Ethnic Identity Running head: ETHNIC IDENTITY MULTIDIMENSIONAL ETHNIC IDENTITY: CROATIANS IN NEW ZEALAND JUDITA F. JANKOVIC GRAHAM M. VAUGHAN University Of Auckland 1 Ethnic Identity 2 Abstract Much of the early research on ethnic identity implied that it is an all-or-none phenomenon. It has often been measured by acculturation scales, predominantly assessing the presence or absence of particular ethnic practices and ethnic behaviour (e.g., Cuellar, Arnold, & Maldonado, 1995; Marin, Sabogal, Marin, Otero-Sabogal, & Perez-Stable, 1987; Mavreas, Bebbington, & Der, 1989). The primary aim of this study was to assess ethnic identity among members of a particular Croatian community (in Auckland, New Zealand) from a social psychological perspective. A questionnaire designed to assess different aspects of ethnic identity and acculturation was administered to 156 individuals of Croatian descent, of first, second, third and fourth generation, aged between 16 and 89 years. Ethnic identity states of the participants were assessed using Berry’s (1989) cross-cultural model. Mancova analyses revealed that NZ-born participants show ethnic cognitive reaffirmation, whereas immigrant participants show host cognitive and affective reaffirmation. The questionnaire items were factor analysed and four factors were extracted using an Oblimin rotation. Correlational analyses supported an orthogonal ethnic identification model. Ethnic Identity 3 Introduction According to Erikson (1968), identity is essentially the answer to the question “Who am I?” and relates to a psychosocial process of an individual’s search for self. Erikson thought that identity was a “process located in the core of the individual and yet also in the core of his communal culture” (p.127), acknowledging that, complementary to personal identity, there were other aspects of identity which developed within a broader context of role relationships. Ethnic identity is one part of an individual’s overall identity and it may include referents that are personal (e.g. name), social (e.g. family) and cultural (e.g. ethnic language and customs) (Aboud,1987). Ethnic identity is a social psychological process (DeVos & Romanucci-Ross, 1995) and can be measured both on an individual level (Phinney, 1993) and a group level (Barth, 1969; Berry, Poortinga, Segall, & Dasen, 1992; DeVos & RomanucciRoss, 1995). Some researchers have recognised that ethnic identity is multidimensional, with many different facets (Bernal & Knight, 1993; Buriel & Cardoza, 1993; Hocoy, 1996; Isajiw, 1990; Keefe & Padilla, 1987; Knight, Bernal, Cota, Garza, & Ocampo, 1993; Liebkind, 1992; Salvaterra, 1994; Saylor & Aries, 1999; Zak, 1973). Ethnic identity has attributes of strength and salience, and is set in the context of the cultural norms and values of an ethnic group (Phinney, 1996). The present study is primarily concerned with the measurement of ethnic identity salience, and some of the cultural values and attitudes pertinent to the Croatian and New Zealand cultures. Berry’s (1980) cross-cultural model, which deals with the adaptation of immigrants and their descendants in a host culture, was used as a starting point to assess the participants’ ethnic states. This model has two key features that distinguish it from previous “melting pot” conceptualisations of the adaptation of ethnic groups in host countries. First, immigrant adaptation is no longer a simple unilinear phenomenon, as the two cultural systems are treated independently. Thus, immigrants can decide to move in and establish relations with the dominant society, without necessarily giving up their own culture. Second, Berry distinguished four styles of adaptation in place of a forced choice between assimilating or not assimilating. According to the cross-cultural model, ethnic identity reflects the way that people use the host ethnic group to locate and express themselves as social beings. The four styles of adaptation are assimilation, integration, separation and marginalisation. Assimilation occurs when the host society’s attitudes and behaviours are acquired, while the heritage culture and identity are relinquished. Separation is the reverse of the assimilation style: the Ethnic Identity 4 host society is rejected, while there is a desire to maintain one’s heritage culture and identity. Integration emphasises active participation in the host society, as well as maintenance of the heritage culture and identity. Individuals who embrace an integrationist style may be considered bicultural. Marginalisation implies that the individual culturally and psychologically rejects both the host society and his or her own culture. Berry acknowledged both inter-individual and intra-individual variability of the adaptation styles. Although people adopt different styles, they may also switch between styles over time. Adaptation styles may manifest differently across various domains of behaviour and social life. For example, an individual may opt for economic assimilation in work, linguistic integration by the way of bilingualism and marital separation by the way of endogamy (Berry, Poortinga, Segall, & Dasen, 1992a). There are limitations to the cross-cultural model. For example, both an assimilationist and an integrationist may have the same level of ethnic identity, but may differ in deciding whether it is worth manifesting or not. The model is somewhat arbitrary and nonrepresentative, as it assumes