CC BY
37Journal of Siberian Federal University. Humanities & Social Sciences 4 (2014 7) 579-596
УДК 394.942
Ethnic Tolerance Scale Development: Renovation of Integrated Approach1
Gershons Breslavs*
Baltic Psychology and Management University College 65 Bruninieku Str., Riga, LV1011, Latvia
Received 12.12.2013, received in revised form 13.02.2014, accepted 20.02.2014
Many theoretical models have contributed to thefield of intergroup relations, but the concept of social tolerance is still very ambiguous despite huge progress in social psychology and relatedfields in the recent 50 years. The concept seems too complicated and researchers prefer to analyze tolerance/ intolerance components: stereotypes, prejudice, perception biases and social discrimination which seem to be more one-dimensional and available for the study. At the same time, without such data social policy in ethnic integration cannot be well goal-directed. Social and personality psychology lack integrated criteria for the assessment of ethnic tolerance/intolerance and the above-mentioned one-dimensional partial measures of ethnic tolerance cannot compensate this deficit, since they reveal only some aspects of intergroup attitudes. The elaboration of such a holistic personality measure for the assessment of intergroup relations in Latvia was the goal of this work1. The five-stage process of scale development has been used. Results have shown that a valid and reliable instrument for ethnic tolerance assessment has been elaborated. The model of elaboration could be applied in all countries.
Keywords: social tolerance, ethnic tolerance, ethnic attitudes, ethnic stereotypes, ethnic prejudices, social discrimination, scale of ethnic tolerance.
The study based on the project Elaboration of a method of societal tolerance monitoring supported by the EUgrant subdivision "Forwarding social integration in Latvia" (Project agreement Nr. 2005/017495-03-01/3-1/50).
Tolerance and relative concepts' meaning
Some disadvantages followed also from the borderline position of this topic between social psychology and psychology of personality. Personality researchers view tolerance as an intergroup phenomenon more or less outside the field, but social psychologists try to avoid its study understanding badly-controlled internalization process of attitudes to out-groups and xenophobic
© Siberian Federal University. All rights reserved
* Corresponding author E-mail address: [email protected]
traits shaping. Few psychologists understand the necessity to combine personality and social-psychological approaches describing prejudices and similar constructs (Ekehammar, Akrami & Fan, 2009).
One of the few books devoted to tolerance was published more than 40 years ago. The author of the book defined tolerance as the lack of prejudices (Martin, 1964, 11 lp.) and discussed the joint phenomenon tolerance-prejudice. At
the same time the author doubted if a person with high tolerance would be well-adjusted to interpersonal relationships (Martin, 1964, p. 119). The doubts are based on two arguments. First, in competitive society tolerance to rivals could result in losing in busness competition. Second, it is difficult for an individual to accept behavior of out-group representatives if it interferes with the person's life style and priorities. It is difficult to expect tolerance to gypsy artists from the neighbours of a multy-apartment buiding if they display their musical giftedness at night. According to social exchange theory (Thibaut, & Kelley, 1959), people's estimation of relationships depends on moral and material reinforcement in accordance with the estimation of their own contribution. It means that tolerance cannot be one-sided disposition for a long time, because interference from newcomers who ignore local traditions and disturb the life of the majority will result in the decrease of tolerance and increase of prejudices.
The recent international conflict around cartoons published in the Danish newspaper on Muhammed-terrorist could serve as a good example of religous and ethnic tolerance decrease initiated by local traditions' violation. It was initiated by the Muslim community leaders' appeal to Muslim countries and resulted not only in the deterioration of relationships between communities in Denmark but also in deterioration of relationships between European and Muslim countries around the world. The leaders of the Muslim community required respect for their own religious traditions, at the same time neglecting the cultural traditions of the host country. It is obvious that such conflicts stimulate not striving to mutual understanding in the majority and minority communities, but rather mutual distrust and discrimination. Sometimes the majority actions such as improvement of safety in public space by forbidding female Muslim dresses
veiling the face (the niqab or burka) in public places in France and Belgium can lead to the same outcomes, despite the dresses perturbing the social norms of European culture.
At the same time, attitudes to national, ethnic or religous outgroups are mediated by the socio-historical context, including the history of communities' relationships, especially if recently in their countries serious frictions, war conflicts and/or violence between these groups occurred (Barrett & Oppenheimer, 2011). The „enemy image" can be shaped from early childhood and supported and stabilized later by perception biases and prejudices to disliked outgroups distributed by ethnic or religious homogenous environment. In early adolescence extreme ethnocentrism and xenophobia can be formed, which is highly resistant to change (Barrett & Oppenheimer, 2011). In its turn, the „enemy image" can stimulate permanent interpersonal and intergroup conflicts without objective contradictions between these individuals or groups. The analysis and control of socio-historical, socio-political and nurture factors do not exclude the study of dispositional personality factors.
One of the first empirical and theoretical works trying to explain reasons of intolerance and prejudice by personality's dispositions and traits was the famous book of Theodor Adorno and his colloborators on the authoritarian personality (Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950). The main scale of their questionnaire was F-scale which shows the tendency to blindly obey authorities, discriminate minority out-group representatives and comply with dogmas. According to the authors' ideas, the higher points on the scale show a person's tendency to base personal relationships on prejudices and to discriminate people from out-groups different in racial, ethnic, gender, or other biological or cultural features. If these features in a particular personality are dominant, his/her behaviour will
be intolerant more or less independently of social environment.
