Chapter 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups in Mexico and Chile Anja Eller, Huseyin Cakal, and David Sirlopu This chapter explores the relationships between different levels and different categories of social identity, intergroup contact, and various health-related measures among a sample of mestizos (i.e. people of combined European and Native American descent) and a different sample of indigenous people in Mexico as well as indigenous and non-indigenous samples in Chile. In theoretical terms, we build on two connected theories: social identity theory (SIT; Tajfel & Turner, 1979) and self-categorisation theory (SCT; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). SIT posited a distinction between personal and social identity. People derive the latter primarily from group memberships (e.g. gender, ethnic group, nationality) and generally strive for a positive social identity, which can be achieved by means of favourable comparisons with relevant out-groups. In certain cases, this in-group bias can lead to prejudice, stereotyping, infrahumanisation, and discrimination against out-groups. Relatedly, SCT proposes categorisation of the self at different levels of abstraction (personal, group, humanity, etc.) and points out that the group and its norms become a frame of reference of how to behave for its members. In Mexico and Chile important identities are based on ethnicity (mestizo or nonindigenous Chilean vs. indigenous; subordinate level) and nationality (Mexican or Chilean; superordinate level). A more inclusive superordinate level of identification A. Eller (*) National Autonomous University of Mexico, Mexico City, Mexico e-mail: eller@unam.mx H. Cakal University of Exeter, Exeter, UK D. Sirlopu Universidad del Desarrollo, Santiago, Chile © Springer International Publishing Switzerland 2016 S. McKeown et al. (eds.), Understanding Peace and Conflict Through Social Identity Theory, Peace Psychology Book Series, DOI 10.1007/978-3-319-29869-6_19 295 296 A. Eller et al. would be Latin American, which is often invoked as a differentiation from the USA and Canada and is also utilised. We should also note that in both Mexico and Chile ethnicity is confounded with socio-economic class: indigenous people tend to be at the bottom of the social ladder while mestizos and non-indigenous Chileans, respectively, are situated higher up. Jetten, Haslam, Haslam, Dingle, and Jones (2014) argue that traditionally health outcomes have been studied at the individual level and group memberships have been left out of the equation. However, groups and social identities are very important in the realm of health—for example, people get sick in groups (say, on a cruise ship) and they get well in groups (as patients on a particular hospital ward). More recently, the link between social identity and relationships on health—something that Jetten, Haslam, and Haslam (2011) refer to as the social cure—has been recognised by a growing number of researchers, such that there is now a burgeoning literature. Haslam, Jetten, Postmes, and Haslam (2009) compiled a list of articles, from 1970 onwards, whose titles, abstracts, or keywords jointly reference “social/ organisational/ethnic identity/identification” and “health and/or well(-)being”. The trend line shows a logarithmic increase in the number of publications that include those terms while there was a quadratic increase in the number of times these publications have themselves been cited. To provide some examples of this novel literature, social identification has been found to buffer individuals from the negative impact of a range of stressors, including illness (Haslam, Jetten, & Waghorn, 2009; Muldoon, Lowe & Schmid, this volume). Large, representative community samples demonstrated social identification to predict life satisfaction and general well-being (Helliwell & Barrington-Leigh, 2011). A review study by Cruwys, Haslam, Dingle, Haslam, and Jetten (2014) has shown social connectedness to be implicated in the development, progression, and treatment of depression. St Claire and He (2009) demonstrated that older adults who self-categorise as “elderly people” are more likely to think that they suffer from hearing loss and require a hearing aid, independent from objective measures of their actual hearing acuity. Social identification has even been shown to affect the progression of HIV in gay men. Disease progression was significantly faster among those gay men who were sensitive to rejection because they were not able to maintain certain social identities (Cole, Kemeny, Taylor, Visscher, & Fahey, 1996). In an educational context, Bizumic, Reynolds, Turner, Bromhead, and Subasic (2009) observed that teachers’ and students’ identification with their school strongly predicted participants’ decreased anxiety and depression. Finally, a longitudinal study during a religious mass gathering (Prayag Magh Mela) in India found that shared identity with other pilgrims impacted positively on self-assessed health (Khan et al., 2014). Taken together, this growing body of evidence already is impressive, and it is likely to expand considerably in the coming years. What has been left rather unexamined, however (but see Jasinskaja-Lahti, Liebkind, & Solheim, 2009), is the interplay of different levels of categorisation/identification on health-related variables. As outlined in SCT, categorisation is a malleable and context-dependent process and daily social life is sufficiently complex as to allow for two (or more) simultaneous salient categories and identification therewith. In the present research, we investigate main and interactive effects of subordinate (ethnic) and superordinate (national) identification on health. 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 297 What is more, the literature employs terms relating to social identity and concepts relating to social contact interchangeably. However, it is very important to clearly define these concepts and tease them apart such that it emerges whether it is the identification with a social group or category or the contact with its members or a combination of both that affects health-related outcomes. In this vein, Jetten et al. (2014) have defined the key concepts of social connectedness, social networks, social relations, social integration, social isolation, and social support. In isolated studies, social identity and social contact have been differentiated. For instance, Sani and colleagues (Sani, Herrera, Wakefield, Boroch, & Gulyas, 2012) conducted two studies with different social groups (i.e. the family and an army unit) and found that group identification was better than social contact at predicting mental health. In contrast, a meta-analysis on more than 300,000 people, followed for an average of 7.5 years, showed that people with adequate social relationships (i.e. contact) had a 50 % great likelihood of survival compared to those with poor or insufficient relationships. Remarkably, the magnitude of this effect compares to that of smoking cessation, and it is larger than many other established risk factors for mortality (e.g. obesity, physical inactivity; Holt-Lunstad, Smith, & Layton, 2010). Thus, it emerges very clearly that both social identity and social contact play pivotal roles in health-related outcomes. In our research, we conceptualise these terms as independent variables, such that their separate and joint impact on healthrelated variables can be assessed. Moreover, we focus on intergroup (as opposed to intragroup, as in the literature discussed above) contact. Social groups and categories, based on nationality, ethnicity, political party