Discussion

EthnicIdentityScale.pdf

Home >Human Resource Management homework help >Discussion

The Development of the Universal Ethnic Identity Scale (UEIS) for use in an Australian contextap_40 238..248 Rosanna Rooney,1 Drew Nesdale,2 Robert Kane,1 John Hattie,3 and Romesh Goonewardene4

1School of Psychology and Speech Pathology, Curtin University, 2School of Psychology, Griffith University, 3School of Teaching, Learning, and Development,

University of Melbourne, and 4School of Architecture, Landscape, and Visual Arts, The University of Western Australia

The aim of the present research was to develop a measure that assessed the level of ethnic identification of adults from diverse ethnic backgrounds, as well as individuals’ level of ethnic identification with a host culture following migration. In the first study, an ethnically diverse sample of 275 people (138 males and 137 females) responded to a range of new items, as well as items drawn from related scales. Factor analysis revealed three dimensions of ethnic identity that were labelled, “Pride in Ethnic Background and Language,” “Liking for Traditional and Social Activities of my Ethnic Group,” and “Sense of Belonging to This (host) Country.” This factor structure was evaluated in a second study with 1,007 (552 males, 455 females) adults from culturally and linguistically diverse backgrounds. A two-factor structure provided the most parsimonious fit to the data but further investigation of the factor structure within ethnically diverse samples was recommended. The results are discussed in relation to theories of social identity as well as research on ethnic identification and adaptation.

Key words: adaptation; belonging; ethnic identity; migration.

What is already known on this topic

1 Ethnic identity is associated with adaptation in migrant groups. 2 The MIEM is an existing ethnic identity measure. 3 The MEIM-R is an existing ethnic identity measure.

What this paper adds

1 An ethnic identity measure that can be used in Australia. 2 An ethnic identity measure that can be used with adults in

Australia. 3 An ethnic identity measure that can be used with South East

Asian groups in Australia.

Geographic relocations of ethnic populations have occurred throughout history. People from almost all cultural backgrounds have moved in small to very large groups over short to very great distances to take up residence in new geographic and cultural locations. Regardless of their reasons for relocation, one conse- quence for all migrating groups is that they are faced with the task of adapting to the demands of a new culture and coping with the attendant psychological effects (Berry, Phinney, Sam, & Vedder, 2006; Nesdale, Rooney, & Smith, 1997). Further complicating their situation, however, is the fact that immigrants typically arrive in a host country having been enculturated with the language, norms, beliefs, mores, and traditional ways of their home country. Clearly, there is considerable room for misunder- standing, incongruity, confusion, embarrassment, and anxiety when traditional ways are juxtaposed against new ways.

Berry and his colleagues (Berry, 1997; Berry et al., 2006; Berry, Poortinga, Segall, Dasen, 1992; Sam & Berry, 1993; Sands & Berry, 1993) have emphasised the central role of the immi- grant’s maintenance or rejection of his ethnic culture in the

acculturation strategy s/he adopts after immigration. Depending upon whether the individual chooses to (1) maintain or (2) reject his/her cultural identity and characteristics, and (3) main- tain or (4) reject relationships with other groups in the host culture, Berry and colleagues consider that immigrants adopt an acculturation strategy of integration (1–3), assimilation (2–3), separation/segregation (1–4), or marginalisation (2–4). Consis- tent with their typology, research has revealed that while some immigrants do seek to assimilate, the preferred and more psy- chologically healthy strategy of the majority of immigrants is integration; that is, to retain (at least, elements of) their home/ ethnic culture and to engage with and adopt the new culture (Berry et al., 2006; van Oudenhoven & Eisses, 1995; Schmitz, 1995). In addition, Phinney, Horenczyk, Liebkind & Vedder (2001) have proposed an interactional model suggesting the central importance of both identification with ethnic and national cultures for the impact on psychological adaptation in immigrants.

On the one hand, immigrant’s maintenance of their ethnic identification (i.e., the extant to which he/she endorses the attitudes, values, beliefs, and behaviours of his/her cultural or ethnic group) post-immigration and indeed, even their progres- sive adoption of aspects of the new host culture, are unsurpris- ing in the light of social identity theory (SIT; Tajfel & Turner, 1986) and related research. The theory emphasises the impor- tance to individuals of their identification with particular social groups, the effect it has on their levels of self-esteem, and the

Correspondence: Rosanna Rooney, Department of Psychology, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia. Fax: +618 9266 2464; email: [email protected]

Accepted for publication 29 May 2011

doi:10.1111/j.1742-9544.2011.00040.x

bs_bs_banner

238

impact it has on their inter- and intra-group cognitions, atti- tudes, and behaviour (e.g., Hogg, Turner, Nascimento-Schulze & Spriggs, 1986; Lemyre & Smith, 1985).