Current studies also showed the positive link between right authoritarian political orientation and prejudices (Ekehammar, Akrami, Gylje, Zakrisson, 2004). The hostility shown by authoritarians appears to be directed primarily toward members of outgroups, that is to say nonmainstream or unconventional. They are more ethnocentric and prejudiced, showing greater antipathy toward members of most ethnic groups to which they do not belong (Altemeyer 1988, 1996). Authoritarianism is a strong predictor of prejudice, but it is not the only predictor. The same is true about conservatism as a personal trait (Fazio, Jackson, Dunton, Williams, 1995), and Jim Sidanius and Felicia Pratto (1999) revealed a complementary social-political tendency called the social dominance orientation (SDO) that can predict racial prejudices (Pratto, Sidanius, Stallworth, Malle, 1994; Akrami, Ekehammar, & Araya, 2000). SDO assumes the necessity of social hierachy for individuals and groups and the tendency to emphasize or increase social inequality (Sidanius, & Pratto, 1999).
Individuals who are high in SDO are hypothesized to accept 'hierarchy-enhancing legitimizing myths' which justify social practices that enhance or sustain social inequality, while individuals low in SDO are more likely to endorse 'hierarchy-attenuating legitimizing myths' and justify social practices that reduce inequality (Sidanius & Pratto, 1999, Ch. 4). In short, the theory of SDO argues that group-based hierarchies reproduce and reinforce themselves via individuals who have a general preference for hierarchical structures over egalitarian ones. Research has generally supported the theory's hypothesized correlation between SDO and specific forms of prejudice, stereotyping, and acceptance of legitimizing myths. For example, SDO was positively correlated with anti-Black
racism, sexism, nationalism, cultural elitism, political-economic conservatism, belief in meritocracy, pro-military attitudes, and favoring punitive legal policies (Pratto, et al., 1994). SDO was negatively correlated with favoring civil-rights policies, social welfare, environmentalism, and noblesse oblige.
Some adherents of personality-trait-theory considered social dominance orientation and right authoritarian political orientation as the main factors of prejudice development (Reynolds, Turner, Haslam, & Ryan, 2001). The follower of this approach Whitley unambiguously refers these two concepts to reasons for prejudices (Whitley, 1999). Despite the difficulty to show cause-and-effect relations between these variables and prejudices, data show that SDO is mediated by the influence of in-group's leading position on prejudices (Guimond, Dambrun, Michinov, & Duarte, 2003). It is in accordance with the data that belonging to leading in-groups predicts higher SDO than belonging to subordinated in-groups and that experimentally manipulated status of group significantly influences SDO (Sidanius, & Pratto, 1999).
Although SDO could be considered as a more general phenomenon than the preference of a particular leading group, it considers social status of individuals and groups' vertical aspect only, but does not capture the huge amount of other interpersonal and intergroup relationships, including the process of enculturation through integration by Berry (Berry, Trimble, & Olmedo, 1986; Berry, Poortinga, Segall, Dasen, 2002) and all types of more or less equal interaction between majority and minority group members.
An important psychological variable mediating inter-group tolerance is the type of attachment. Some data show that dispositional attachment anxiety and avoidance are inversely related to compassion, but attachment security (both on subliminal and supraliminal levels) is
positively related to compassion and altruistic behavior. The authors concluded that attachment security could promote prosocial values and tolerance (Mikulincher & Shaver, 2007).
Tolerance concept's definition and study
At the same time, the amount of intergroup conflicts including collective violence has not decreased in the 21st century, and psychologists need more holistic (comprehensive) measures to assess attitudes to out-groups that could predict individual predisposition to involvement in such conflicts and violence.
The last edition of the APA Dictionary gives the following second and third meanings of the notion tolerance: "2. Acceptance of others whose actions, beliefs, physical capabilities, religion, customs, ethnicity, nationality, and so on differ from one's own. 3. A fair and objective attitude toward points of view different from one's own" (APA Dictionary, 2007, p.944).
It seems important to combine both aspects of the definition because tolerant intergroup perception should include both: a) acceptance of others (with different social belonging and background); b) the fair and objective attitude to opinions of others. The study of majority-minority in-group interaction has shown that minority-effected social change left a group stronger when that change increased the group's tolerance than when the group experienced conversion (Prislin, & Filson, 2009).
Taking into account the priority of the first meaning of tolerance (habituation to drugs) in psychology, it is very difficult to find the use of the concept in the second meaning (Mummendey, & Wenzel, 1999; Prislin, & Filson, 2009). In its turn, the reason for this priority could be explained by the ambiguity and multilevel character of the phenomenon of social tolerance. In social-political discourse all reasons of social
conflicts are sometimes explained by the lack of social tolerance.
Central to the analysis of tolerance is the question of how members deal with intergroup difference. If the outgroup's difference is judged to be nonnormative and inferior, devaluation, discrimination, and hostility are likely responses toward the outgroup. Judging the outgroup's difference to be normative or positive leads to acceptance and appreciation of this group (Mummendey, & Wenzel, 1999). For example, in the Latvian establishment ethnic Latvians with American background are perceived more positively than local residents and their speech accent is viewed as more prestigious.