affiliation, gender, sexual preference, university affiliation and many other criteria, are omnipresent in daily life. Additionally, a globalised world heightens people’s mobility and makes contact across group boundaries more and more common and likely. Intergroup contact theory (Allport, 1954) is probably the most widely used technique for reducing prejudice and improving intergroup relations. It stipulates that intergroup contact needs to be qualified by certain conditions, such as equal status in the situation and friendship potential, in order to have beneficial effects on intergroup relations. During the past 60 years, hundreds of studies have attested to the theory’s effectiveness, using different methodologies, participant populations, bases of group membership, etc. (Brown & Hewstone, 2005; Eller & Abrams, 2003, 2004; Pettigrew & Tropp, 2006). More recently, it has been suggested that indirect contact is also effective at reducing prejudice and improving intergroup relations. The extended contact hypothesis (Wright, Aron, McLaughlin-Volpe, & Ropp, 1997) proposes that the mere awareness that an in-group member has an out-group friend can improve intergroup relations. Associations between extended contact and lower prejudice have been observed among (racial) majority and minority group participants in the USA (Wright et al., 1997), Catholics and Protestants in Northern Ireland (Paolini, Hewstone, Cairns, & Voci, 2004), friends of international exchange students (Eller, Abrams, & Zimmermann, 2011), and other intergroup contexts (see Dovidio, Eller, & Hewstone, 2011). Taking together the direct and extended contact literature, while the protective influence of relationships within a social group on health is well documented (HoltLunstad et al., 2010), it has hitherto been neglected whether the beneficial effects of direct and extended intergroup contact can expand to health-related variables. 298 A. Eller et al. Context of the Present Research Both Latin American countries of Mexico and Chile have a colonial past and are home to indigenous groups albeit at different percentages of their population. According to recent census results, in Mexico 60 % of the overall population identify themselves as of mestizo descent while 30 % of the population identify as indigenous. In Chile, however, 88 % of the overall population identify as of White European descent while only 11 % identify as indigenous (Cakal, Eller, Sirlopú, & Pérez, under review; Van Cott, 2007). Straightforward interpretation of these numbers, however, is problematic as in both countries the census defines “indigenous” by means of people’s self-categorisation, and this might be a problem because of the lower status of indigenous people in both countries and hence, the unwillingness of indigenous or mestizo people to identify as such. As Ruiz-Linares et al. (2014) argue, the actual distribution of indigenous versus European-descent (or mestizo) individuals in Chile might be closer to an equal distribution of both ethnicities. In Mexico, ever since colonial times, relations between the numerical majority of mestizos and the minority of indigenous people have been ambivalent at best and fraught with difficulties, discrimination, and violence at worst. Despite a national sense of pride in their Native American history and notwithstanding the fact that more than 80 % of the Mexican population has some indigenous physical features, there is widespread prejudice, stereotypes, and discrimination on the part of mestizos toward indigenous people. According to the National Discrimination Survey (ENADIS, 2010), a quarter of people living in Mexico City openly discriminate against indigenous people. Being the target of social exclusion and discrimination can have serious consequences on psychological and even physical health, as the literatures on stigmatisation, bullying, and ostracism have clearly shown (e.g. Quellet-Morin et al., 2013). Chile’s largest ethnic group are the Mapuche. Current statistics shows that 11 % of Chileans consider themselves to be part of one of the 11 recognised ethnic groups in Chile (CEPAL, 2012). Of these, the vast majority is identified with being Mapuche (84 %). The nature of the intergroup relationship between White Chileans and Mapuche has not changed fundamentally. For instance, Merino and Quilaqueo (2003) found that a vast majority of non-indigenous Chileans show prejudice and hold negative stereotypes toward Mapuche in their everyday discourse. Moreover, the state has been increasingly less tolerant toward the Mapuche. For instance, a Mapuche individual who is presumed guilty of attacks on non-indigenous Chilean property or another violent acts can be punished with strong legal penalties under the provision of Anti-Terrorist Law passed during the military regime (De la Maza, 2014). We believe that it is particularly important to study such under-researched populations as indigenous people. They rank as lowest in quality of life as indicated by much higher rates of poverty, exclusion, and child mortality rates than the mainstream societies (Hall & Patrinos, 2014). By studying indigenous populations, we also respond to recent criticism on the perils of studying WEIRD (Western, educated, and from industrialized, rich, and democratic countries) populations and generalising to other non-WEIRD populations (Henrich, Heine, & Norenzayan, 2010). Study 1 consisted of a sample of mestizos 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 299 and a sample of indigenous participants, responding to the superordinate category Mexican. Study 2 involved a sample of non-indigenous Chileans (who could be White or mestizo; see difficulties of categorisation discussed above) and a sample of indigenous people responding to the superordinate category Chilean and another sample of nonindigenous Chileans responding to the superordinate category Latin Americans. This was done in order to assess the effects of superordinate categories that differ in level of inclusiveness (cf. Turner et al., 1987). Hypotheses In light of the positive effects of identification, on the one hand, and social contact, on the other, on psychological and physical health, as outlined above, our global prediction is that subordinate and superordinate identification as well as direct and extended intergroup contact will be associated with better current health, improved health over the previous year, better psychological health, and lower incidence of stress-related diseases. If we were to make more refined predictions, we would hypothesise that the impact of intergroup contact on improved health would be weaker for minority than majority groups (Tropp & Pettigrew, 2005). Building on Sani’s (2011) argument that curative relationships are built on trust, cooperation, and mutual support, we would also expect contact with other indigenous groups to be more beneficial than contact with the majority groups for indigenous participants. Moreover, it is likely that subordinate identification as indigenous will have more (positive) impact on health-related variables for minority group participants while superordinate identification on the national or perhaps even regional level (Study 2) will have stronger effects for majority group participants (see Staerklé, Sidanius, Green, & Molina, 2010). Study 1: Mexico Method This questionnaire-based study consisted of two samples in Mexico City, one of 237 self-categorised mestizos (44.1 % men; mean age 21.5 years) and another sample of 152 self-categorised indigenous people (37.1 % men; mean age 37.1 years). Participation was voluntary; mestizos were recruited via social media websites, completed