside the emotional value and meaning that is attributed to this sense of belonging.
By using other groups as a reference point, the concept of relationship is included in the defi nition of stereotypes. The concept of ethnocentrism (in confrontations with other groups) emerges, in which the intragroup (group of belonging) is valued more positively than the exogroup (groups to which one does not belong). From this point of view, stereotypes are part of the confl ict between intragroups and exogroups.
As a theoretical concept ethnocentrism emerged through the study of behaviors and relationships among groups (Sumner, 1906). Ethnocentrism is linked to key concepts such as xenophobia, racism and intercultural competencies. In general, ethnocentrism is a social phenomenon that exists in any group of individuals and involves group belonging, the consideration of oneself as a superior form of life as compared to those from groups (Sharma et al., 1995). This approach is supported by various authors. For example, Aguilera (2002) defi nes ethnocentrism as an attitude by which a group believes itself to have a central position in comparison to other groups; ie a group values its own achievements and features more positively. Thus, every social and cultural group is ethnocentric to a lesser or greater degree. Ethnocentrism can be understood (according to scientifi c discipline) as the cultural sociocentrism of a human group based on culture.
Indeed, the defi nition of ethnocentrism as a relationship between different groups, leads to a double functioning of the concept. In the approaches applied by researchers, ethnocentrism has been assessed in two ways, according to function in terms of intragroups or exogroups. Giner et al. (1998: 277) highlights such double functionality: "ethnocentrism is an attitude that considers the world and others from a cultural and ethnic perspective.
[…] Therefore, it is a basic process that supports collective solidarity and identity and, at the same time, establishes differences and inequalities regarding others, especially foreigners or immigrants".
From an intragroup perspective, ethnocentrism is a concept that is more robust and resistant than other expressions of identity. Several authors consider it to be a psychological or cultural constant in societies. From an anthropological point of view, Lévi-Strauss presents ethnocentrism as natural and intrinsic for humankind, resulting from "the desire of each culture to be resistant to surrounding cultures and to be different. In order for cultures to not disappear, they should remain, in some sense, resistant" (cited by Geertz, 2000: 70).
Furthermore, from an evolutionary psychology point of view, ethnocentrism is defi ned as favoritism towards a group and disregard or hostility towards other groups. It is also considered as one of the mechanisms that exists in all cultures, thus explaining human behaviour from an evolutionary perspective (Yamamoto y Araújo, 2009). In this sense, ethnocentric can be understood as "everything relating to not only ethnicity (a highly disputed concept), but also to psychosocial identifi cation" (Fierro,1987: 158). This perspective considers that ethnocentrism has its foundations in a psychological attitude that emerges among individuals when they are in unexpected situations, and which consists of rejecting cultural forms (traditions, uses, norms…) for being different or remote from other, closer forms with which they identify (Aguilera, 2002). Jones and Smith (2001) suggest the importance of studying the difference between ethnic identity and civic national identity. Ethnic identity is solid regardless of globalization, mass migrations and cultural diversity. Irrational feelings of belonging to a group are more permanent than are civic or rational expressions of belonging.
From an intragroup perspective, ethnocentrism maintains social cohesion and loyalty to group principles. Ethnocentrism serves to preserve culture, solidarity, loyalty, cooperation, defense and survival within a group (Caruana, 1996;Luque-Martínez et al., 2000).
Internal group consistency is created based on a notion of difference and often inferiority towards other groups. It is the basis of all colonization and territorial expansion processes. It is obvious that ethnocentrism can lead to phenomena such as violent nationalism or racism (Aguilera, 2002). In fact, ethnocentrism is a starting point (by means of the defi nition of intragroup) for the exclusion of groups that are considered to be different. For this reason, the empirical relationship between ethnocentrism and xenophobia  and/or racism is consistent and permanent from a comparative research perspective. Racism has been traditionally defi ned as a "process of marginalizing, excluding and discriminating against those defi ned as different based on their skin color or ethnic identifi cation" (Wetherell, 1996: 178). This does not imply that ethnocentrism is the only source of racism. Several authors highlight how racism is based more on class discrimination than on purely ethnic (Myrdal, 1944;Colectivo IOÉ, 1998) or cultural (Van Dijk, 1987;Solé et al., 2000;Chacón, 2005) rejection. However, it is equally true that ethic origin is most frequently the element that gives rise to racial and discriminatory prejudice, contributing to the degree of ethnic and cultural dissimilarity of the immigrant population compared to the autochthonous population (Brücker et al., 2002:123).