Dissimilarity or "foreignness" has a Janus-faced character as it may elicit either attraction or aversion (Graumann, 1992). When the outgroup's difference is evaluated negatively, perhaps as a challenge or threat to the ingroup's opinions and attributes and hence to the ingroup itself, the outgroup should experience devaluation and discrimination. When the outgroup's difference is, however, evaluated positively, for instance as enrichment or as a variation that in a more abstract sense still confirms the ingroup's views, then the difference should be accepted and the differing outgroup should be treated positively (Mummendey, & Wenzel, 1999). It is possible to find ambivalent attitudes to outgroups: acceptance on the conscious level, but rejection on the unconscious level leads to the minimization of informal contacts with members of outgroups.
In the recent work tolerance was defined as out-groups' perception and estimation without prejudices and opinions based on in-group criteria (values, norms, traditions) only and as acceptance of human (cultural, religious, gender, age, opinions, etc.) differences that are not harmful to other groups' members (Breslavs, Abele, Derjabo, Pisinska & Roze, 2008). The
latter condition seems to be quite important because unconditional acceptance could lead to maladjustment -tolerance to criminal or self-uncontrolled groups could be dangerous for the functioning of a society.
Sometimes in political discourse the notion secure tolerance is used for such understanding of tolerance. The borderline between secure and insecure seems diffuse and ambigious because interests of different groups of residents could be very different, and political establishment has its own interests not coherent with the interests of the majority of citizens, which limits the possibilities of objective expertise in the case of dangers to some social groups. It would be particularly complicated in the case of cultural threats. For example, ethnic Latvians in Latvia, including the establishment, consider that the Russian language of minorities is a real threat to the Latvian language and culture. A special law and many state language control institutions were set up on the implicit basis of this opinion. Attempts to change the status of the Russian language (today it is just a foreign language) are perceived by the Latvian ethnic establishment and mass media as an anti-state and destructive activity.
Theories in the field of intergroup relations can be divided into two groups: general and specific. General theories try to answer the question: why are attitudes to in-groups and outgroups members not equivalent? Only two main theories can be mentioned: Cognitive or Social Identity Theory (SIT) (Tajfel & Turner, 1986), and Functional Theory of Intergroup Relations (FTIR) by Muzafer Sherif (Sherif, 1966).
Tajfel's Social identity theory (SIT) is based on Cooley and Mead's idea on group real or imagined membership as the basis for social identity (Mead, 1934/1967) and on Festinger's theory of social comparison. Taking into account that social identity is an important part of a
person's identity and people prefer a positive self-image or self-concept, Henry Tajfel proposed that similarly with self-concept we tend to assess our own group and its members better than another comparing group and its members irrespective of how strong our affiliation with our own group is (Tajfel, 1979; Tajfel, & Turner, 1986). Previously, Tajfel and his colleagues in Bristol had shown that social categorization per se is sufficient to generate intergroup discrimination, applying minimal group technique (Tajfel, 1970; Tajfel, Billig, Bundy, and Flament, 1971). It means that our belonging (real or imagined) to a group more or less automatically brings about the phenomena of in-groupfavoritism and out-group discrimination or derogation. At the same time SIT considered that identification with one's in-group, not formal belonging only, is a precondition for positive in-group evaluation.
The second approach proposed in the FTIR is that group members' intergroup attitudes and behavior will reflect the real interests of their group vis-à-vis other groups and the type of interaction (Sherif, 1966). Sherif organized a three-stage field experiment on intergroup relationship in a summer camp for boys around 12 years in the 1950-ies. The full design included: group formation in the first few days, when new friends were distributed in out-groups and groups were separated; intergroup competition, when groups were involved in a 'win-lose' game competition with attractive prizes for winners (pen-knifes); and intergroup cooperation at the third stage, when an important joint task was proposed and both groups were forced to combine their efforts. The results showed that the intergroup relations on the second stage became very hostile, with permanent conflicts and mutual accusations with very big in-group favoritism and out-group derogation, but on the third stage out-group derogation was reduced together with intergroup conflicts (Sherif, 1966).
Specific theories attempt to explain particular phenomena and factors of intergroup relationships and attitudes. For example, the optimal distinctiveness model of social identity (Brewer, 1991) holds that group identification is the product of opposing tendencies for the inclusion in the group and differentiation from others. When in-groups become larger, their rules and institutions become rigid and the moral order is seen as absolute rather than relative, the in-group's moral superiority become incompatible with tolerance for differences. If out-groups do not subscribe to the same rules and are perceived as the source of threat to the in-group, the indifference would be replaced by denigration and contempt (Brewer, 1999).
W. G. Stephan and C.W. Stephan (1996) proposed the Integrated Threat theory (ITT) which classifies reasons for negative attitudes to out-groups into four major types: realistic threat, symbolic threat, intergroup anxiety, and negative stereotypes. Rather than conceptualizing the relationship between threat and attitudes as stemming from either competition or value conflict, ITT proposes that both can influence out-group attitudes simultaneously. Within the ITT, a realistic threat includes perceptions of competition, conflicting goals, and threats to physical and economic well-being of the in-group. A symbolic threat is similar to ideas underlying symbolic racism, where a threat arises from a conflict in values, norms, and beliefs between groups.
Although intergroup anxiety and negative stereotypes have not been traditionally considered to be forms of intergroup threat, W. G. Stephan and colleagues argued that they also reflect concerns about negative outcomes from intergroup relations and therefore constitute unique types of threat and a source of a conflict (W. G. Stephan & Stephan, 1996, 2000). Intergroup anxiety involves feelings of uneasiness and awkwardness
in the presence of out-group members because of uncertainty about how to behave toward them, which makes interactions with out-groups seem threatening (W. G. Stephan & Stephan, 1985). Intergroup anxiety has been demonstrated to be a predictor of out-group attitudes and bias (Ho & Jackson, 2001; Islam & Hewstone, 1993b; Voci & Hewstone, 2003) and furthermore, people who tend to be generally high in anxiety often exhibit higher levels of prejudice (Hassan, 1978). Negative stereotypes generate threat by creating negative expectations concerning the behavior of out-group members. Stereotypes have long been associated with negative out-group attitudes and readiness to react in a negative way (Eagly & Mladinic, 1989; Spencer-Rodgers & McGovern, 2002).