only two referent cultures — the ethnic group and the mainstream, host group. There are many ethnic groups in multicultural societies; hence a participant may have more than two reference groups. The psychological nature of ethnic identity has been ambiguous in previous research, for example, in equating ethnic categories with ethnic groups. There is a difference between groups of people with a common feature, such as the same birthplace or hair colour, and a group whose members interact in a meaningful way, share a sense of belonging, in addition to having a common feature (Hogg & Abrams, 1988). The former is not truly a psychological group, whereas the latter is (McKay & Lewins, 1978). Thus, a common cultural pattern alone does not define an ethnic group (Francis, 1947, cited in Hutnik, 1991). The current study also tests an orthogonal identification model by measuring separately each participant’s level of ethnic identity with Croatian and New Zealand cultures. Independence of identification with two or more cultures is assumed in this model. Multidimensionality of ethnic identity is also assumed, emphasising that the degree to which new cultural traits are accepted and traditional cultural traits are lost varies from trait to trait (Keefe & Padilla, 1987). Thus, unlike the assumption of the cross-cultural model, a bicultural person is not necessarily assumed to be highly adapted in both cultures. The implication that Ethnic Identity 5 follows is that studies that compute single scores for acculturation and ethnic identity (e.g. Suinn, Rickard-Figueroa, Lew, & Vigil, 1987) cannot capture variability among members. METHOD Instruments Various methods have been employed in measuring ethnic identity, most using crosssectional quantitative surveys (Barona & Miller, 1994; Bond & Yang, 1982; Knight, Bernal, Garza, Cota, & Ocampo, 1993; Masuda, Matsumoto, & Meredith, 1970). However, qualitative techniques have been also used, such as a Twenty Statement Test (TST) (e.g. Kuhn & McPartland, 1954;), scales (e.g. Rosenthal & Hrynevich, 1985), vignettes (e.g. Pruegger & Rogers, 1993) and a card sorting task (e.g. Berry, Kim, Power, Young, & Bujaki, 1989). In one TST approach (Hutnik, 1986; 1991), participants were timed to respond to the question “Who am I?” on a sheet containing 20 numbered blanks. The written instructions do not direct the respondent to answer in a particular way; the TST is a general open-ended instrument. Rosenthal and Hrynevich’s (1985) Ethnic Identification Scale compared Greek, Italian and Anglo Australian adolescents with respect to their own identities and the way their ethnic identity relates to the other ethnic groups. It required participants to tick the items with which they associate, or how much the specific item describes them — for example, an average Australian kid, people who are religious, people whose parents won’t give them any independence, Australians who are proud of being Australian, and so on. The interview focused on perceived similarities and differences between Anglo and non-Anglo Australians. The current study employed a cross-sectional survey intended for a sample of the Croatian community in Auckland, both immigrants and descendants born in New Zealand. A quantitative survey was designed for the purposes of this study with 30 items in the form of statements. The questionnaire also included an open-ended question on self-identification. Participants had the opportunity to express final comments at the end of the questionnaire. Most of the items were statements using a 5-point Likert-scale to indicate the level of agreement or disagreement with the statements. The statements referred to behaviours, attitudes and affect related to ethnic identity. The self-identification item asked the participants to write down the group label they most like to use to refer to the group they like to associate in terms of ethnicity. The questionnaire was designed in accordance with the guidelines of the University of Auckland Human Subjects Ethics Committee and was pilot Ethnic Identity 6 tested. The final form of the questionnaire consisted of three parts: (a) socio-cultural practices; (b) ethnic identity; (c) general information. 1 There are several critical issues pertinent to ethnic identity surveys that were considered and dealt with in our questionnaire (see Table 1). TABLE 1 Generic Issues And Specific Examples Pertinent To Ethnic Identity Surveys Generic Issue Specific example Forced choice vs. bipolar vs. separate items — for example, a forced Questionnaire wording of the and choice item Are you Aboriginal or Australian? used by Sommerlad and Berry degree of concordance (1970) reflects the bipolar model of ethnic identity promoted at the time (cited between theoretical in Hutnik, 1991). Some studies used more than two options, but a bipolar assumptions and continuum was still assumed as identifications with the host and the ethnic operational measurement culture was not measured separately. For example, in Johnston’s study (1965, cited in Ho, 1995), immigrants were asked to identify themselves on a scale ranging from completely Australian, more Australian than Dutch, equally Australian and Dutch, more Dutch then Australian to completely Dutch. Italian and Polish immigrants in Australia were assessed in a similar way (Taft, 1961; Wiseman, 1971 cited in Ho, 1995). The present study assesses identification with Croatian culture and New Zealand culture separately, using two, rather than a single question, i.e.: Ethnic selfidentification issues 1. To what extent do you consider yourself as a Croatian? 