In some respects, racism and xenophobia are extreme consequences of ethnocentrism. Ethnocentrism can hide symbolic racism with subtle representation strategies, defending traditional moral values and with a certain resentment towards the favors obtained by "others" (Solana, 1999, cited in Villanueva, 2001. Ethnocentrism expressed as xenophobia can come about as a consequence of social changes. For example, Giner et al. (1998: 277) claims that "the convergence of ethnocentric values with economic and political power interests contributes to justifying any imposed action: colonialism, language imposition, as well as a stigmatizing ideological attitude: xenophobia and racism".
Ethnocentrism has been studied mostly in terms of a negative perspective of the "other". In particular, it is associated with the study of migration and cultural contact. Multicultural cohabitation and intercultural competencies are two different dimensions used to analyze group dynamics. Malgesini and Giménez (2000) highlight the difference between interculturality and multiculturalism. They cite Moreno (1991: 15) and indicated that multiculturalism "covers a characteristic reality of certain societies in which national and ethnic groups coexist in a differentiated manner in the same area", whereas interculturality "means interaction, exchange, openness and effective solidarity: recognizing values, the ways of life, and symbolic representations, whether referring to the same or different cultures". Similarly, Froufe (1994: 164) defi nes multiculturalism as "the presence of two or more ethnicities and their coexistence in the same society and the same area", and interculturality as the search for "exchange, reciprocity, interaction, mutual relation and solidarity between different ways of understanding life, values, history and social behaviors, etc. in conditions of peer infl uence" (Guichot, 2002;Dietz, 2003;Calvo et al. ,2002).
It is obvious that ethnocentrism carries out a double function, depending on the group that is considered. When its double function is used as a reference, ethnocentrism and xenophobia can be perceived as two sides of the same coin.
There are authors who criticize the continued existence of ethnocentrism as a basic element of the defi nition of belonging, related to migrations and communication. Clifford Geertz suggests that cultural diversity is becoming blurred as "we are increasingly living in an enormous collage [...] the world is coming at each of its local points to look more like a Kuwaiti bazaar than like an English gentleman' s club" (Geertz, 1996: 56). In this sense, Todorov (1991: 95) claims that "a humankind that has discovered universal communication will be more homogeneous than a humankind that does not know anything about it; this does not mean that all differences would be eliminated. Assuming so would imply that societies would be simply the product of mutual ignorance". The relationship between ethnocentrism and the consequences of multicultural societies and the processes of interculturality are especially relevant areas of study, in particular taking into account possible social crises and ways to avoid confl icts in the future. Borboa (2006) proposes that knowledge of "others" enables the understanding and comprehension of cultural aspects that can seem negative at fi rst, but that it becomes a dialogue for understanding and harmonious cohabitation, and marginalizes ethnocentrism, prejudice and discrimination. Authors including Altarejos and García (2003) support the idea that ethnocentrism and cultural relativism are two attitudes that are detrimental to interpersonal communications between people from different cultures.
The relationship between stereotypes and ethnocentrism is important because the former are expressions of the later. Ethnocentrism "can be seen as a diffi culty of thinking about the difference on an intellectual level; on an emotional level, it can refer to feelings of strangeness, fear, hostility, etc. (Rocha, 1984: 7) where ethnocentrism often implies quite a violent apprehension of "others". It can also imply, in the majority of cases, that a misrepresented and manipulated image of others is made. As Pureza (2002:2) summaries: "Ethnocentrism is the inability to look at the world in the eyes of others".