the questionnaire online, and were entered into a draw for several monetary awards. Indigenous participants were recruited from indigenous communities in Mexico City, interviewed by a research assistant, and received some monetary compensation for their time. Unless otherwise indicated, variables were measured by five-point Likert-type scales. In terms of predictor variables, participants were asked about their direct 300 A. Eller et al. contact with the respective other group: contact as friends (number); quantity of contact in daily life, visits to people at their homes, via the internet (1 = never—5 = all the time; cf. Cakal, Hewstone, Schwar, & Heath, 2011; Eller, Abrams, & Gomez, 2012); and quality of contact (cooperative, positive, respectful, important for you, makes you feel accepted, enjoyable; 1 = little—5 = much; Eller et al., 2011). These measures were combined into a single direct-contact score, which loaded on two factors for mestizos, explaining 39.3 % and 19.1 % of the variance, respectively (Cronbach’s alpha = 0.79) and two factors, explaining 43.7 % and 17.7 % of the variance, respectively, for indigenous people (α = 0.85). We also asked participants about their extended contact with the respective outgroup (cf. Wright et al., 1997): how many in-group members they knew who had close out-group friends as well as whether the relationship between the in-group member and their out-group friend was cooperative, positive, respectful, important for the person, made them feel accepted, and enjoyable (1 = little—5 = much). These measures were combined into a single extended-contact score, which loaded on one factor for mestizos, explaining 68.4 % of the variance (α = 0.90) and one factor, explaining 69.2 % of the variance, for indigenous people (α = 0.91). In the case of indigenous participants only, they were additionally asked about their quantity and quality of contact with other indigenous groups in Mexico, using equivalent measures to the ones reported above. These measures were combined into a single other-indigenous-contact score, which loaded on two factors, explaining 46.7 % and 21.6 % of the variance, respectively (α = 0.87). In terms of moderator variables, we further asked our respondents about their level of identification with the subordinate category (mestizo or indigenous, fouritem scale, five-point Likert-type scale from do not agree at all to agree completely) and the superordinate category (Mexican), for example, “I’m proud to be mestizo [Mexican]” (Eller & Abrams, 2003). These measures were combined into a single subordinate identity score, which loaded on one factor for mestizos, explaining 72.6 % of the variance (α = 0.87) and one factor, explaining 79.1 % of the variance, for indigenous people (α = 0.90) as well as a single superordinate identity score, which loaded on one factor for mestizos, explaining 84.4 % of the variance (α = 0.94) and one factor, explaining 87.7 % of the variance, for indigenous people (α = 0.95). Finally, in terms of outcome variables, participants rated their current health (1 = bad—5 = excellent); their current health compared to that a year earlier (five-point Likert-type scale, from much better than a year ago to much worse than a year ago, reverse-scored); their psychological health (5-items, 1 = never—5 = always, e.g. during the past year, how often have you felt anxious and nervous? How often have you felt disenchanted and sad?; Goldberg et al., 1997); and whether they had suffered any of a seven stress-related conditions (back pain, insomnia, headaches, colds, etc.) during the past year. The measure of psychological health was combined into a single score, which loaded on two factors for mestizos, explaining 40.4 % and 29.7 % of the variance (α = 0.71) and one factor for indigenous people, explaining 67.3 % of the variance (α = 0.75). For the indigenous sample, two items were excluded from the final score because they reduced scale reliability substantially. The number of stressrelated conditions checked by participants was added up to arrive at a single score. 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 301 Results and Discussion Descriptive statistics are shown in Table 19.1. For mestizo participants, direct contact with indigenous people was just below the mean while extended contact was comparatively high. Both types of identity were above the mean, but mestizos’ subordinate identity was lower than the superordinate national identity. Current health, health compared to a year ago, and psychological health were all above the mean for mestizos but still not especially high. Finally, the number of stress-related diseases was relatively low at 2.62 out of a maximum of 7. For indigenous participants, all three contact congregate measures were above the mean and direct and extended intergroup friendships were very high. Both subordinate and superordinate identities scored very high, as well. Indigenous participants rated their current health to be rather poor; however, health compared to a year earlier and psychological health were both above the mean and the number of stress-related illnesses was rather low. After mean centring direct and extended contact (Aiken & West, 1991), we used simultaneous multiple regression analysis to test the effects of direct and extended contact, subordinate and superordinate identity, and their interactions on the four health-related variables. In the mestizo sample (see Fig. 19.1), direct contact with indigenous people was associated with less psychological health, β = −0.03, t(234) = −1.96, p = 0.051. Higher subordinate identity was related to a deterioration of health compared to a year earlier, β = −0.06, t(234) = −2.99, p = 0.003, while Table 19.1 Means and standard deviations for mestizos and indigenous participants in Mexico (Study 1) Measure Out-group friendsa Direct out-group contactb Out-group friends of IG membersc Extended out-group contactb Contact w/other indigenous groups Subordinate identityb Superordinate identityb Current health ratingb Health compared to 1 year agob Psychological healthb Stress-related diseasesc Mestizos 1.02 (4.42) 2.89 (0.73) 1.92 (5.71) 3.95 (1.01) – 3.62 (1.08) 4.16 (1.03) 3.24 (0.95) 3.43 (0.97) 3.51 (0.73) 2.62 (1.64) Indigenous 9.57 (16.31) 3.48 (0.96) 14.55 (28.43) 3.72 (1.21) 3.66 (0.92) 4.60 (0.79) 4.58 (0.91) 2.77 (0.87) 3.48 (0.99) 3.40 (1.18) 2.29 (1.72) Note: IG in-group. Number, open-ended b Five-point scale c Number between 0 and 7 Standard deviations are presented in parentheses. Out-group friends and direct out-group contact are presented separately; for further analysis, these variables were z-transformed and combined. Out-group friends of in-group members and extended out-group contact are presented separately; for further analysis, these variables were z-transformed and combined a A. Eller et al. 302 Direct contact Health -.03+ Extended contact -.06** Health compared to 1 year ago subordinate ID low: -.07, ns subordinate ID high: -.25** Subordinate identity Psychological health .04* Superordinate identity .05** Stress-related diseases Fig. 19.1 Path diagram to show the results of regression analyses for mestizo participants in Mexico (Study 1). Note: Solid lines represent significant paths; interrupted lines represent interactions; numbers are standardised partial regression coefficients (ß). +p<0.06; *p <0.05; **p < 0.01; ***p < 0.001 higher superordinate identity was associated with improved health over time, β = 0.04, t(234) = 2.04, p = 0.04, as well as with better psychological health, β = 0.05, t(234) = 3.09, p = 0.002. We also found a significant interaction between extended contact and subordinate