Tolerance scale elaboration
Integrated Threat theory was selected as appropriate for the understanding of the concept of tolerance (Stephan & Stephan, 1996). The five-stage process of ethnic tolerance scale elaboration comprises: a) phenomenological data collection through structural interviews and main topics description about existing positive and negative ethnic stereotypes, forms of cooperation and competition, points of misunderstanding, types of attitudes; b) formulation of the set of statements (93) from this data collection according to the Threat theory and pilot study; c) preparation of the first 89-item inventory with six subscales; d) data collection with the 89-item inventory and factor analysis of the data resulting in the reduction of statements to 50 with three subscales (negative attitude + negative stereotypes, positive stereotypes, and positive attitude); e) two-stage data collection with the 50-item inventory and 18-items' antipathy scale with contrast samples, and quantitative analysis (including factor, variance and correlational analysis) of the data.
Structural interviews included four parts: 1) description of Latvian and Russian ethnic groups' essential traits including similarities and differences between Latvians and Russians; 2) criteria of good or bad intergroup relationships on the individual and societal level; 3) factors of good or bad intergroup relationships, including language skills, segregation, and differences in the interpretation of Latvian history; 4) approaches to improving these relationships on the individual and societal level. The data of the interviews were used on the next stage to formulate 93 statements, four of which were eliminated after a pilot study with six Latvian and six Russian-speaking participants.
All 89 statements were distributed on six factors based on previous theoretical assumptions: 1) negative attitudes to the ethnic outgroup, including blatant prejudices and realistic threats; 2) negative stereotypes; 3) symbolic threats; 4) subtle prejudices, including intergroup anxiety; 5) positive stereotypes; 6) positive attitudes. Taking into account the complicated topic of the questionnaire four -point Likert scale with two-type estimations was used: a) agree- partly
agree - partly disagree - disagree; b) regularly -sometimes - seldom - never.
At this stage 128 participants were involved (see Table 1).
Factor analysis on the first data collection stage showed that in the Latvian sample the first factor explains accordingly 17.321 % of the total variance, the following 7.616 %, 5.476 %, 4.613 %, 4.036 %, and 3.496 % of the total variance.
Component factor analysis showed more or less strong three factors-structure only.
Taking into account the Latvian-Russian asymmetry in the types of threats, blatant and subtle prejudices, results of internal consistency analysis using Cronbach's a and the results of the component factor analysis, this ethnic tolerance scale was reduced on the next stage to 50 items and three factors only - a) negative attitudes to ethnic outgroup, including negative stereotypes and prejudices; b) positive attitudes, c) positive stereotypes.
In the process of transformation, selection and rejection of weak items along with psychometric criteria were used, as well as criteria of similarity between coherent items in the Latvian and Russian version. For example,
Table 1. Participants on the 89-item questionnaire (demographic variables)
Samples Latvian sample Russian sample
female male both female male both
N 47 20 67 27 34 61
Age (average) 35.8 34.4 35.4 39 37.4 38.1
Mono-ethnic + mixed-ethnic family 38V + 9J 15 V + 5J 53V + 14J 17V + 10J 22 V + 12J 39V + 22J
Education (university) 25 7 32 12 22 34
Education (high school or secondary school) 22 13 35 15 12 27
Location
Riga 18 7 25 14 20 34
Riga's region or other city 18 9 27 5 9 14
Country-side 11 4 15
Other countries 8 5 13
Scree Plot
Component Number
Fig 1. 89-item tolerance's scale Varimax rotation on 6 factors (Russian sample)
Scree Plot
OWDlfflWU'K^i' —lOWCTJCDhJCJlCO——I О W CT> (D СЛ Co
Component Number
Fig 2. 89-item tolerance's scale Varimax rotation on 6 factors (Latvian sample)
in the factor of positive attitudes to out-group were maintained such items as " Savu brrvo laiku es labprat pavaditu aro krievu komppnija" ("I readily spend leisure time in Russian company too") in the Latvian veosionand "Я охотно проведу время в латышской компании»"
("I willingly epend leisure time in Latvian company") (Breslavs, et al., 2008).
Participants
On (he next stage data were collectrdOrom thesample of 265 pasticipants representing two
Table 2. Participants on the 50-item questionnaire (demographic variables) Social workers and medical personnel were coded as group 1.1. , policemen, lowers and military personnel as group 1.2.