MEASUREMENT OF ETHNOCENTRISM AND STEREOTYPES
Different methodological approaches to measuring ethnocentrism which depend on the discipline of the study in question. A pioneer study coordinated by Adorno et al. (1950) represents a psychometric study of ethnocentric prejudice. This study found a large relationship between ethnocentrism and authoritarianism (Adorno et al., 1950). As Fierro (1987:158) indicates, citing Rokeach and Fruchter (1956), "it can be assumed that the correlations of ethnocentric prejudice are approximately the same as authoritarian ideology. Among them, closed-mindedness and mental rigidity are highlighted, as well as paranoid predisposition and disregard towards others and oneself". In these fi rst measurements of the psychological features of the authoritarian personality, the scales explore dimensions that are possibly closer to racism or xenophobia than ethnocentrism itself, as defi ned from a cultural point of view.
One of the diffi culties in measuring ethnocentrism is "its high vulnerability to the bias of social desirability […] due to the stigma that leads to its admission in societies such as the Spanish society, where any declaration or behavior against the constitutional principles of equality and non-discrimination are censured and even punished (Cea D'Ancona, 2009: 21). Thus, "indirect indicators are preferred over direct ones, as the latter are more exposed to the bias of social desirability and error of non-response. Other indicators that measure affection, attraction or rejection towards people with a different ethnic-national-cultural origin should be included" (Cea D'Ancona, 2009: 39).
In terms of indirect strategies for measuring ethnocentrism, there has been interest from other fi elds of social science, for example international marketing, which measures ethnocentrism as a determinant of disposition towards purchasing foreign products (Klein y Ettenson, 1999;Luque-Martínez et al., 2000;Granzin and Painter, 2001;Kaynak and Kara, 2002;Fernández et al., 2003;Balabanis and Diamantopoulos, 2004;Marín, 2005;Grier et al., 2006;Russell and Russell, 2006;Riefl er and Diamantopoulos, 2007;Jiménez et al, 2008). In these marketing studies, an ethnocentric person would potentially consider buying foreign products to be immoral and unpatriotic as it could damage the national economy and generate a downturn in employment (Shimp and Sharma, 1987;Klein and Ettenson, 1999).
In sociology, the use of public opinion surveys gives less accurate measurement than psychometric scales which can administer multiple items to a limited number of individuals. Surveys involve a few items applied to a large number of individuals. In this research, we use a sociological approach to measure the most characteristic features of ethnocentrism. As Giner et al. (1998: 277) summarizes, "ethnocentrism is an attitude that considers the world and others from a cultural and ethnic perspective. […] By creating a comparative category, ethnocentrism organizes the social reality hierarchically and establishes criteria of superiority and inferiority with regard to people' s own lifestyles and those of others". Ethnocentrism consists of a vision of the world, according to which the person' s own group is considered to be the center of everything, and every other group is perceived through one' s own values, models and defi nition of life. In this sense, the ethnocentrism variable used here concerns the hierarchical order of cultures and lifestyles.
Both variables considered here, ethnocentrism and stereotypes, have been studied previously by these authors (Alaminos A. and Alaminos P., 2012a y 2012b). We used the phrasing of PEW in the case of ethnocentrism and a cumulative scale in the case of stereotypes. With regard to stereotypes, there are important elements, such as higher or lower levels of exposure to the cultural groups considered, that refer to both daily interactions and acceptance. We ensured that the positions are different in theory, where the transmission of stereotypes through socialization combine the importance of an experience or cultural contact, even when produced by means of news or information from the media. In the absence of information regarding the degree of cultural contact between those who respond to the survey and the Muslim population, we must consider that the structure of answers internally contain heterogeneity. This heterogeneity of exposure and socialization can be the specifi c subject of analysis, with the aim of assessing the effect of knowledge on the presence of stereotypes (Alaminos and López, 1999).
The data used this research came from the Global Attitudes Survey carried out in 2011 by the PEW Foundation. The survey addresses how non-Muslims associate or do not associate certain attributes with being a Muslim. In certain societies religion confers an identity, regardless of the degree to which it is practiced. In the case of Spain, being Catholic confers an intragroup identity, while being Islamic makes one a part of the exogroup. This is a consequence of historical and cultural socialization (Alaminos and Penalva, 2012). There are 10 attributes which are considered in the survey, six of which are negative and four positive. Ultimately, the use of a negative or positive attribute has an effect on the cognitive framework with which the respondent considers the questions.