identification on psychological health, β = −0.04, t(234) = −2.08, p < 0.04. Analyses of simple slopes (Cohen, Cohen, West, & Aiken, 2003) showed that extended contact was significantly associated with decreased psychological health when participants’ subordinate identity was high, β = −0.25, t(121) = −2.84, p = 0.005, but not when it was low, β = −0.07, t(114) = −0.70, p = 0.49 (see Fig. 19.1). In the indigenous sample (see Fig. 19.2), direct contact with mestizos was associated with improved health over the previous year, β = 0.07, t(134) = 2.15, p = 0.03, and with a decrease in stress-related diseases, β = −0.17, t(134) = −2.22, p < 0.03. Contact with other indigenous groups was related to improved health over the previous year, β = 0.21, t(134) = 2.06, p = 0.04, as well as to lower psychological health, β = −0.42, t(134) = −3.29, p < 0.001. A higher superordinate, Mexican identity was associated with lower ratings of current health, β = −0.08, t(134) = −2.74, p = 0.007, but also with higher psychological health, β = 0.07, t(134) = 1.99, p < 0.05. We also found a significant interaction between extended contact and superordinate identification on stress-related diseases, β = −0.23, t(134) = 2.15, p = 0.05. The association between extended contact and stress-related diseases was not significant when superordinate identity was low (β = −0.12, t(134) = −0.69, p = 0.48), but it was significant and negative when superordinate identity was high (1SD; β = −0.47, t(134) = −2.98, p = 0.003). 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… Direct contact Health .07* -.08** Extended contact Contact with other indigen. groups Health compared to 1 year ago .21* -.42*** Psychological health Subordinate identity Superordinate identity 303 superord. ID low: -.12 ns .07* superord. ID high: -.47** -.17* Stress-related diseases Fig. 19.2 Path diagram to show the results of regression analyses for indigenous participants in Mexico (Study 1). Note: Solid lines represent significant paths; interrupted lines represent interactions; numbers are standardised partial regression coefficients (ß). *p < 0.05; **p < 0.01; ***p < 0.001 The relation between contact with other indigenous groups on improved health over the previous year was qualified by a three-way interaction: contact × superordinate identity × subordinate identity, β = 0.09, t(139) = 1.94, p = 0.049. More specifically, more contact with other indigenous people and stronger identification as Mexican was associated with poorer health when indigenous identity was low (−1SD; β = −0.15, t(139) = −1.92, p = 0.046). This association, however, became non-significant when indigenous identity was high (1SD, β = 0.001, t(139) = −0.086, p = 0.93). Unpacking this three-way interaction, we observed that contact with other indigenous people positively predicted improved health over previous year when superordinate identity as Mexican was low (−1SD; β = 0.22, t(140) = 2.41, p = 0.02) but not when superordinate identity was high (1SD; β = 0.15, t(140) = 1.59, p = 0.11). In summary, direct out-group contact had negative effects on (psychological) health among mestizos while it had positive effects on (physical) health among indigenous participants. For these participants, contact with other indigenous groups was associated with positive physical but negative psychological outcomes. Moreover, this contact had negative effects on physical health at low subordinate identification, and it had positive effects on health at low superordinate identification. For mestizos, health over the previous year was predicted positively by superordinate identification but negatively by subordinate identification. For indigenous participants, superordinate identification was related to lower physical but higher psychological health. Finally, extended contact had beneficial outcomes for mestizos when subordinate identification was low and for indigenous participants when superordinate identification was high. 304 A. Eller et al. Study 2: Chile Method This questionnaire-based study consisted of three samples. Two samples, composed of non-indigenous Chileans, were recruited through universities from Concepción, a city located in the south of Chile. One of them comprised 209 Chileans (64.6 % men; mean age 19.5 years) responding to Chile as the superordinate identity. The another comprised 205 Chileans (71.1 % men; mean age 20.9 years) responding to Latin America as superordinate identity. Finally, the Mapuche sample comprised 180 participants of different professional occupations (61.4 % men; mean age 36.7 years), responding to Chile as the superordinate category. All participants of the three samples completed anonymous questionnaires and decided to participate voluntarily. Non-indigenous Chilean people were recruited in a university classroom settings by four research assistants. Mapuche people were recruited from some indigenous communities located in rural zones of Concepción, and a research assistant with Mapuche origins interviewed them. All variables were equivalent to Study 1. Direct and extended contact excluded contact as (extended) friends because these variables decreased scale reliability substantially. The direct-contact score loaded on two factors for the first sample of non-indigenous Chileans, explaining 54.4 % and 20.3 % of the variance, respectively (Cronbach’s alpha = 0.85), three factors for the second sample, explaining 34.9 %, 24.7 %, and 12 % of the variance, respectively (α = 0.79), and two factors (40.7 % and 22.9 %) for indigenous participants (α = 0.81). The extended-contact score loaded on one factor for the first sample of non-indigenous Chileans, explaining 76.5 % of the variance (α = 0.93), on two factors (70.9 % and 14.4 %) for the second sample (α = 0.90), and one factor, explaining 70.0 % of the variance, for indigenous people (α = 0.91). Contact with other indigenous groups in Chile (only indigenous participants) loaded on two factors, explaining 52.8 % and 23.5 % of the variance, respectively (α = 0.87). The subordinate identity score (as mestizo or indigenous) loaded on one factor for the first sample of non-indigenous Chileans, explaining 79.9 % of the variance (α = 0.92), on one factor (78.5 %) for the second sample (α = 0.91), and one factor, explaining 79.8 % of the variance, for indigenous people (α = 0.90). The superordinate identity score (as Chilean or Latin American) loaded on one factor for the first sample of non-indigenous Chileans, explaining 83.9 % of the variance (α = 0.94), on one factor (83 %) for the second sample (α = 0.93), and one factor, explaining 88.0 % of the variance, for indigenous people (α = 0.95). Stress-related conditions included diabetes and gastrointestinal disorders for the first sample; heart problems, asthma, colds, gastrointestinal disorders, and selfdefined “other” diseases for the second sample, and colds and cancer for the third sample. The measure of psychological health loaded on two factors for the first sample, explaining 40.1 % and 30.3 % of the variance (α = 0.64), one factor (73.5 %) for the second sample (α = 0.64), and one factor for indigenous people, explaining 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 305 61.8 % of the variance (α = 0.69). For the second sample, three items were excluded from the final score and for the third sample two items were excluded because they reduced scale reliability substantially. Results and Discussion Descriptive statistics are shown in Table 19.2. For the mestizo samples, direct contact with indigenous people was just below the mean while extended contact was comparatively higher. Both types of identity were above the