Izlases Latvian Russian
Prof.groups 1.1. 1.2. 1.1. 1.2. 1.1. 1.2. 1.1. 1.2. 1.1. 1.2. 1.1. 1.2.
female male both female male both
Sample features
N 58 54 0 32 58 86 28 25 12 42 40 67
Age (average) 42.4 21.8 - 23.1 42.4 22.3 44.6 22.5 40.8 25.2 43.4 24.2
Mono-ethnic + mixed-ethnic family 48+10 43+11 26+6 48+10 69+17 14+14 13+12 9+3 25+17 23+17 38+29
Education (university) 26 2 - 8 26 10 16 1 10 3 26 4
Education (high or secondary school) 32 52 24 32 76 12 24 2 39 14 63
Location
Riga 18 17 - 5 18 22 17 13 8 18 25 31
Riga's district or other city 25 30 - 14 25 44 10 12 4 15 14 27
The country-side 15 7 - 13 15 20 9 9
From other countries - 1 1
ethnic populations and two types of different professional groups, 251 were accepted as valid (144 Latvians - 58 social workers and 86 policemen, military personnel, and lawyers; and 107 Russian-speaking - 40 social workers and medical personnel, 67 - policemen, military personnel, and lawyers) (see Table 2). In the retesting procedure 57 Latvian and 66 Russian-speaking respondents participated.
Factor component analysis on the second data collection stage showed the reduction of six factors to three main factors only in the Latvian sample: The first factor (negative attitudes + stereotypes) explaining 23.942 % of the total variance. The second factor (positive stereotypes) - explaining 11.537 % of the total variance. The third factor (positive attitudes) explaining 5.641 % of the total variance (see Fig. 3).
The three factor structure of the 50-item ETS was confirmed in the Latvian sample, explaining 41.12 % of the total variance
Three factors with Eigen value>2 are distinguishable on the Scree Plot.
Analysing Fig. 3, two main factors in the Latvian sample are clear, but the third factor is ambiguous.
Quantitative data of factor analysis showed the first two factors' priority. Accordingly, these factors explain 23.942 % and 11.537 % from the data variations, but the third factor - 5.641 %.
Analysing Fig. 4, two main factors are clear in thee Russian sample too, but the third factor looks ambiguous. Quantitative data of factor analysis showed the first two factors' priority. Accordingly, these factors explain 24.665 % and 8.669 % from data variations, but the third factor - 6.046 %.
Scree Plot
l l l l l l l l l l l l l l l l l l l l l l l l l
4 4 4 4 4
Component Number
Fig. 3. 50-item tolerance's scale Varimax rotation on 6 factors (Latvian sample)
Scree Plot
^ 7.5-
ra >
£
01 5?
UJ 5.0-
Fig. 4. 50-item tolerance scale Varimar rotation on 6 factors (Russian sample)
The first factor, characterizing negative attitudes andprejudices to ethnic ouC-group, is the strongest but tlie nert two positive factors are nrnr so strong and diffeaentiated.
According to tte (tudf's plan, two ways of construct validity verification were used: the comparison of data from two functioaallf
rontrasting psofessifnal gcoups chat could be potentially different also on social tolerance; and positive correlation with a similar measute (coavf rgant vdidizatiof.
Tre comparisor of two contrasting groups was made using ANOVA approach (see Table 3). The tf sufts of the Latvian sampte show tiaat only f88-
10.0
2.5
0.0
the difference on the second factor (positive stereotypes) is significant ( F = 5.16, p = 0.025), while other diferences are in the predicted direction but not significant. The small diferences can be explained by the interaction between professional, gender and age factors. In group 1.1. (social workers) the average age was twice higher (M = 42.8, SD = 9.4) than in group 1.2. (M = 21.9, SD = 1.8), and group 1.1. was female mainly.
The results of the Russian sample show that only the difference on the first factor (negative attitudes and stereotypes) is significant (F = 5.35, p = 0.023) ), while other diferences are in the predicted direction but not significant (see Table 4.). The small diferences can be explained by the interaction between professional and age factors. In group 1.1. (social workers) average age is twice higher (M = 43.54, SD = 8.89) than in group 1.2. (M = 24.39, SD = 4.83).
Taking into account the main matter of interest for intolerance 18-item scale My antipathies was chosen that had been previously elaborated and verified (Breslav, 2011; Breslavs, Tjumeneva, 2008). Responses to each of these
items are made on a 9-point Likert scale ranging from does not apply to me (1) to strongly applies to me (9). The passive hate subscale consists of eleven statements as "...activities make me anxious" and "Obviously feet get me far from ...", the active hate subscale consists of seven statements, as "I want to punish such people as. for public grievance" and "I became enraged when such people as .vaunt their achievements". The internal-consistency reliability of the Hate subscales on Latvian data showed good results too - Cronbach's Alpha 0.913 for passive hate and 0.827 for active hate for the Latvian-speaking sample, 0.915 & 0.853 for the Russian-speaking sample accordingly.
The study hypothesis on this stage presupposed the positive link between the first factor (negative attitudes +stereotypes) and hate, but negative links or lack of the link between hate and the second and third factors. The results (see Table 5. and Table 6.) confirmed this hypothesis. In the Latvian sample the first factor has positive correlation with both hate types (0.218 with passive hate and 0.240 with active hate), the
Table 3. 50-item scale variance (F Fisher) analysis results (Latvian sample, n=102)
ANOVA Latvian-50 F Sig.
Tolerance_fact1 Between Groups 0.0747 0.785177
Tol_fact2 Between Groups 5.157399 0.025291
Tol_fact3 Between Groups 2.493912 0.117445
Hate_sum Between Groups 0.954645 0.330898
Table 4. 50-scale variance analysis (F Fisera) results Russian sample (n=100)
ANOVA Russian version -50 F Sig.