2. To what extent do you consider yourself as a New Zealander? (Note: the same example for questionnaire wording given above also applies to the generic issue of self-identification.) Collapsed coding may fail to detect subtle psychological correlates associated with self-identification labels that have been located under the same code category. Some researchers treat ethnic labels, (such as Japanese) and bicultural labels, (such as Japanese American) as equivalent responses (e.g. Stephan & Stephan, 1989). However, other studies have pointed to significant psychological differences among ethnic groups based on the various ethnic labels that they themselves use. For example, the diverse labels used by Ethnic Identity 7 participants of Mexican origin, such as Mexican, Chicano, Latino, Hispanic and Mexican American have different psychological correlates and meanings attached to them (Buriel & Cardoza, 1993; Keefe & Padilla, 1987). Procedure In our study, a questionnaire was mailed to 350 potential participants in the greater area of the city Auckland (population approximately 1 million). Snowball sampling was utilised: key informants of Croatian descent who are involved in the network of the Croatian community in Auckland distributed the questionnaires to potential participants. The respondents were anonymous, and the criteria for their participation were age, ethnic descent and place of residence. Selection checks were made after the survey by noting the participation rates within the demographic section of each of the questionnaires. The overall response rate was 45%, with 156 participants completing and returning the questionnaire. Participants’ ages ranged from 16 to 89 years, with an average age of 40. Participants The largest single generation group represented was the second generation, consisting of 71% participants. Overall, there were more males than females (82 males and 73 females). The male to female ratio was approximately even within each of the generation groups, with the smallest proportion of females in the first generation. Among the first (immigrant) generation, the average age on arrival in New Zealand was 24 years and the average length of residence was 19 years. These two statistics have relatively large standard deviations, indicating the heterogeneity of the immigrant sample in these respects. The group of potential participants to whom a questionnaire has been sent may have been a representative sample. However, the group of individuals who returned their questionnaire (45%) may not have been representative. The issue of representativeness is a general limitation of mail out questionnaires. Hence, any surveys with less than 100% return rate have to be interpreted with caution, rendering generalizations as limited to the survey sample only. RESULTS Ethnic Identity 8 Mean scores of ethnic identity Ethnic identity is a state and is experienced as a matter of degree, rather than a simple either/or construct (Hutnik, 1986) and, because it consists of different facets, is multidimensional (Oetting & Beauvais, 1991). In the present study, the questionnaire was designed to measure various facets of Croatian and New Zealand identity. Ethnic identity scores have not been calculated as single total scores in the manner of some previous research (e.g. Keefe & Padilla, 1987; Pawliuk et al., 1996; Suinn, Rickard-Figueroa, Lew, & Vigil, 1987); one aggregated score does not allow for nor capture a construct that is probably multidimensional. Instead, the means and standard deviations of different ethnic identity items are presented in Table 2. On closer examination of the means between the immigrant and the New Zealand-born participants, slight discrepancies exist regarding some items. For example, the first generation more strongly agrees that Croatian heritage is important for them than the New Zealand-born generations. On the other hand, New Zealand-born generations more strongly agree that being a New Zealander is important for them, as well as with the extent of being a New Zealander. TABLE 2 Mean Raw Scores and Standard Deviations of Ethnic Identity Items N Item (questionnaire item; section) NZ-born o. Total generations First sample generation (2nd, 3rd and 4th) M S M S M S ean D ean D ean D Importance of Croatian heritage (4; 1) 1.89 1.78 1.61 .67 1.87 .99 2. Importance of region of origin (5; 1) 2.16 1.18 2.15 1.15 2.15 1.19 3. Importance of particular religious affiliation (6; 1) 2.74 2.05 2.73 2.89 2.74 1.22 4. Importance of being a New Zealander (7; 1) 2.09 .98 2.29 .93 1.97 .99 5 Comfort with both cultures (8; 1) 1.71 2.23 .61 2.46 2.16 2.82 1.23 1.78 1.11 1 . 2. 