mean, but for the first sample, curiously, non-indigenous Chileans’ subordinate identity was higher than the superordinate national identity (for the second sample, it was the other way around). Current health was at the mean of the scale while health compared to a year ago and psychological health were both above the mean for non-indigenous Chileans but still not especially high. Finally, in the first sample 5 % of participants had diabetes whereas 35 % had gastrointestinal problems. For indigenous participants, all three contact congregate measures were above the mean and direct and extended intergroup friendships were very high. Both subordinate and superordinate identities scored very high, as well. Current health was at the scale mean and health compared to a year earlier was substantially better, as was psychological health. Table 19.2 Means and standard deviations for indigenous and non-indigenous Chileans (Study 2) Measure Direct out-group contacta Extended out-group contacta Contact w/other indigenous groupsa Subordinate identitya Superordinate identitya Current health ratinga Health compared to 1 year agoa Psychological healtha Stress-related diseasesb Diabetesc Gastrointestinal disordersc Coldsc Cancerc NI Chileans (first sample) 2.94 (0.80) 3.76 (1.19) – NI Chileans (second sample) 2.86 (0.72) 3.39 (1.12) – Indigenous 4.43 (0.63) 4.24 (0.87) 3.69 (0.87) 3.75 (1.08) 3.46 (1.16) 3.01 (0.99) 3.70 (0.98) 3.63 (0.70) – 0.05 (0.22) 0.35 (0.48) – – 3.68 (1.03) 3.76 (1.00) 2.98 (1.02) 3.56 (1.02) 3.89 (0.70) 1.03 (0.96) – – – – 4.70 (0.70) 4.28 (1.02) 2.99 (1.30) 4.13 (1.18) 3.61 (0.79) – – – 0.16 (0.36) 0.006 (0.07) Note: IG in-group, NI non-indigenous five-point scale b number between 0 and 5 c 0: disease not present, 1: disease present Standard deviations are presented in parentheses a 306 A. Eller et al. Direct contact Extended contact Health Health compared to 1 year ago .37+ Psychological health Subordinate identity Superordinate identity .30** Diabetes -.38*** Gastrointestinal disorders Fig. 19.3 Path diagram to show the results of regression analyses for first non-indigenous sample in Chile (Study 2). Note: Numbers are standardised partial regression coefficients (ß). +p <0.06; *p < 0.05; **p < 0.01; ***p < 0.001 The regression analyses yielded few results for the first mestizo sample (see Fig. 19.3). Current health, health over the previous year, and psychological health were not predicted by contact or identity variables. Extended contact was marginally associated with an increase in gastrointestinal disorders, β = 0.37, t(201) = 1.90, p = 0.059, and subordinate (mestizo) identity was related to higher incidence of diabetes, β = 0.30, t(201) = 2.95, p = 0.004, while superordinate (Chilean) identity was related to lower incidence of diabetes, β = −0.38, t(201) = −3.73, p < 0.001. The effect of contact with indigenous people on improved health over the previous year was qualified by a three-way interaction, β = −0.18, t(197) = −2.08, p = 0.04. More specifically, more contact with indigenous people and stronger identification as mestizo was associated with improved health when superordinate identity as Chilean was low (−1SD; β = 0.58, t(197) = 2.33, p = 0.02). When superordinate identity was high (1SD), however, this association became nonsignificant (β = 0.10, t(197) = −1.17, p = 0.28). Unpacking this complex relationship, contact with indigenous people positively predicted health over the previous year (β = 0.20, t(203) = 2.21, p = 0.027) when subordinate identification as mestizo was high (1SD) but not when subordinate identification as mestizo was low (−1SD; β = 0.11, t(203) = 1.05, p = 0.24). In the second mestizo sample (see Fig. 19.4), direct contact with indigenous people was associated with better psychological health, β = 0.13, t(195) = 2.25, p = 0.026, and with fewer stress-related diseases, β = −0.75, t(195) = −3.37, p = 0.001. Extended contact was marginally related to better health over the previous year, β = 0.12, t(195) = 1.78, p = 0.076, and to better psychological health, β = 0.09, t(195) = 2.15, p = 0.03. A higher incidence of stress-related diseases was 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 307 superordinate ID low: .21* superordinate ID high: -.12 ns Direct contact Health subordinate ID low: -.25* subordinate ID high: .20* Extended contact .12+ Health compared to 1 year ago .09* .13* Subordinate identity Psychological health -.75*** -.14** Superordinate identity .10+ Stress-related diseases Fig. 19.4 Path diagram to show the results of regression analyses for second non-indigenous sample in Chile (Study 2). Note: Solid lines represent significant paths; interrupted lines represent interactions; numbers are standardised partial regression coefficients (ß). +p < 0.08; *p < 0.05; **p < 0.01; ***p < 0.001 predicted by lower subordinate (mestizo) identity, β = −0.14, t(195) = −2.70, p = 0.008, and marginally, by higher superordinate (Latin American) identity, β = 0.10, t(195) = 1.81, p = 0.072. We also found two two-way interactions, both relating to current health: direct out-group contact × superordinate identity (β = −0.16, t(201) = −2.56, p = 0.01) and extended contact x subordinate identity (β = 0.22, t(191) = 2.99, p = 0.003). More specifically, direct contact was positively associated with current health (β = 0.21, t(201) = 2.21, p = 0.03) when superordinate identity was low (−1SD). However, when superordinate identity was high (1SD) this association became negative and non-significant (β = −0.12, t(201) = −1.24, p = 0.21). As for the extended contact × subordinate identity interaction, contact was significantly and negatively associated with current health (β = −0.25, t(191) = −2.28, p = 0.028) when subgroup identity was low (−1SD). When subordinate identity was high (1SD), however, contact positively predicted current health (β = 0.20, t(201) = 1.98, p = 0.05). In the indigenous sample (see Fig. 19.5), direct contact with non-indigenous Chileans was associated with better current health, β = 0.07, t(145) = 2.35, p = 0.02, while extended contact related to better psychological health, β = 0.06, t(145) = 2.26, p < 0.03. A higher subordinate (indigenous) identity was associated with improved health over the previous year, β = 0.06, t(145) = 2.36, p = 0.02, as well as with a lower incidence of colds, β = −0.27, t(145) = −2.96, p = 0.004, and cancer, β = −0.33, t(145) = −3.26, p = 0.001. A higher superordinate (Chilean) identity was related to higher psychological health, β = 0.04, t(145) = 2.06, p = 0.04. 