Tolerance_fact1 Between Groups 5.355669 0.022743
Tol_fact2 Between Groups 0.065097 0.799149
Tol_fact3 Between Groups 2.039226 0.156467
Hate Between Groups 1.86805 0.174826
Table № 5. Ethnic Tolerance correlations with Hate Latvian Latvian sample (r Spearman)
Variables Intolerance - negative attitudes + stereotypes (1) Tolerance - positive stereotypes (2) Tolerance- positive attitudes (3)
1. Tolerance factor
2. factor - 0.119
3. factor - 0.679*** 0.330***
Hate-passive 0.218* -0.114 -0.256**
Hate-active 0.240** -0.029 -0.190*
* - p < 0.05; ** - p < 0.01, *** - p < 0.001
Table № 6. Etnic Tolerance correlations with Hate Russian Latvian sample (r Spearman)
Variables Intolerance - negative attitude + stereotypes (1) Tolerance - positive stereotypes (2) Tolerance- positive attitudes (3)
1. factor
2. factor - 0.406***
3. factor -0 594*** 0.596***
Hate-passive 0.207* 0.085 -0.075
Hate-active 0.190* -0.003 0.062
* - p < 0.05; ** - p < 0.01, *** - p < 0.001
second factor is not linked significantly, but the third factor has negative correlation with both types of hate (Table 5).
In the Russian sample the first factor has positive correlation with both hate types (0.207 with passive hate and 0.190 with active hate), but the second & third factors are not linked significantly with both types of hate (Table 6). The differences between the two samples can be explained by more homogenous data of the Russian sample on the second and third factors represented in high intercorrelations of these factors. The Latvian sample data show that positive stereotypes are higher than positive attitudes to the out-group and that intercorrelation is not significant.
The tolerance scale reliability was verified using two approaches: 1) internal consistency as a scale's homogeneity, verified with the most popular equation by Cronbach's a; 2) retesting
procedure as a scale's stability of data. The data on Cronbach's a showed good scale's reliability for the Latvian and Russian versions (see Table 7.). For the Russian version Cronbach's a is a little higher but both versions look strong enough.
The second approach to the tolerance scale's reliability verification through retesting also revealed good results (see Table 8). All correlations (r Spearman) are significant and sufficiently high, which shows the 50-item tolerance scale high reliability as stability of results, taking into account that the retesting procedure was realised approximately three months after the first assessment.
In figures 5a,b,c, all three factors testing-retesting correlation for the Latvian sample are represented. In figures 6a,b,c, - all three factors testing-retesting correlationfor the Russian sample are represented. All correlations are very high and significant.
Table № 7. 50-item tolerance scale's internal consistency (Cronbach's a)
Factors/Samples a Cronbach (Latvian sample) a Cronbach (Russian sample)
1 .908 .930
2 .846 .863
3 .740 .833
Table № 8. 50-item tolerance scale testing-retesting correlations
F actors/Versions ]^^"tviaLn(5'7 respondents) Russian (66 respondents)
1 .698*** 839***
2 .688*** .834***
3 723*** 898***
** * -p< 0.001
Korelaciju diagramma latviesu izlasei (1f) Korelaciju diagramma latviesu izlasei (2f) Korelaciju diagramma latviesu izlasei (3f)
100 90 80 70 = 60 ;s 50 2 40 ~ 30 20 10 *H 60 50 «40 £30 10 ♦ 45 40 35 «30 J 25 '120 ^¡15 10 5 * ♦
0 3200 20 40 60 80 100 i.testesana 10 20 30 40 50 60 i.testesana 0 5 10 15 20 25 30 35 40 45 i.testesana
(a) (b) (c)
Fig. 5 a,b,c. Testing-retesting tolerance scale Latvian version correlations for the first factor (a), second factor (b), and fhird factor (c)
Korelaciju diagramma krievu izlasei (if) Korelaciju diagramma krievu izlasei (2f) Korelaciju diagramma krievu izlasei (3f)
70 90 60
60 50 J 40 £30 sf * 80 70 0 50 £4 0 ♦ 50 „40 £30
c-^20 10 ♦ ♦ 20 10 10
1 2 3 4 5 6 7 8 i.testesana 10 20 30 40 50 60 70 80 90 1.teste?ana 0 10 20 30 40 50 60 I.testesana
(a) (b) (c)
Fig. 6 a,b,c. Testing-gefestrng Solerance scale Russian version's correlations tor the first factor (c), second factor (b), and third factor (c)
Results
The data have shown significant gender differences on the second factor (positive stereotypes) only (male - M = 35.67, SD = 4.76; female - M = 39.19, SD = 5.15) (see Table 9.). The diferences on the other factors are not significant.
Similar results are revealed in the Russian sample where no significant gender diferences are found at all (See Table 10).
The comparison of professional groups has shown significant diferences in the Latvian sample on the second factor only (F = 5.157, p = .025) (See Table 11). The lack of other diferences could be explained by big age diferences between groups 1.1. (M = 42.8, SD = 9.4) and 1.2. (M = 21.9, SD = 1.8).