36 6. Extent of being a New Zealander (9a; 1) 2.19 1. Ethnic Identity 9 27 7. Extent of being a Croatian (9b; 1) 1.79 .95 1.60 .78 1.91 1.04 8. Feelings about being a New Zealander (10a; 1) 1.76 .77 1.98 .81 1.62 .71 9. Feelings about being a Croatian (10b; 1) 1.67 .66 1.61 .52 1.71 .74 10. Important for my children to be aware of their 1.84 2.65 1.50 .57 1.74 .80 Croatian cultural heritage (1; 2) 11. Croatians should marry other Croatians (2; 2) 3.62 1.17 3.61 1.16 3.62 1.18 12. Croatians should marry within their religious 3.63 1.17 3.60 1.22 3.65 1.14 1.94 3.33 1.65 .63 2.13 4.26 1.62 .67 1.65 .60 1.60 .71 1.58 .91 1.52 .86 1.62 .94 affiliation (3; 2) 13. Croatians should visit their country of origin, if possible (4; 2) 14. Significant others are aware of my Croatian heritage (1; 3) 15. No embarrassment mentioning Croatian heritage to others (2; 3) 16. It is harmful for Croatians to stick together (3; 3) 2.18 1.09 2.20 1.14 2.18 1.04 17. Importance of feeling part of the Croatian 2.24 1.00 2.06 .92 2.35 1.04 3.17 1.20 3.61 .98 2.87 1.25 2.24 1.11 2.17 .98 community in NZ (4; 3) 18. Others describe me as a typical New Zealander in every aspect of my life (5; 3) 19. Knowledge seeking about ethnic traditions, 2.20 customs, history (6; 3) 20. 1. 03 I have talked to others to learn more about my 2.38 1.07 2.26 1.06 2.46 1.07 1.70 .67 1.55 .67 1.81 .65 ethnic background (7; 3) 21. Knowledge/familiarity regarding recent events in Croatia (1; 4) 22. Feeling of pride in having Croatian heritage (2; 4) 1.58 .65 1.66 .65 1.53 .65 23. Feeling of unity with other Croatians in NZ (3; 4) 2.17 .98 2.27 .98 2.10 .98 24. Sense of belonging to my ethnic group (4; 4) 1.95 .87 2.00 .97 1.92 .80 NOTE: The questionnaire sections are: Ethnic Identity (1), Cultural Values (2), Community Belongingness (3) and Ethnic Pride (4). The Likert scales were coded 1-5 from Strongly agree (M=1) to Strongly disagree (M=5), with negatively worded items reverse coded. Self-identification and other factors With respect to self-identification, most participants chose the ethnic label (62%), followed by the bicultural label (18%) and the host label (15%) (see Table 3). A principal component Ethnic Identity 10 factor analysis was applied to the data using an Oblimin rotation and four factors were extracted. Variance explained for the total sample ranged from 25 percent for factor 1 down to six percent for factor 4 (see Table 4). Oblimin rotation was used, rather than Varimax, as it was anticipated that some factors may be correlated to a certain extent. TABLE 3 Labels Most Preferred by Participants to Describe Their Ethnic Group (%) Generation Label First First: First: pre-1900 post-1900 Second Third Fourth New Zealand- Tota l born Ethnic 85 83 92 54 21 29 47 62 Bicultur 5 5 4 27 36 14 28 18 Host 5 7 - 16 36 43 21 15 Multiet - - - - 7 14 2 1 5 5 4 3 - - 2 4 al hnic Other NOTE: The “other” category includes related ethnic labels (e.g. ethnic groups from the former Yugoslavia), unrelated ethnic labels (e.g. Australian, Hungarian), and non-ethnic terms (e.g. cosmopolitan). Bold typeface indicates the greatest percentage preference for each generation group and for the total sample. Factor 1 reflects an internal identification with Croatian culture and acknowledgement of Croatian identity as part of an individual’s personal identity. It also signifies the importance attached to this identification, and positive feelings attached to it. Other elements that comprise this factor reflect the importance of maintaining certain aspects of ethnic identity, such as the native language, community belongingness and visits to the homeland. Factor 2 also reflects internal identification with Croatian culture and positive feelings associated with this identification. Unlike factor 1, however, additional elements reflect an Ethnic Identity 11 external and group orientation of pride and acknowledgement of ethnic identity to others, as well as unity and a sense of belongingness. Factor 3 reflects a New Zealand identity and is contained in items that measure an internal cognitive and affective identification with, and the importance of, being a New Zealander. It also reflects a looking-glass self, that is, the extent to which individuals see themselves as significant others perceive them, specifically with respect to being a New Zealander in everyday life. The looking-glass self concept originates from the theory of reflected appraisals (Cooley, 1902 cited in Liebkind, 1992; see a modern discussion of the looking glass self in Vaughan & Hogg, 2002). Factor 4 emphasizes ethnic and religious endogamy. This is reflected in the importance attached to keeping boundaries through same ethnicity and same religion marriages, hence preventing exogamy (ethnic or religious intermarriages). The notion of keeping the boundaries through endogamy is also conveyed through the importance of feeling part of the Croatian community (Table 4). TABLE 4 Summary of Factors Factor (with range of loading) 1. Croatian identity A (.57-.75) 2. Croatian identity B (.54-.74) Eigenvalue 7.15 3.31 Description (with % explained variance) The strength and importance of Croatian identity (25%). The strength and pride in acknowledging Croatian identity (11%). The degree to which individuals feel and identify like New 3. New Zealand identity (.64-.80) 2.30 Zealanders and the extent of importance attached to this identity (8%). The extent to which individuals feel Croatians should marry 4. Religious and ethnic endogamy (.43-.63) 1.69 within their own ethnic group and religious affiliation, the importance of feeling part of the Croatian community and the salience of one’s religious affiliation (6%). NOTE: The factor loadings, eigenvalues and variance are based on the pattern matrix SPSS output of an Oblimin rotation. Ethnic Identity 12 Ethnic and host identity states Guided by Sayegh and Lasry (1992), two identification scores that were measured separately determined participants’ ethnic identity states based on the self-reported level of identification