308 A. Eller et al. superordinate ID low: .31** superordinate ID high: .22 ns Direct contact Health superordinate ID low: .11 ns superordinate ID high: -.15 ns superordinate ID low: .17 ns .07* Extended contact Health compared to 1 year ago Contact with other indigen. groups Subordinate identity Superordinate identity superordinate ID high: -.14 ns .06* .06* Psychological health -.27** Colds .04* -.33*** Cancer Fig. 19.5 Path diagram to show the results of regression analyses for indigenous participants in Chile (Study 2). Note: Solid lines represent significant paths; interrupted lines represent interactions; numbers are standardised partial regression coefficients (ß). *p < 0.05; **p < 0.01; ***p < 0.001 The relation between direct contact with Chileans on improved health over the previous year was qualified by a marginal interaction with superordinate identity, β = −0.08, t(145) = −1.82, p = 0.07. More specifically, contact was significantly and positively associated with improved health over the previous year when superordinate identity was low (−1SD; β = 0.31, t(147) = 2.80, p = 0.006); but only marginally when superordinate identity was high (β = 0.22, t(139) = 1.79, p = 0.07). In addition to the interactions we discussed earlier, we found two more interactions. First interaction was between contact with other indigenous groups and superordinate identification on current health, β = −0.13, t(145) = −2.99, p = 0.003. Although not significant, the association between contact and current health changed direction at different levels of superordinate identity, at low identity contact was related to better health, (−1SD: β = 0.17, t(145) = 1.09, p = 0.27), while at high identity contact was associated with worse health (1SD; β = −0.14, t(145) = −1.15, p = 0.25). The second interaction was between extended contact and superordinate identification on improved health over the previous year, β = 0.08, t(145) = 1.81, p = 0.07. Similarly, the association between extended contact and improved health over the previous year changed at different levels of superordinate identity; at low identity contact was related to improved health (−1SD: β = 0.11, t(145) = 0.81, p = 0.41) while at high superordinate identity contact was associated with decreased health (1SD; β = −0.15, t(145) = −1.25, p = 0.21). Moreover, there was a three-way interaction: direct contact × subordinate identity × superordinate identity on health over the previous year, β = 0.22, t(144) = 2.69, 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 309 p < 0.008. More specifically, more contact with the mestizo out-group and stronger identification as Chilean were associated with deteriorated health when subordinate identity as indigenous was low (−1SD; β = 0.34, t(144) = −2.74, p = −0.007). When subordinate identity was high (1SD), however, this association became non-significant (+1SD, β = −0.001, t(144) = −0.74, p = 0.99). Unpacking this complex relationship, contact with mestizos positively predicted health over the previous year (β = 0.20, t(144) = 2.80, p = 0.006) when superordinate identification as Chilean was low (−1SD), but not when it was high (1SD; β = 0.22, t(144) = 1.79, p = 0.08). In summary, with the exception of extended contact in the first mestizo sample, direct and extended out-group contact had exclusively positive effects on physical and psychological health for the second mestizo sample and the indigenous sample. Again with the exception of the first mestizo sample, higher subordinate identification (mestizo or indigenous, respectively) had exclusively positive effects on physical health—in the indigenous sample, it was even associated with a lower incidence of cancer. When the superordinate identity was Chilean, it related to better physical health (first sample) and better psychological health (third sample). When the superordinate identity was Latin America, it was associated with more stress-related diseases (second sample). In combination with contact, low superordinate identity consistently produced more positive effects than high superordinate identity whereas low subordinate identity consistently produced more negative effects on health than high subordinate identity. General Discussion The research presented in this chapter explored the impact of social identity and intergroup contact on psychological and physical health among samples of ethnic majority and minority groups in Mexico and Chile. This research is innovative in that (a) it examined the effects of different levels of categorisation/identification, (b) it differentiated social identity and social contact as discrete predictors of health, (c) it assessed the effects of intergroup contact on health-related outcomes, and (d) it employed under-researched, non-WEIRD populations (see also Khan et al., 2014). Our analyses revealed a very complex pattern of results that confirmed our predictions in most cases but contradicted them in some other instances. In both studies, direct and extended out-group contact had exclusively beneficial effects on physical and psychological health among indigenous participants while their impact for mestizo participants was more mixed. Contact with other indigenous groups had positive physical but negative psychological outcomes (Study 1). Tropp and Pettigrew’s (2005) meta-analytic results relating to contact between majority and minority groups showed stronger effects for the former than the latter. On the one hand, the fact that contact had such positive effects on health among indigenous participants is very positive. On the other hand, the fact that 310 A. Eller et al. contact with indigenous groups in some cases worsened health for majority members might be due to the fact that contact with indigenous groups is not approved of, not normative among majority members. The negative psychological effect of contact with other indigenous groups might be explained by the fact that through this contact indigenous people are reminded of their inferior position in society and their plight. In general, it is intriguing that for indigenous participants in Mexico psychological and physical health were not affected uniformly through contact. These non-expected findings notwithstanding, it is remarkable that intergroup contact had any effects whatsoever—and more so, mostly positive ones— on health-related outcomes. The impact of the identification variables was mostly in line with our predictions. Subordinate identification had detrimental effects for mestizos in Study 1 while it showed mixed impact for indigenous people in Study 1 and only beneficial effects for the second and third samples in Study 2. Specifically, it decreased stressrelated diseases in non-indigenous Chileans and reduced the incidence of colds and even cancer in indigenous participants. Superordinate identification had positive impact among mestizos but mixed effects among indigenous participants in Study 1. In Study 2, superordinate identification had positive outcomes for White and indigenous participants when the superordinate category was Chile but not when it was Latin America. Interacting with contact variables, we found low superordinate and high subordinate identities to be beneficial in terms of improved health for all three samples. This is in line with the Hewstone–Brown model (Brown & Hewstone, 2005) of categorisation during contact, which holds that clear group distinctions during intergroup contact will result in better intergroup outcomes than identifying with a common, superordinate category. The positive effects of subordinate identity on health, mainly among indigenous samples, can be elucidated with Schmitt and Branscombe’s (2002) rejectionidentification model. In it, the authors argue that the shared social identity of members of stigmatised groups provides a basis for giving, receiving, and benefiting from social support that provides individuals with the emotional, intellectual, and material resources to cope with and resist the injustice of discrimination, prejudice, and stigma (see Haslam, Jetten & Waghorn, 2009). A longitudinal study of immigrants from Russia to Finland exemplifies these processes. It demonstrates how the maintenance of identification with an ethnic group of origin and the development of identification with a new national group combine to define people’s experiences of discrimination and the resulting stress (Jasinskaja-Lahti et al., 2009). A study by Oyserman, Fryberg, and Yoder (2007) proves instructive to explain the often negative effects of superordinate identification on indigenous people’s health. These authors found that members of ethnic minority groups that do not identity with mainstream society often react against health-related messages from the majority. The predominantly positive