The comparison of professional groups has shown significant diferences in the Russian
Table 9. 50-item tolerance scale's data (Latvian sample)
Factors Male (n=24) Female (n=78)
l.factor M 73.17 72.97
SD 12.06 12.4
2.factor M 35.67** 39.19
SD 4.76 5.15
3.factor M 29.92 31.38
SD 5.03 4.96
Table 10. 50-item tolerance scale's data (Russian sample)
Factors Male (n=51) Female (n=49)
1.factor M 49.37 51.35
SD 10.82 12.67
2.factor M 58.57 57.39
SD 8.57 9.69
3.factor M 34.29 33.02
SD 6.77 6.54
Table 11. 50-item tolerance scale's data on Professional comparison (Latvian sample)
Factors Social workers and medical personnel (n=45) Policmen, lawers and military personnel (n=57)
1.factor M 72.64 73.32
SD 11.92 12.62
2.factor M 39.67* 37.33
SD 5.33 5.01
3.factor M 31.91 31.91
SD 4.07 4.07
Table 12. 50-item tolerance scale's data on professional comparison (Russian sample)
Factors Social workers and medical personnel (n=39) Policmen, lawers and military personnel (n=61)
l.faktors M 53.67* 48.21
SD 13.33 10.16
2.faktors M 58.28 57.80
SD 10.06 8.52
3.faktors M 32.49 34.43
SD 7.14 6.28
sample on the first factor only (F = 5.356, p = .023) (See Table 12).
The lack of other diferences could also be explained by big age diferences between groups 1.1. (M = 43.5, SD = 8.9) and 1.2. (M = 24.4, SD = 4.8).
Discussion
The first stages of the new measure's elaboration have shown the possibility to assess the tolerance construct in its social-psychological aspect. On the initial stages of the measure's elaboration the main themes of out-group perception of Russian-speaking Latvians by Latvian Latvians and of Latvians by Russians. As expected, these themes were not the same for Latvian and Russian participants. For example, Latvians were more sensitive to language issues but Russian-speaking participants to discrimination issues. To minimize the diferences between two versions of the scale the same structure and the same number of items were elaborated. It means the necessity to use assymetric criteria in the process of weak items' transformation, selection and rejection.
Statements were shaped in coherence with these themes content and previous studies on intergroup relations, especially with the Integrated Threat theory (Stephan & Stephan, 1996). Elaboration of the first six factors' 89-item tolerance scale with six subscales (realistic
threat, symbolic threat, intergroup anxiety and negative stereotypes, positive stereotypes and positive attitude) and its verification showed non-coherence of the main factors with the factors proposed by the ITT. Factor analysis results of the 89-item tolerance scale showed that it is impossible to differentiate ethnic out-group anxiety, ethnic out-group threats and negative stereotypes about ethnic out-group (Breslavs, et al., 2008).
In the next version of the scale these factors were combined into one factor - negative attitudes. On the next stage of data collection and the scale's verification six factors were reduced to three factors' (negative attitudes to ethnic outgroup + negative stereotypes; positive stereotypes; and positive attitudes to outgroup) 50-item scale with three subscales. Factor analysis results of the 50-item tolerance scale showed that the first factor is strong enough for both versions, but the second and third factors are not so strong, which leaves the opportunity to minimize the scale's structure to two main factors (negative and positive). The opportunity to minimize this scale to 38-item two-factor tolerance scale should be verified in further studies.
The results of both versions the 50-item scale's reliability look good enough including good internal consistency (the Latvian version -a = .908, .846, .740 for accordingly, the first,
second and third factors; the Russian version -a =.930, .863, .833) and high testing-retesting correlation (from .688 to .898). This tolerance version showed positive correlation of the negative subscale with the hate scale but negative correlation (or lack of correlation) of the positive subscales with the hate scale that had been expected by the test's elaborators. The comparison of two contrast professional samples looks not so decisive, which can be explained by the big age difference between selected professional groups. The verification of construct validity could be continued to get more robust results with the right authoritarian political orientation and the SDO measures (Altemeyer, 1996; Pratto, et al., 1994) as well as aggression measure (Buss, & Perry, 1992).
Conclusion
Results of the 50-item tolerance scale elaboration look quite promising.
The data have shown good reliability (internal consistency and retesting stability) and moderate construct validity of the 50-item scale's two Latvian and Russian versions. But taking into account different intercorrelations between the second and third factors in the Latvian and Russian versions, the higher stability of two-factor structure (negative attitudes + negative stereotypes - positive attitudes + positive stereotypes), the tolerance scale could be reduced to the 38-item scale with two subscales.
At the same time, the concept of tolerance in its social-psychological sense seems to be insufficiently elaborated theoretically. Many related concepts, such as stereotypes, prejudices, social discrimination and derogation should be revised and placed into a theoretical model for understanding the processes of intergroup relationships, taking into account individual variables that can improve or disturb these processes.
References
1. Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The authoritarian personality. New York: Harper & Brothers.
2. Akrami, N., Ekehammar, B. & Araya, T. (2000). Classical and modern racial prejudice: A study of attitudes toward immigrants in Sweden. European Journal of Social Psychology, 30, 521-532.
3. Altemeyer, B. (1988). Enemies of freedom: Understanding right-wing authoritarianism. San Francisco: Jossey-Bass.
4. Altemeyer, B. (1996). The authoritarian specter. Cambridge, MA: Harvard University Press.
5. APA Dictionary of Psychology, G.R. VandenBos (Ed.), (2007).Washington, DC: American Psychological Association.
6. Barrett, M. & Oppenheimer, L. (2011). Findings, theories and methods in the study ofchildren's national identifications and national attitudes. European Journal of Developmental Psychology, 8: 1, 5-24.
7. Berry, J. W; Poortinga, Y. H; Segall, M. H; Dasen, P. R. (2002). Cross-cultural psychology: Research and applications (2nd ed.). Cambridge: Cambridge University Press.
8. Berry, J.W., Trimble, J.E.& Olmedo, E.L. (1986). Assessment of acculturation. In W.J. Lonner, & J.W Berry (Eds.), Field methods in cross-cultural research. (pp.290-327).Beverly Hills, CA: Sage.