with being a Croatian and being a New Zealander. When identification with both Croatian and New Zealand culture is strong (comprised of “agree” and “strongly agree” scores of the Likert scale used), the participant is categorised as having an integrationist or acculturative ethnic identity state. When the two possible identifications are weak (“disagree”, “strongly disagree” and “uncertain”), the ethnic identity state is marginalisation. An assimilationist style is characterised by strong identification with New Zealand and a weak or uncertain identification with Croatian culture. Finally, separation or a dissociative ethnic identity state is characterised by a strong identification to Croatian culture with a weak or uncertain identification with New Zealand culture. The greatest proportion of participants exhibited the integrated ethnic identity state, replicating other previous research findings (Berry, Poortinga, Segall, & Dasen, 1992b; Hutnik, 1991; Isajiw, 1990; Raza, 1997; Rosenthal & Hrynevich, 1985). Croatian and New Zealand cognitive identification Measures that are orthogonal are not correlated. Researchers who have tested for orthogonality of ethnic and host identities have generally supported the orthogonal model (e.g. Rosenthal 1984; Hutnik, 1986; Ho, 1995; Cuellar, Arnold, & Maldonado, 1995; Sayegh & Lasry, 1993; Zak, 1973). In practice, the correlation between the identities measured in these studies was not actually zero. For example, Zak (1973) tested several geographically different samples of participants and the correlation between their Jewish and American identities ranged from .08 to .15. Nevertheless, they concluded that low correlations are an indication that the participants’ Jewish and American identities are effectively independent. Bipolar measures are indicated by negative correlations. However, regardless of the direction of correlations (whether positive or negative), the degree of correlation determines whether the relationship between two variables is meaningful or not. In practice, correlations of <.3 are usually considered meaningless (George & Mallery, 1999). Orthogonality was examined by correlating the items measuring New Zealand and Croatian identification. Table 5 includes the product-moment correlations between identification scores as a Croatian and/or as a New Ethnic Identity 13 Zealander, along with the means and standard deviations of each, both for the total sample and for each generation. TABLE 5 Means, Standard Deviations, and Correlations between Croatian and New Zealand Identification Statistic Identification Generation Firs Second Third Fourth t New Total Zealand-born Croatian 1.60 1.70 2.60 2.57 1.91 1.79 New 2.82 1.93 1.43 1.00 1.78 2.19 Croatian 0.78 0.84 1.40 1.27 1.04 0.95 New Zealand 1.23 1.20 0.65 0.00 1.11 1.27 -.26* -.04 -.15 -- -.13 -.23** M Zealand SD r NOTE: The mean item score was based on a five-point response scale ranging from 1 to 5 for “Very much” to “Not at all”. * p < .05. ** p < .01 Croatian and New Zealand affective identification Two separate items that assessed affective attachment with being a Croatian and/or a New Zealander were analysed in the same manner as the identification items. Table 6 shows the product-moment correlations, means and standard deviations for the two affective attachment items. The results are similar to those for the identification items; the correlation for the degree of affective identification is generally weak, ranging from -.10 to .02. TABLE 6 Means, Standard Deviations and Correlations between Croatian and New Zealand Affective Attachment Statistic Affective attachment Generation Ethnic Identity Fir Second Third Fourth st New Zealand- 14 Total born Croatian 1.61 1.63 2.29 1.29 1.71 1.67 New Zealand 1.98 1.67 1.57 1.14 1.62 1.76 Croatian 0.52 0.64 0.99 0.49 0.74 0.66 New Zealand 0.81 0.72 0.76 0.38 0.71 0.77 -.10 .00 0.02 -- -.02 -.04 M SD r NOTE: The mean item score was based on a five-point response scale ranging from 1 to 5 for “Very much” to “Not at all”. Cognitive and affective identification correlations A matrix of the correlations between identification scores and affective attachment scores is shown in Table 7. These results indicate that the two sets of scores are statistically and substantially correlated. Participants who identify strongly as Croatian also feel positively Croatian, and those who identify strongly as New Zealanders feel positively New Zealanders. TABLE 7 Correlations between Identification Scores and Affective Attachment Scores Identification as a New Identification as a Affective attachment to Zealander Croatian being a New Zealander -.30* -.04 -.15 Identification as a -.11 Croatian -.13 -.23** .57** -.12 .63** -.02 Affective attachment to .72** .06 being a New Zealander .72** .02 .64** -.06 Ethnic Identity .64** Affective attachment to being a Croatian 15 -.12 -.19 .66** -.10 -.14 .77** .00 -.23 .80** -.02 .03 .55** .19 -.14 .70** .02 -.17* .69** -.04 NOTE: Correlations within each box are ordered from top to bottom as follows: first generation, second generation, third generation, third and fourth generations combined, New Zealand-born (second, third and fourth generations combined), and total sample. * p < .05. ** p < .01 Identification and affective attachment scores for one culture are not mutually exclusive to those for the other culture. The correlations across the cultures are slight or close to zero. The