effects of superordinate identification on health among majority group members makes sense when considering Mummendey and Wenzel’s (1999) in-group projection model (IPM). Grounded in SCT, Mummendey 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 311 and Wenzel argue that in-group and out-group members compare themselves on dimensions that define the superordinate category. The groups’ evaluation depends on their relative prototypicality of the superordinate category. In-group members tend to see their own group as more prototypical and hence, they view the outgroup as a deviation from the norm. As shown empirically, high-status groups often legitimise status differences between the groups by pointing to their greater perceived prototypicality of the superordinate category (Weber, Mummendey, & Waldzus, 2002). Applying this reasoning to our research, non-indigenous majority group members in Mexico and Chile are bound to see themselves as more prototypical than indigenous participants of the categories of Mexicans and Chileans. Not only do they have numerical majority but they also hold higher status positions that need to be justified. Moreover, at least in the case of Mexico, where superordinate identification had solely positive effects among mestizos, the superordinate, national category provides a more “natural” basis for identification than the subordinate majority ethnicity. This is different for minority group members for whom their ethnic group membership is actually more important than their nationality, as evidenced by the descriptive statistics. It is also interesting to note that in Study 2, superordinate identification produced beneficial effects for both majority and minority groups when the category was Chile but not when it was Latin America. Latin America includes other Spanishspeaking White people, but more importantly, it also includes other indigenous minority groups with whom indigenous participants in Chile might have formed a psychological connection. In contrast, it appears that the category of Latin America is simply too inclusive. Findings might have been different if the frame of comparison and intergroup contact related to specific other Latin American countries. Related to these issues Jetten et al. (2014, p. 112) ask “is this relationship, group or role important in defining who I am? We argue that only when the answer to this question is “yes” will social connectedness have the capacity to predict outcomes such as health behaviour and well-being.” Similarly, Sani (2011) discusses (interpersonal) relationships that do not cure. A substantial number of studies have shown that negative relationships impair physical and mental health (e.g. Holt-Lunstad, Uchino, Smith, & Hicks, 2007). Similar processes might hold for intergroup relations. Our research had certain limitations. Firstly, we relied on self-reported data, which is open to subjective interpretations, particularly with regard to the healthrelated questions. Some people might tend to over report their incidence of certain diseases while others might under report this. Objective data, such as individual health records or physiological data, would have been more accurate. A second limitation is our use of cross-sectional data, which cannot establish a causal chain of events. Therefore, while it is plausible that more and better contact and higher identification improves health, it is equally feasible that participants that suffer from diseases lose social connections and identify less with their groups and categories. This circularity between contact and health was captured nicely by Putnam (2000) who noted that “It is not just that because we are well we are more likely to partici- 312 A. Eller et al. pate in group life, but also that because we participate in group life we are more likely to be well” (in Haslam, Jetten & Waghorn, 2009, p. 15). To conclude, our research was the first to show the (mostly) beneficial effects of different levels of identification as well as intergroup contact on health-related outcomes. This broadens both the social cure and the intergroup contact literatures considerably. Given the ubiquity of social groups/categories and, following from this, of intergroup contact, our findings are very encouraging. Future research should focus on more objective measures of identity, contact, and particularly, health variables. Moreover, longitudinal research could not only illuminate how the relations among identity, contact, and health transform over time but also how the variables relate to each other causally. Acknowledgements We are grateful to Benjamin Arditi for his comments on an earlier version of this chapter. We thank Andres Perez, Fernanda Mata, Rocio Castro, Katia Rivera, and Miguel Freyre for their help with data collection, translation of materials, and bibliographic searches. Preparation of this chapter was facilitated by PAPIIT/UNAM grant IN303112-3 to the first author. Different aspects of the same dataset have been presented in Cakal, H., Eller, A., Sirlopu, D., & Perez, A. (under review). Intergroup relations in Latin America: Intergroup contact, common ingroup identity and activism among indigenous groups in Mexico and Chile. Journal of Social Issues. References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley. Bizumic, B., Reynolds, K. J., Turner, J. C., Bromhead, D., & Subasic, E. (2009). The role of the group in individual functioning: School identification and the psychological well-being of staff and students. Applied Psychology, 58(1), 171–192. doi:10.1111/j.1464-0597.2008.00387.x. Brown, R., & Hewstone, M. (2005). An integrative theory of intergroup contact. Advances in Experimental Social Psychology, 37, 255–343. doi:10.1016/S0065-2601(05)37005-5. Cakal, H., Eller, A., Sirlopú, D., & Pérez, A. (under review). Intergroup relations in Latin America: Intergroup contact, common ingroup identity and activism among Indigenous groups in Mexico and Chile. Journal of Social Issues. Cakal, H., Hewstone, M., Schwar, G., & Heath, A. (2011). An investigation of the social identity model of collective action and the “sedative” effect of intergroup contact among Black and White students in South Africa. British Journal of Social Psychology, 50, 606–627. doi:10.1111/j.2044-8309.2011.02075.x. CEPAL (2012). Desigualdades territoriales y exclusión social del pueblo mapuche en Chile: Situación en la comuna de Ercilla desde un enfoque de derechos. (Territorial inequality and social exclusion of the mapuche people in Chile: The situation in the commune of Ercilla from a standpoint of rights). Santiago de Chile: Comisión Económica para América Latina y el Caribe (CEPAL). Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Hillsdale: Erlbaum. Cole, S. W., Kemeny, M. E., Taylor, S. E., Visscher, B. R., & Fahey, J. L. (1996). Accelerated course of human immunodeficiency virus infection in gay men who conceal their homosexual identity. Psychosomatic Medicine, 58, 219–231. 