The design of the survey in Spain in 2011 is described by the PEW Foundation as follows. Country: Spain. Sample design: Random Digit Dial (RDD) probability sample representative of telephone households (about 99% of Spanish households) stratifi ed by region and proportional to population size. Mode: Telephone adults 18 plus. Languages: Spanish/Castilian. Fieldwork dates: March 22 -April 5, 2011. Sample size: 1,000. Margin of Error: ±3.5 percentage points. Representative: Telephone households (including cell phone only households).
The cultural ethnocentrism variable reference is written so that respondents do not feel pressure to provide a socially desirable response. Thus, the question states that "our people are not perfect, but our culture is superior to others". The answer options were "completely agree", "considerably agree", "considerably disagree", and "completely disagree". Table 1 shows the answers of the Spanish public to the question that measures the degree of ethnocentrism. In 2011, 9.7% claimed to completely agree with the statement that our culture is superior to others, 34% considerably agreed, 38% considerably disagreed and 16.8% completely disagreed. As a whole, the percentage of individuals that reject the idea of Spanish cultural superiority is higher than those who accept it. Nonetheless, both perceptions are quite balanced in terms of percentage. Thus, those who "completely agree" plus those who "considerably agree" come to 45%, while the sum of those who "considerably disagree" and "completely disagree" total 54%. In this sense, it cannot be said that the majority of the Spanish population feels culturally superior.
Table number 2 shows the associations of different attributes to being a Muslim. The attributes are "generous", "violent", "greedy", "fanatic", "honest", "selfi sh", "immoral", "arrogant", "tolerant", and "respectful towards women". In terms of the concept of being "generous", 52% consider that it is not related to being a Muslim, whereas 37% do associate generosity with being a Muslim. In the case of "violent", 61% of the respondents associate being violent with being a Muslim. A much more positive opinion appears in the case of "greedy". Around 59% do not associate the adjective "greedy" with being a Muslim, however 32% do. About 80% relate to term "fanatic" with the image of Muslims; 17% do not. Another positive attribute is the concept of "honesty"; 45% associate honesty with Muslims, and 41% do not.
The category of being "honest" shows the highest percentage of "do not know" responses, with 11% indicating not knowing to what extent the attribute is or is not associated with being a Muslim. The attribute "selfi sh" is associated with being a Muslim by 49% of respondents, while 41% do not make this association. About 58% of the respondents do not consider the attribute of "immoral" to be associated with being a Muslim. In the case of "arrogant", 47% relate it the image they have of Muslims, with 46% claiming there is no association. Finally, there are more dramatic percentages for tolerance and respect for women, 74% believe Muslims are not tolerant and 88% believe they do not respect women.
The attribute of tolerance is closely related to respecting women. In this sense, the dimension of gender is the most distinctive feature attributed to Muslim. This is something that was already highlighted in previous papers Alaminos P. 2012a, 2012b) both for other western countries, as well as countries with a Muslim majority in terms of how the population considers the West. It is obvious that these two attributes require specifi c treatment in how specifi c features and characteristics of the Muslim stereotype are identifi ed.
The aim of this work is to assess the possible relationship between the degree of ethnocentrism of the respondent and Muslim stereotypes (with a negative or positive orientation). Stereotypes are defi ned by a group of both positive or negative attributes. Respondents must express whether they believe they are associated or not with the Muslim image.  For all of the attributes, there is a relationship between the degree of ethnocentrism and the positive or negative stereotypes with regard to Muslims, with the exception of respect towards women and tolerance. In these two attributes, the more generalized opinion is that they are not associated with the Muslim image, regardless of the respondent' s degree of ethnocentrism.
In order to continue exploring the relations between ethnocentrism and sterotypes, we used two additional strategies in order to measure positive or negative stereotypical orientation towards Muslims. First, we looked at it the dimensionality present in the set of attributes through factorial analysis. From the dimensional analysis of this set of attributes (Alaminos A. y Alaminos P, 2012a), we concluded that there was a main dimension that coordinated the variability of the majority of the considered attributes, with the exception of "tolerant" and "respectful towards women".