On the other hand, however, there has been surprisingly little research into issues of ethnic identity maintenance, change, and adoption in relation to adult immigrants by cross-cultural or social identity researchers (Liebkind, 1996; Sands & Berry, 1993). However, some researchers have focussed on the devel- opment and implications of ethnic identity in ethnic adolescents (often second generation; e.g., Liebkind, Jasinskaja-Lahti & Solheim, 2004; Phinney, 1992; Phinney & Chavira, 1995; Phinney, Chavira, & Williamson, 1992; Phinney & Ong, 2007; Spencer, Icard, Harachi, Catalano, & Oxford, 2000).

A major issue limiting such research has been the urgent need for a reliable and valid instrument that can measure the ethnic identification of adults from a range of different ethnic backgrounds. While some researchers have simply based ethnic identity on self-definition (e.g., Ullah, 1985, 1987), others have attempted to use more psychometrically based measures, typically based on factor analysis (e.g., Driedger, 1976; Hogg, Abrams, & Patel, 1987; Kwan, 2000; Leclezio, Louw-Potgieter & Souchon, 1986; Phinney, 1992; Rosenthal & Hrynevich, 1985; Whittler, Calantone, & Young, 1991). Factor analysis of items hypothesised to measure ethnic identity enables researchers to identify dimensions that are crucial to the assessment of the concept. The range, number, and type of items that researchers have initially hypothesised to assess ethnic identity have varied a great deal in these studies (e.g., Hogg et al., 1987; Leclezio, Louw-Potgieter & Souchon, 1986; Phinney, 1992; Rosenthal & Hrynevich, 1985; Whittler et al., 1991). In addition, given the wide range of groups that have been investigated in a relatively small group of studies, it is not surprising that the number of factors, the nature of the factors, and subsequent factor names have varied. The type of factor extraction technique used in such studies has commonly been principal components with occasional exceptions, for example, Whittler et al. (1991), who used a maximum likeli- hood method but this measure was not suitable for a range of ethnically diverse people. The use of maximum likelihood enables the researcher a greater chance of examining under- lying dimensions of concepts being measured rather than simply summarising what exists in the data. In sum, a measure of ethnic identity that is suitable for adults from a range of cultural backgrounds, which also uses a maximum likelihood method of factor extraction is needed.

To date, the only factor analytically derived instrument that approaches the criterion of versatility among ethnic groups was the Multigroup Ethnic Identity Measure (MEIM) developed by Phinney (1992) and more recently the MEIM-R (Phinney & Ong, 2007). Confirmatory factor analysis of the MEIM has revealed that the fit of the two-factor model of the MEIM is only moderate with a goodness of fit index (GFI) = 0.84 although further investigation is needed in more diverse groups (Ponterotto, Gretchen, Utsey, Stracuzzi, & Saya, 2003). Phinney and Ong (2007) made modifications and provided evidence of a better fit for the two-factor model with the MEIM-R. However, that instrument was originally designed for use with adolescents and strongly focuses on measuring aspects of the development of ethnic identity, as well as positive feelings about the in-group.

In addition, the “Other Group Orientation” factor, which can be used alongside ethnic identification, includes items which are not specific about which group a person may be identifying with. Similarly, the scale developed by Rosenthal and Hyrnevich (1985), was explicitly designed to measure and compare the salient dimensions of ethnic identity for Greek and Italian adolescents. Aside from these approaches, Driedger (1976) has carried out two studies, which have sought to develop a scale to measure the ethnic identity of members of a number of Euro- pean groups. However, the participants were all university stu- dents (at the University of Manitoba) and were mainly from second-generation ethnic backgrounds.

The aim of the present studies was to develop a psychometri- cally sound measure of ethnic identity that could be used with people from of a wide variety of ethnic groups. The instrument was designed to tap into the identity issues important to first- and second-generation immigrants, both in relation to their home and specifically their host country. Consequently, it was expected that issues relating to ethnic identity would be more salient as the boundaries between recently arrived groups and the host country would be more obvious in terms of aspects such as cultural practices and language. The underlying theoretical framework guiding the research was based on SIT (Tajfel & Turner, 1986), which specifies that a measure of (ethnic) identity should include cognitive, behavioural, and affective components. According to this theory, ethnic identity is one component of a person’s social identity, others being, for example, social identity based on gender, occupation, and reli- gion. Ethnic identity is linked to self-esteem because social identity is that part of an individual’s self-image that derives from his or her knowledge of membership in a social group or groups (e.g., ethnic group) together with the value and emotional significance attached to that group. When indivi- duals identify with an ethnic group, they accept its concomi- tant values and behaviours as their own and this, in turn, comprises a source of their self-esteem. According to SIT, apart from establishing which ethnic group an individual identifies with (self-definition), for example, which group is “my ethnic group,” conative, behavioural, and affective elements should always be included in a measure of identity (Tajfel & Turner, 1986). These elements have been incorporated in at least two studies measuring ethnic identity (Hogg et al., 1987; Ullah, 1987).