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 313 Cruwys, T., Haslam, S. A., Dingle, G. A., Haslam, C., & Jetten, J. (2014). Depression and social identity: An integrative review. Personality and Social Psychology Review, 18(3), 215–238. doi:10.1177/1088868314523839. De la Maza, F. (2014). Between conflict and recognition: The construction of Chilean indigenous policy in the Araucanía region. Critique of Anthropology, 34(3), 346–366. doi:10.1177/03082 75X14531836. Dovidio, J. F., Eller, A., & Hewstone, M. (2011). Improving intergroup relations through direct, extended and other forms of indirect contact. Group Processes and Intergroup Relations, 14(2), 175–192. doi:10.1177/1368430210390555. Eller, A., & Abrams, D. (2003). “Gringos” in Mexico: Cross-sectional and longitudinal effects of language school-promoted contact on intergroup bias. Group Processes and Intergroup Relations, 6, 55–75. doi:10.1177/1368430203006001012. Eller, A., & Abrams, D. (2004). Come together: Longitudinal comparisons of Pettigrew’s reformulated intergroup contact model and the common ingroup identity model in Anglo-French and MexicanAmerican contexts. European Journal of Social Psychology, 34, 229–256. doi:10.1002/ejsp.194. Eller, A., Abrams, D., & Gomez, A. (2012). When the direct route is blocked: The extended contact pathway to improving intergroup relations. International Journal of Intercultural Relations, 36(5), 637–646. doi:10.1016/j.ijintrel.2012.03.005. Eller, A., Abrams, D., & Zimmermann, A. (2011). Two degrees of separation: A longitudinal study of actual and perceived extended international contact. Group Processes and Intergroup Relations, 14(2), 175–191. doi:10.1177/1368430210391120. ENADIS. (2010). Retrieved from http://www.conapred.org.mx/index.php?contenido=pagina&id=424& id_opcion=436&op=436. Goldberg, D. P., Gater, R., Sartorius, N., Ustun, T. B., Piccinelli, M., Gureje, O., et al. (1997). The validity of two versions of the GHQ in the WHO study of mental illness in general health care. Psychological Medicine, 27(1), 191–197. doi:10.1017/S0033291796004242. Hall, G. H., & Patrinos, A. (2014). Latin America. In G. H. Hall & A. Patrinos (Eds.), Indigenous peoples’ poverty and development (pp. 344–358). Cambridge, UK: University Press. Haslam, S. A., Jetten, J., Postmes, T., & Haslam, C. (2009). Social identity, health and well-being: An emerging agenda for applied psychology. Applied Psychology: An International Review, 58, 1–23. doi:10.1111/j.1464-0597.2008.00379.x. Haslam, S. A., Jetten, J., & Waghorn, C. (2009). Social identification, stress, and citizenship in teams: A five-phase longitudinal study. Stress and Health, 25, 21–30. doi:10.1002/smi.1221. Helliwell, J. F., & Barrington-Leigh, C. P. (2011). How much is social capital worth? In J. Jetten, C. Haslam, & S. A. Haslam (Eds.), The social cure: Identity, health and well-being (pp. 55–71). New York: Psychology Press. Henrich, J., Heine, S. J., & Norenzayan, A. (2010). Most people are not WEIRD. Nature, 466, 29. doi:10.1038/466029a. Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLoS Medicine, 7(7), e1000316. doi:10.1371/journal.pmed.1000316. Holt-Lunstad, J. L., Uchino, B. N., Smith, T. W., & Hicks, A. (2007). On the importance of relationship quality: The impact of ambivalence in friendships on cardiovascular functioning. Annals of Behavioral Medicine, 33, 278–290. doi:10.1007/BF02879910. Jasinskaja-Lahti, I., Liebkind, K., & Solheim, E. (2009). To identify or not to identify? National disidentification as an alternative reaction to perceived ethnic discrimination. Applied Psychology, 58(1), 105–128. doi:10.1111/j.1464-0597.2008.00384.x. Jetten, J., Haslam, C., Haslam, S. A., Dingle, G., & Jones, J. (2014). How groups affect our health and wellbeing: The path from theory to policy. Social Issues and Policy Review, 8(1), 103–130. doi:10.1111/sipr.12003. Jetten, J., Haslam, C., & Haslam, S. A. (2011). Why we need a social identity analysis of health. In J. Jetten, C. Haslam, & S. A. Haslam (Eds.), The social cure: Identity, health, and well-being (pp. 3–19). New York: Psychology Press. 314 A. Eller et al. Khan, S. S., Hopkins, N., Reicher, S., Tewari, S., Srinivasan, N., & Stevenson, C. (2014). Shared identity predicts enhanced health at mass gatherings. Group Processes & Intergroup Relations, 18(4), 504–522. doi:10.1177/1368430214556703. Merino, M.-E., & Quilaqueo, D. (2003). Ethnic prejudice against Mapuche in the discourse of members of Chilean society as a reflection of the racialist ideology of the Spanish conqueror. American Indian Culture and Research Journal, 27(4), 105–116. doi:10.17953/aicr.27.4. xu0113854222ln72. Mummendey, A., & Wenzel, M. (1999). Social discrimination and tolerance in intergroup relations: Reactions to intergroup difference. Personality and Social Psychology Review, 3, 158– 174. doi:10.1207/s15327957pspr0302_4. Oyserman, D., Fryberg, S. A., & Yoder, N. (2007). Identity-based motivation and health. Journal of Personality and Social Psychology, 93, 1011–1027. doi:10.1037/0022-3514.93.6.1011. Paolini, S., Hewstone, M., Cairns, E., & Voci, A. (2004). Effects of direct and indirect cross group friendships on judgments of Catholics and Protestants in Northern Ireland: The mediating role of an anxiety-reduction mechanism. Personality and Social Psychology Bulletin, 30, 770–786. doi:10.1177/0146167203262848. Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90, 751–783. doi:10.1037/0022-3514.90.5.751. Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New York: Simon & Schuster. Quellet-Morin, I., Wong, C. C., Danese, A., Pariante, C. M., Papadopoulos, A. S., Mill, J., et al. (2013). Increased serotonin transporter gene (SERT) DNA methylation is associated with bullying victimization and blunted cortisol response to stress in childhood: A longitudinal study of discordant monozygotic twins. Psychological Medicine, 43(9), 1813–1823. doi:10.1017/ S0033291712002784. Ruiz-Linares, A., Adhikari, K., Acuña-Alonzo, V., Quinto-Sanchez, M., Jaramillo, C., Arias, W., et al. (2014). Admixture in Latin America: Geographic structure, phenotypic diversity and selfperception of ancestry based on 7,342 individuals. PLoS Genetics, 10, e1004572. doi:10.1371/ journal.pgen.1004572. Sani, F. (2011). Group identification, social relationships, and health. In J. Jetten, C. Haslam, & S. A. Haslam (Eds.), The social cure: Identity, health, and well-being (pp. 21–38). New York: Psychology Press. Sani, F., Herrera, M., Wakefield, J. R. H., Boroch, O., & Gulyas, C. (2012). Comparing social contact and group identification as predictors of mental health. British Journal of Social Psychology, 51(4), 781–790. doi:10.1111/j.2044-8309.2012.02101.x. Schmitt, M. T., & Branscombe, N. R. (2002). The meaning and consequences of perceived discrimination in disadvantaged and privileged social groups. European Review of Social Psychology, 12(1), 167– 199. doi:10.1002/0470013478.ch6. St Claire, L., & He, Y. (2009). How do I know if I need a hearing aid? Further support for the self-categorisation approach to symptom perception. Applied Psychology, 58, 24–41. doi:10.1111/j.1464-0597.2008.00380.x. Staerklé, C., Sidanius, J., Green, E. G. T., & Molina, L. E. (2010). Ethnic minority-majority asymmetry in national attitudes around the world: A multilevel analysis. Political Psychology, 31(4), 491–519. doi:10.1111/j.1467-9221.2010.00766.x. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks-Cole. Tropp, L. R., & Pettigrew, T. F. (2005). Relationships between intergroup contact and prejudice among minority and majority status groups. Psychological Science, 16, 951–957. doi:10.1111/j.1467-9280.2005.01643.x. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford: Blackwell. 19 Identity, Contact, and Health Among Majority and Minority Ethnic Groups… 315 Van Cott, D. L. (2007). Latin America’s indigenous people. Journal of Democracy, 18, 127–142. Weber, U., Mummendey, A., & Waldzus, S. (2002). Perceived legitimacy of intergroup status differences: Its prediction by relative ingroup prototypicality. European Journal of Social Psychology, 32, 449–470. doi:10.1002/ejsp.102. Wright, S. C., Aron, A., McLaughlin-Volpe, T., & Ropp, S. A. (1997). The extended contact effect: Knowledge of cross-group friendships and prejudice. Journal of Personality and Social Psychology, 73, 73–90. doi:10.1037/0022-3514.73.1.73.