following are weakly related pairs of variables: Croatian identity and New Zealand affect, New Zealand and Croatian affect, New Zealand and Croatian identity, as well as New Zealand identity and Croatian affect. These findings add credibility to the orthogonal model. Generational differences in ethnic and host identity components Mean scores for the two components of ethnic and host identity – cognitive and affective identification – were analyzed by a 2 (identity) x 2 (generation) x 2 (age) MANCOVA with gender as a covariate. No age differences between 2 generation groups – immigrants and NZborn participants – were found; hence age was not added as a covariate in the MANCOVA, but instead as an independent variable with 2 levels (based on a median-split). The mean age differences between the two generation groups was 5 years (Meanimmi=42.53, sd=18.89; MeanNZ-born=37.09, sd=16.44; t153=1.90, p=.059). Ethnic identity MANCOVA. There was a moderate, close to significant multivariate effect for generation (λ=2.59; p=.08, η2=.02), which is due to (young) NZ-born generations identifying themselves ethnically (as Croatians) more strongly than immigrants and older participants (Meanimmi=2.77; MeanNZ-born=1.74; F1,148=4.11; p<.05). This only held significant for the cognitive identification (considering themselves Croatians), but not for the affective identification (extent of positive feeling attached to being Croatian). Host identity MANCOVA. There were both generation and age significant multivariate Ethnic Identity 16 effects. Regarding generation effects (λ=.82; p=.00, η2=.18), immigrant generations (and younger participants) identified themselves with the host identity (as New Zealanders) more strongly than NZ-generations (and older participants), both in terms of a cognitive host identification (Meanimmi=2.77; MeanNZ-born=1.74; F1,149=33.32; p=.00) and an affective host identification (Meanimmi=1.92; MeanNZ-born=1.58; F1,148=10.69; p<.05). Regarding age effects (λ=.93; p=.00, η2=.07), young participants overall identify most strongly with host identity, in terms of both a cognitive identification (Meanyounger=2.32; Meanolder=1.99; F1,149=6.71; p=.01), and an affective identification (Meanyounger=1.88; Meanolder=1.55; F1,149=9.86; p=.00). Ethnic and host identity means across age and generation are shown in Figures 1A through to 1D. FIGURE 1A Ethnic Cognitive Identification between Younger and Older Participants Ethnic cognitive identification 2.0 1.9 1.8 1.7 1.6 Median split (age) 1.5 younger participants 1.4 older participants First generation Generation status NZ-born generations Ethnic Identity FIGURE 1B Ethnic Affective Identification between Younger and Older Participants Ethnic affective identification 1.70 1.68 1.66 1.64 1.62 1.60 1.58 Median split (age) 1.56 younger participants 1.54 1.52 older participants First generation Generation status NZ-born generations 17 Ethnic Identity FIGURE 1C Host Cognitive Identification between Younger and Older Participants Host cognitive identification 3.0 2.5 2.0 Median split (age) 1.5 younger participants 1.0 older participants First generation Generation status NZ-born generations 18 Ethnic Identity 19 FIGURE 1D Host Affective Identification between Younger and Older Participants Host affective identification 2.4 2.2 2.0 1.8 1.6 Median split (age) 1.4 younger participants 1.2 older participants First generation NZ-born generations Generation status DISCUSSION AND CONCLUSIONS Overall, the findings support the proposition that four generations of New Zealanders who are part of a Croatian community identify independently with Croatian and New Zealand cultures. Independent ethnic identities have also been reported by Bottomley (1979), DerKarabetian (1980), Ho (1995), Raza (1997) and Rosenthal and Hrynevich (1985). Correlational analysis of ethnic identity items in the present study, some relating to individuals’ identification with Croatian culture and some to identification with New Zealand culture, were generally unrelated across cultures. Positive identification with Croatian culture is independent of positive identification with New Zealand culture for the participants, other than Croatians in the first (immigrant) generation. The existence of multiple ethnic identities provides support for an orthogonal model of ethnic identity proposed by several researchers (e.g. Berry, 1993; Hutnik, 1991; Rosenthal & Hrynevich, 1985). In support of previous research (e.g. Der-Karabetian, 1980; Zak, 1973), several ethnic identity factors emerged, rather than a single factor. The principal component factor analysis Ethnic Identity 20 revealed four factors. Oblimin rotation was justifiably used, rather than Varimax, as factors one and two (both Croatian identity factors) correlate to a certain extent. The majority of participants in this study had an integrated ethnic identity state (rating themselves highly on both their Croatian and New Zealand identities). This is in line with previous studies using a similar framework (e.g. Berry, Poortinga, Segall, & Dasen, 1992b; Hutnik, 1991; Isajiw, 1990; Raza, 1997; Rosenthal & Hrynevich, 1985). Multivariate analyses provided an interesting finding related to generational identification, both cognitive and affective, with the ethnic and host cultures. Immigrants seem to be motivated and eager to ‘become’ New Zealanders, especially younger immigrants, being more flexible and adaptable to do so. On the other hand, NZ-born participants show a tendency to emphasize their ethnic origins by, paradoxically, ethnically identifying themselves more strongly than immigrants, in terms of (cognitively) considering themselves as Croatians. This is suggestive of NZ-born participants of this study being aware of their ethnic origins and being proud of it. The following conclusions can drawn from the present results: 1. Ethnic identity is multidimensional, as indicated by the four factors extracted by principal factor analysis. 