In the second strategy we built a scale to count the number of times positive and negative attributes were assigned to being a Muslim. This strategy was chosen as it treats lost values in a conservative way. Thus, we created a scaled for orientation scale of negative or positive stereotypes through an additive process, which counts how many times each respondent associated positive or negative attributes to being a Muslim.
These strategies resulted in two variables that measure stereotypes related to being a Muslim, both positive and negative. The variable showing the positive orientation of Muslim stereotypes counts how many times the respondent has considered that he/she does not identify negative attributes or categories as associated with being a Muslim. In other words, if the respondent did not associate the attribute of greed to Muslims, we consider that the orientation is positive. The second variable follows a similar procedure as it counts how many times the respondent associated negative attributes to being a Muslim. In both cases, the values of the variables range between zero (does not associate attributes) and 10 (associates the maximum of positive or negative attributes, as applicable). All cases present in the sample were included, increasing the value of zero (no attributes) of those who answered "do not know" or "do not answer". In other words, not answering is included as a zero. Table 4 shows the distribution of values for the case of positive stereotypes. The data shows that 13% of people have not attributed any positive attribute to being a Muslim, 13.3% associate one attribute, 11.2% two attributes, 13.4% three attributes, 12.6% four attributes, 10.3% fi ve attributes, 8.9% six attributes, 8% seven attributes, 5.6% eight attributes, 2.3% nine attributes and, fi nally, 1.3% 10 positive attributes. In order to assess the consistency of the scale, the number of negative attributes were measured as an additional scale. The following table shows that 2.4% do not attribute any attribute to being a Muslim, 3.7% one attribute, 7.9% two attributes, 10.3% three attributes, 11% four attributes, 13.3% fi ve attributes, 13.6% six attributes, 11% seven attributes, 10.3% eight attributes, 9.6% nine attributes and, fi nally, 7.1% 10 negative attributes. The correlation between the two variables of positive and negative orientation is negative as it is statistically more signifi cant and with a high correlation of -.88. Other variables considered in this study are explained in the structural analysis.

STRUCTURAL ANALYSIS
The main aim of this study is to assess the relationship between the level of ethnocentrism and the presence of stereotypes towards Muslims in Spain. The scale defi ned by the measurement of positive orientation is used as a dependent variable, and the central relationship is estimated by controlling the effect of signifi cant control variables, which are of a structural or sociodemographic nature, such as age, gender, education, political ideology in terms of left and right, and city size. Due to the level of ordinal measurement of the variables (as is the case of ethnocentrism) a polychoric correlation was applied in the estimation and model adjustment. After fi ltering "no answer" originating from variables of income, educational level or ideology, 887 cases were included in the analysis.
Graph 1 presents the structural model taking the dependent variable as a reference of positive orientation measurement with regard to stereotypes associated with being a Muslim. A suitable adjustment can be observed with a P value of .29 for a RMSA of 0.01. Regarding the model as a whole, the specifi cation defi nes the effect of these structural variables based on the degree of ethnocentrism and stereotypes. Once the model was controlled by fi ve variables, a positive association between the degree of ethnocentrism and positive orientation in terms of stereotypes was observed. A standardized coeffi cient of .13 indicates that the lower the respondent' s degree of ethnocentrism, the higher the number of positive attributes he/she associates to with being a Muslim.
The degree of ethnocentrism is explained in this model by the age of the respondent and the educational level. The older the respondent is, the higher the degree of ethnocentrism -0. 25 (we should remember that ethnocentrism is numbered from higher to lower ethnocentrism from 1 to 4). Likewise, the higher the educational level, the lower the opinion regarding one' s own culture being superior to others .16. In this sense, the degree of ethnocentrism is higher when the person is older, and lower when formal education of the respondent is higher. The variables of gender, ideology or city size do not offer a signifi cant empirical relationship to the levels of ethnocentrism.