As well as SIT, items that were included represented a range of the dimensions involved in group identity outlined by Ashmore, Deaux, and McLaughlin-Volpe (2004) and sum- marised by Phinney and Ong (2007). These included self- categorising and labelling (e.g., showing they identified with their group), commitment and attachment (e.g., a sense of belonging to the group), exploration (e.g., reading books written by those from their ethnic group), ethnic behaviours (e.g., speaking the language and eating the food of their ethnic group), evaluation and in-group attitudes (e.g., pride in being a member of their ethnic group), values and beliefs (e.g., marry- ing people from different ethnic groups, traditional child rearing), and national (or American) identity (e.g., items per- taining to the host or in this case “Australian” identity). Clearly, the possible items and dimensions involved for migrants in pluralistic societies is multifaceted and potentially more

R Rooney et al. Universal Ethnic Identity Scale (UEIS)

239

complex than the three areas of cognitive, behavioural, and affective components outlined by Tajfel and Turner (1986). Our approach to constructing the actual measure was also heavily influenced by the four-stage model proposed by Hattie (1981), which includes exploratory analysis, development of factor names that are theoretically sound and valid, assessing goodness of fit with confirmatory factor analysis, and validation on a separate sample of participants. This four-stage model enables the development of a psychometrically sound measure of ethnic identity, which has been lacking in an Australian context as it involves exploratory and confirmatory factor analysis with vali- dation of the factor structure on a different sample of partici- pants. Consequently, in the current study, two phases were involved. Phase 1, using participants from a variety of ethnic groups, incorporated the first three stages of Hattie’s factor ana- lytic approach. These are, first, exploratory analysis to assess the number of factors; second, developing factor names based on theoretical arguments and ensuring that each factor consists of subscales that researchers agree appropriately load on the factor; and third assessing goodness of fit using factor analysis. Given the wide range of items from the literature that were included as possible items measuring ethnic identity, and consequently, given the indefinite number of factors, following Hattie’s (1981) first stage of constructing a measure, the number of factors was not initially specified in Phase 1. This allowed an exploration of the underlying dimensionality of the items. Phase 2, involved the fourth and final stage of Hattie’s (1981) factor analytic approach was designed to confirm the factor structure on a separate sample of subjects.

Phase 1: Determination of Ethnic Identity Factors

Method

Participants

A sample of 275 adults (138 males and 137 females) was selected so as to reflect the diverse range of ethnic groups in Perth, Western Australia. The participants included all those who completed usable questionnaires. The participants included respondents from Vietnamese (n = 30), Chinese (n = 57), Malaysian (n = 6), Indian (n = 7), Sri Lankan (n = 5), Italian (n = 34), Greek (n = 12), Macedonian (n = 12), Portu- guese (n = 9), Polish (n = 9), Yugoslav (n = 8), Chilean (n = 5), and 81 participants from a variety of other groups. They ranged in age from 18 to 65 years, with a mean of 35.6 years. For age group, there were 33% in the 18–28 category, 35% in the 20–39 category, 14.5% in the 40–49 category, and 13.8% in the 50–65- year-old category, with 4% not identifying their age. There were 18.1% respondents who were professional, 1% executive, 14.9% clerical, 24% skilled 9.4% and 32.3% in the other or missing category. In terms of years lived in Australia, 37.4% had been resident for up to 5.9 years, 13.8% for up to 11.9 years, 25.1% for up to 17.9 years, 27% for up to 34.9 years, 4.3% for up to 35+ years, and 4.3% who were missing this information. Participants had spent a varying amount of time in Australia and thus representing first and second generations, which would provide a sample of participants that could be generalised to more populations. The first-generation migrants would be more

likely to speak in their language of origin and would therefore from a SIT perspective (Tajfel & Turner, 1986), require items pertaining to behaviour through language to be included, for example speaking their language of origin, for the ethnic iden- tity questionnaire. For second-generations participants, lan- guage items would be less likely to be endorsed but other items such as those pertaining to culture and traditions can still be strongly endorsed in later generations and indeed have been connected to positive psychological and socio-cultural adapta- tion outcomes (Berry et al., 2006).