9. Breslavs, G., Abele, U., Derjabo, L., Pisinska, I. & Roze, J. (2008). Vai mes esam iecietigi ? Sabiedribas etniskäs iecietibas monitoringa metodikas izveide (Are we tolerant? Elaboration of ethnic tolerance assessment scale for societal monitoring). Riga: „Salus".
10. Breslav, G. (2011). Nenavistj kak predmet psihologicheskogo issledovanija (Hatred as the subject-matter of psychological study). Voprosy Psyhologii, # 2, 138-148.
11. Breslavs, G., Tyumeneva, J. (2008). Is Hate an opposite to Love? The paper presented at the 14th European conference on Personality, Tartu, July 16-20, 2008.
12. Brewer, M.B. (1991). The social self: On being the same and different at the same time. Personality and Social Psychology Bulletin, 17, 475-482.
13. Brewer, M. B. (1999). The psychology of prejudice: Ingroup love or outgroup hate? Journal of Social Issues, 55(3), 429-444.
14. Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality and Social Psychology, 63, 452-459.
15. Ekehammar, B., Akrami, N. & Fan Y.-W. (2009). Ethnic Prejudice: A Combined Personality and Social Psychology Model. Individual Differences Research, Vol. 7 (4), 255-264.
16. Ekehammar, B., Akrami, N., Gylje, M. & Zakrisson, I. (2004).What matters most to prejudice: Big Five personality, Social Dominance Orientation, or Right-Wing Authoritarianism? European Journal of Personality, Vol. 18 (6), 463-482.
17. Fazio, R. H., Jackson, J. R., Dunton, B. C., Williams, C. J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline. Journal of Personality and Social Psychology, 69, 1013-1027.
18. Graumann, C. F. (1992, May). Xenophobia: A challenge for social psychology. Invited lecture presented at the 9th East-West Meeting of the European Association of Experimental Social Psychology in Munster, Germany.
19. Guimond, S., Dambrun, M., Michinov, N., & Duarte, S. (2003). Does social dominance generate prejudice? Integrating individual and contextual determinants of intergroup cognitions. Journal of Personality and Social Psychology, 84, 697-721.
20. Martin, J.G. (1964). The tolerant personality. Detroit, MI: Wayne State University Press.
21. Mikulincer, M. & Shaver, P.R. (2007). Boosting Attachment Security to Promote Mental Health, Prosocial Values, and Inter-Group Tolerance. Psychological Inquiry, v.18 (3), 139-156.
22. Mummendey, A., & Wenzel, M. (1999). Social discrimination and tolerance in intergroup relations: Reactions to intergroup difference. Personality and Social Psychology Review, 3, 158-174.
23. Pratto, F., Sidanius, J., Stallworth, L. M., Malle, B. F. (1994). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67, 741-763.
24. Prislin, R. & Filson, J. (2009). Seeking Conversion Versus Advocating Tolerance in the Pursuit of Social Change. Journal of Personality & Social Psychology, Vol. 97 (5), 811-822.
25. Reynolds, K. J., Turner, J. C., Haslam, S. A., & Ryan, M. K. (2001). The role of personality and group factors in explaining prejudice. Journal of Experimental Social Psychology, 37, 427-434.
26. Stephan, W. G., & Stephan, C. (1996). Predicting prejudice. International Journal of Intercultural Relations, 20 (3/4), 409-426.
27. Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy and oppression. Cambridge, England: Cambridge University Press.
28. Thibaut, J.W. & Kelley, H. H. (1959). The Social Psychology of Groups. New York:Wiley.
29. Whitley, B. E. (1999). Right-wing authoritarianism, social dominance orientation, and prejudice. Journal of Personality and Social Psychology, 77, 126-134.
Разработка шкалы этнической толерантности: восстановление интегрального подхода
Г. Бреслав
Балтийский институт психологии и менеджмента Латвия, 1011, Рига, ул. Бруниниеку, 65
Несмотря на значительный прогресс в социальной психологии в последние 50 лет, понятие «социальная толерантность» остается весьма туманным. Для исследователей оно выглядит слишком комплексным, и они предпочитают анализировать отдельные аспекты этой толерантности - стереотипы, предрассудки, искажения восприятия, социальную дискриминацию, которые выглядят более одномерными и доступными для изучения. В то же время представляется, что без сведений о терпимости/нетерпимости к разным социальным группам в целом невозможно целенаправленно строить социальную политику, направленную на гуманистическую интеграцию общества. В психологии личности и социальной психологии на сегодняшний день отсутствуют интегрированные критерии для оценки, в частности этнической толерантности, и известные одномерные и частные показатели такой толерантности не могут компенсировать дефицит в этой области.
Целью данной статьи была разработка методики целостной диагностики межэтнических установок как личностной диспозиции в Латвии. Разработка такой методики представляла процесс из пяти этапов, в результате чего создана валидная и надежная шкала измерения этнической толерантности. Данная модель разработки шкалы толерантности может быть использована в любой стране с учетом конкретного социально-культурного контекста.
Ключевые слова: социальная толерантность, этническая толерантность, этнические установки, этнические стереотипы, этнические предрассудки, социальная дискриминация, шкала этнической толерантности.
Исследование проводилось в рамках проекта "Разработка метода мониторинга социальной толерантности" при содействии отделения ЕС по грантовой поддержке "Развитие социальной интеграции в Латвии" (Соглашение # 2005/017-495-03-01/3-1/50).