There is a negative relationship between gender (of -.14) and positive stereotypes, according to which women tend to have a more negative Muslim image than men. In addition, regarding ideology, the coeffi cient is -.12, thus the more right-wing the individual is, the less positive attributes are associated with the idea of being a Muslim. Two other variables with explanatory empirical value of positive orientation towards being a Muslim in terms of stereotypes are the level of formal education and city size. The higher the level of education, the higher the number of positive attributes associated with the Muslim image (.16.). City size shows a relationship of 0.07 with positive stereotypes. Thus, the larger the habitat size, the greater the probability that the respondent will express a higher number of positive attributes. In short, relations are those that are theoretically foreseeable, where a higher level of education, a more urban habitat, a progressive left-wing ideology or males associate a better valuation of the Muslim image. However, it is important to note that this model does not suggest that ideology, gender or habitat size have a direct relationship to the level of ethnocentrism.

Graph 1. Structural Model Ethnocentrism and Positive Stereotypes (standardized coeffi cients)
Source: own elaboration based on the PEW Global Attitudes Survey, 2011

Graph 2. Structural Model Ethnocentrism and Positive Stereotypes (coeffi cients t-values)
Source: own elaboration based on the PEW Global Attitudes Survey, 2011.
As seen in graph 2, all relationships offer a signifi cant t value. Finally, it can be concluded that there is a signifi cant relationship between the degree of ethnocentrism and the positive orientation of stereotypes about being a Muslim. This relationship remains signifi cant after carrying out the control with different relevant variables, in terms of ideology and socio-demographics.

CONCLUSIONS AND DISCUSSION OF RESULTS
First, it should be noted that this research is based on secondary data. In that sense, analysis designs should consider existing data, and in particular what questions are asked, as well as the answers offered, and the scales used. It is not easy to fi nd accessible data that offer the possibility of empirically measuring theoretical concepts such as ethnocentrism, together with the possibility of evaluating stereotypes. In that sense, the usual strategy is to use proxy variables to measure, indirectly, the concepts considered.
In that sense, the measurement of the concepts considered are limited by the operationalization of the data. For that reason, it is important to check alternative measurement designs in order to test and choose the most reliable, considering the existing data. In this study we considered two alternative operationalizations scales for the measurement of stereotypes. The fi rst design considers a factorial dimension to summarize the attributes associated with the Muslim. The second uses a cumulative scale according to the type of answer (positive or negative regarding the attribute). The measurement of the dimension "Muslim stereotype" showed problems related to the missing values while the alternative design (accumulative scale) maximize the inclusion of cases and remove the possibility to adjust a model based on subpopulations. The disadvantage of this strategy is that the scale may suffer a reduction of the homogeneity, originate for the inclusion of non-parallel items. The attributes "tolerance" and "respect towards women", are tightly related and defi ne a subscale regarding the remaining attributes. In any case, this design is a conservative approach to test the statistic association, because the relations detected among the variables are weakened by this non strictly parallelism in the items considered.
Even in this adverse measurement conditions, the empirical analysis shows a statistically signifi cant relationship between the degree of ethnocentrism of the respondents and the positive orientation with regard to the stereotypical attributes of the Muslim image. This relationship remains signifi cant after controlling it by relevant theoretical variables. This way, ethnocentrism is explained by variables such as age or education, while stereotypes show a higher association to the variables of political ideology, city size, education and gender. The income level variable does not offer a signifi cant association within the model. Adorno, T. W., Frenkel-Brunswik, E., Levinson, D.J., Sanford, R. N. (1950). The Authoritarian Personality. Nueva York: Harper. Aguilera Portales, Rafael Enrique (2002), "El problema del etnocentrismo en el debate antropológico entre Clifford Geertz, Richard Rorty y Lévi-Strauss", en Gazeta de Antropologia, núm. 18, Granada, España. Alaminos-Fernández, A. F. y Alaminos-Fernández, P. (2012) "Estereotipos sobre los "occidentales" en países musulmanes". VIII Jornadas Internacionales de Modelado Estructural en el Análisis de la Realidad Social. Universidad de Alicante. Alicante.