2. Croatian and New Zealand identities are independent for those generations born in New Zealand, and slightly negatively related for first generation immigrants. The negative relationship for the first generation participants may be due to culture shock and acculturative stress associated with immigration-related adaptation (Tsai, Ying, & Lee, 2000). 3. The independence of identification with the New Zealand and Croatian cultures supports an orthogonal identification model and suggests that the participants identify independently with the two cultures. 4. NZ-born participants show cognitive reaffirmation of their ethnic origins by identifying themselves as Croatians more strongly than immigrants in this study, which is suggestive of ethnic awareness and acknowledgement present among a segment of the Croatian community in New Zealand. 5. Immigrants show host identity reaffirmation in terms of both cognition and affect, suggestive of willingness to adapt to the host culture. Ethnic Identity 21 The most important finding of our research suggest that the orthogonal model is valid for native-born generations, that is, for ethnic descendants. The relationship between the Croatian and New Zealand identifications for the first or immigrant generation is weak and slightly negative. We cannot assume that each generation group is homogeneous. For example, within the first generation sub-sample, the standard deviations are large regarding the age of arrival to New Zealand and the years spent in New Zealand. A consideration of cohorts, reflecting different historical contexts, is important. We hope that the present study contributes to an understanding of ethnic identity in terms of its multidimensional and orthogonal characteristics. However, there are several issues that warrant further empirical and theoretical investigation of the ways that they might relate to ethnic identity, such as acculturation, place identity, contextual variations in self-reported ethnicity, and the manner in which ethnic identity changes over time. NOTE 1. A copy of the questionnaire is available from either author on request. Ethnic Identity 22 REFERENCES Aboud, F. E. (1987). The development of ethnic self-identification and attitudes. In J. S. Phinney & M. J. Rotheram (Eds.), Children’s ethnic socialization: Pluralism and development (pp. 32-55). Newbury Park, CA: Sage. Barona, A., & Miller, J. A. (1994). Short acculturation scale for Hispanic youth (SASH-Y): A Preliminary report. Hispanic Journal of Behavioral Sciences, 16(2), 155-162. Barth, F. (1969). Ethnic groups and boundaries. Boston: Little Brown. Berry, J., Poortinga, Y., Segall, M., & Dasen, P. (1992). Ethnic groups and minorities. New York: Cambridge University Press. Bernal, M. E., & Knight, G. P. (1993). Introduction to ethnic identity. In M. E. Bernal & G. P. Knight (Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities (pp. 1-10). New York: State University of New York Press. Berry, J., Poortinga, Y., Segall, M., & Dasen, P. (1992a). Acculturation and culture contact. In J. Berry, Y. Poortinga, M. Segall ,& P. Dasen (Eds.), Cross-cultural psychology: Research and applications(pp. 271-291). New York: Cambridge University Press. Berry, J., Poortinga, Y., Segall, M., & Dasen, P. (1992b). Ethnic groups and minorities. New York: Cambridge University Press. Berry, J. W. (1993). Ethnic identity in plural societies. In M. E. Bernal & G. P. Knight (Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities .(pp. 271-296). Albany, NY: State University of NY Press. Berry, J. W., Kim, U., Power, S., Young, M., & Bujaki, M. (1989). Acculturation attitudes in plural societies. Applied Psychology: An International Review, 38, 185-206. Bond, M. H., & Yang, K. S. (1982). Ethnic affirmation versus cross-cultural accommodation. The variable impact of questionnaire language on Chinese bilinguals from Hong Kong. Journal of Cross-Cultural Psychology, 13(2), 169-185. Bottomley, G. (1979). The study of social processes: Some problems of theory and method. In P. R. de Lacey & M. E. Poole (Eds.), Mosaic or melting pot: Cultural evolution in Australia .(pp. 71-76). Sydney: Harcourt Brace Jovanovich Group Pty Ltd. Ethnic Identity 23 Buriel, R., & Cardoza, D. (1993). Mexican American ethnic labeling: an intrafamilial and intergenerational analysis. In M. E. Bernal & G. P. Knight (Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities (pp. 197-210). Albany, NY: State University of NY Press. Cuellar, I., Arnold, B., & Maldonado, R. (1995). Acculturation rating scale for Mexican Americans-II: A revision of the original ARSMA scale. Hispanic Journal of Behavioral Sciences, 17, 275-304. Der-Karabetian, A. (1980). Relation of two cultural identities of Armenian-Americans. Psychological Reports, 47, 123-128. DeVos, G. A., & Romanucci-Ross, L. (1995). Ethnic Identity: A Psychocultural Perspective. In L. Romanucci-Ross & G. A. DeVos (Eds.), Ethnic Identity: Creation, conflict , and accommodation (pp. 349-381). Walnut Creek: AltaMira Press. George, D., & Mallery, P. (1999). SPSS for Windows: Step by step. Needham Heights, MA: Allyn & Bacon. Ho, E. S. (1995). The Challenge of Culture Change. The Cross-Cultural Adaptation of Hong Kong Chinese Adolescent Immigration in New Zealand. Unpublished A doctoral thesis, The University of Waikato, Waikato. Hocoy, D. (1996). Empirical distinctiveness between cognitive and affective elements of ethnic identity and scales for their measurement. In H. Grad, A. Blanco & J. Georgas (Eds.), Key issues in cross-cultural psychology (pp. 128-137). Lisse, Netherlands: Swets and Zeitlinger. Hogg, M. A., & Abrams, D. (1988). Social identifications. A social psychology of intergroup relations and group processes. London: Routledge. Hutnik, N. (1986). Patterns of ethnic minority identification and modes of social adaptation. Ethnic and Racial Studies, 9, 150-167. Hutnik, N. (1991). Ethnic minority identity. Oxford: Clarendon Press. Isajiw, W. W. (1990). Ethnic identity retention. In R. Breton, W. W. Isajiw, W. E. Kalbach, & J. G. Reitz (Eds.), Ethnic identity and equality: Varieties of experience in a Canadian city. (pp. 34-91). Toronto: University of Toronto Press. Ethnic Identity 24 Keefe, S. E., & Padilla, A. M. (1987). Chicano ethnicity. Albuquerque: University of New Mexico Press. Knight, G. P., Bernal, M. E., Cota, M. K., Garza, C. A., & Ocampo, K. A. (1993). Family socialization, identity and behaviour. In M. E. Bernal & G. P. Knight (Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities (pp. 105129). New York: State University of New York Press. Kuhn, M. H., & McPartland, T. S. (1954). An empirical investigation of self-attitudes. American Sociological Review, 19, 68-76. Liebkind, K. (1992). Ethnic identity-challenging the boundaries of social psychology. In G. M. Breakwell (Ed.), Social psychology of identity and the self concept (pp. 147-185). San Diego: Surrey University Press. Marin, G., Sabogal, F., Marin, B. V., Otero-Sabogal, R., & Perez-Stable, E. J. (1987). Development of a short acculturation scale for Hispanics. Hispanic Journal of Behavioral Sciences, 9, 183-205. Masuda, M., Matsumoto, G. H., & Meredith, G. M. (1970). Ethnic identity in threegenerations of Japanese Americans. The Journal of Social Psychology, 81, 199-207. Phinney, J. S. (1993). Ethnic identity development in adolescence. In M. E. Bernal & G. P. Knight (Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities. New York: State University of New York. Mavreas, V., Bebbington, P., & Der, G. (1989). The structure and validity of acculturation. Social Psychiatry and Psychiatric Epidemiology, 24, 233-240. McKay, J., & Lewins, F. (1978). Ethnicity and the ethnic group: a conceptual analysis and reformulation. Ethnic and Racial Studies, 1, 412-425. Oetting, E. R., & Beauvais, F. (1991). Orthogonal cultural identification theory: The cultural identification of minority adolescents. International Journal of the Addictions, 25, 655685. Pawliuk, N., Grizenko, N., Chan-Yip, A., Gantous, P., Mathew, J., & Nguyen, D. (1996). Acculturation style and psychological functioning in children of immigrants. American Journal of Orthopsychiatry, 66, 111-121. Ethnic Identity 25 Phinney, J. S. (1993). A three-stage model of ethnic identity development in adolescence. In M. E. Bernal & G. P. Knight (Eds.), Ethnic identity: Formation and transmission among Hispanics and other minorities (pp. 61-79). New York: State University of New York. Pruegger, V. J., & Rogers, T. B. (1993). Development of a scale to measure cross-cultural sensitivity in the Canadian context. Canadian Journal of Behavioural Science, 25, 615621. Raza, F. (1997). Ethnic identity, acculturation, and intergenerational conflict among second generation New Zealand Indians. Unpublished Master of Arts Thesis, The University of Auckland, Auckland. Rosenthal, D., & Hrynevich, C. (1985). Ethnicity and ethnic identity: A comparative study of Greek-, Italian-, and Anglo-Australian adolescents. International Journal of Psychology, 20, 723-742. Salvaterra, D. L. (1994). Becoming American: Assimilation, pluralism, and ethic identity. In T. Walch (Ed.), Immigrant America: European ethnicity in the United States. New York: Garland Publishing, Inc. Sayegh, L., & Lasry, J. C. (1993). Immigrants' adaptation in Canada: Assimilation, acculturation, and orthogonal cultural identification. Canadian Psychology, 34, 98-109. Saylor, E. S., & Aries, E. (1999). Ethnic identity and change in social context. Journal of Social Psychology, 139, 549-566. Sommerlad, E. A, & Berry, J. W. (1970). The role of ethnic identification in distinguishing between attitudes towards assimilation and integration of a minority racial group. Human Relations, 23, 23-29. Stephan, C. W., & Stephan, W. G. (1989). After intermarriage: Ethnic identity among mixed heritage Japanese-Americans and Hispanics. Journal of Marriage and the Family, 51, 507519. Suinn, R. M., Rickard-Figueroa, K., Lew, S., & Vigil, P. (1987). The Suinn-Lew Asian SelfIdentity Acculturation Scale: An initial report. Educational and Psychological Measurement, 47, 401-402. Ward, C., & Rana-Deuba, A. (1999). Acculturation and adaptation revisited. Journal of Cross-Cultural Psychology, 30, 422-442. Ethnic Identity 26 Zak, I. (1973). Dimensions of Jewish-American identity. Psychological Reports, 33, 891-900.