Scale development

Items for the ethnic identity questionnaire were prepared on the basis of previous research (e.g., Driedger, 1976; Phinney, 1990, 1992; Rosenthal & Hrynevich, 1985) and theory (Lewin, 1948; Tajfel & Turner, 1986) as well as through discussions with tertiary students from a range of ethnic minority groups and with professionals working with members of culturally diverse groups. Item selection was also influenced by Tajfel and Turner’s (1986) view that a measure of (ethnic) identity should include cognitive, behavioural, and affective components. Items included in the scale reflecting a cognitive component included identifying being a member of their ethnic group, behaviours related to aspects such socialising with members of their ethnic group and eating the food of their ethnic group, while affective items included items related to feelings such as a sense of pride or anger towards their ethnic group. In addition, a review of factor analytic studies of ethnic identity revealed that common factors include social interaction, in-group identification (cul- tural beliefs and practices, social interaction, importance of ethnic background, pride, and affirmation), out-group identifi- cation and denial, and language. Finally, items that were included are consistent with a range of the dimensions involved in group identity outlined by Ashmore et al. (2004) and sum- marised by Phinney and Ong (2007). These included self- categorising and labelling, commitment and attachment, exploration, ethnic behaviours, evaluation, and in-group atti- tudes, values and beliefs, and national (or American identity). On this basis, 50 statements were prepared to measure aspects of a number of dimensions including sense of pride and belong- ing to an ethnic group (in-group identification), sense of pride and belonging to Australia (out-group identification), attempts at hiding ethnic group membership (denial), social and business interaction, participation in various cultural activities, impor- tance of ethnic background, use of the ethnic language, affective indices of ethnic identity using words such as “anger” and “annoys,” degree of perceived difference between self and others in Australia and between self and others in own ethnic group, child-rearing practices, and intermarrying. Given the wide range of items and factors outlined in the literature and theory, to allow for an unknown number of dimensions and in line with Hattie’s first exploratory stage where the dimension- ality of the factor structure could be explored, the number of factors was not specified.

Each item was worded as a statement to which the subject was required to rate her/his strength of agreement on a 5-point bipolar scale ranging from “strongly agree” to “strongly dis- agree.” As well as including items with negative affect such as

Universal Ethnic Identity Scale (UEIS) R Rooney et al.

240

“anger,” 14 items were reworded in the negative so as to reduce the incidence of positive demand characteristics. Within each item, the subject’s ethnic group was described as “my ethnic group” so that each item was relevant to all ethnic groups. Given the number of different languages spoken by the partici- pants, the questionnaire was written in English and the partici- pants were required to be proficient in reading and writing English, which was decided upon by experienced migrant workers who handed out the questionnaire and checked by the researcher by the clarity of written expression on the question- naires. The completed scale was read by a number of individuals who had English as a first or second language in order to enhance the clarity and focus of the items.

Procedure

Prospective participants were approached at various ethnic social clubs and migrant resource centres and it was explained that the questionnaire was part of a research project investigat- ing ethnic attitudes and interests. The importance of the research was described and any questions about the project were answered. Participants then filled out the questionnaire in a quiet location in the centre. The researcher clarified any issues of meaning as they arose and had access to individuals who could provide special language assistance, should it be required. The questionnaire took approximately 20 min to complete and participants were thanked for taking part upon completion.

Results

Participants’ responses to the 50 items of the proposed Universal Ethnic Identity Questionnaire were factor analysed using a maximum-likelihood method of factor extraction. The aim was to test the dimensionality of the items and to allow for a range of factors outlined by Tajfel and Turner (1986) and other researchers (Ashmore et al., 2004; Phinney & Ong, 2007), in line with Hattie’s first exploratory phase of scale development, the number of factors was not specified. Five factors provided the best fit to the data and accounted for 32% of the total variance. The fifth factor had two items loading at approxi- mately 0.4 but there was no obvious meaning, and thus, this factor was not included in the further analysis. As the oblique solution indicated that the correlations between factors were small (range -0.31 to 0.15, mean = 0.11), the more parsimoni- ous varimax solution was used. Factor names were tentatively given to each of the four factors.

Following Hattie (1981), during this part of the process, five additional independent researchers were asked to assign the 50 items to the tentatively named factor labels. The aim here was to ensure face validity of factors by having respondents agree that each factor consists of items that appropriately load on the factor. The individuals who participated were all university academics and all had PhDs. This process was repeated twice, each time modifying the names and explanations for the purpose of greater clarification because the respondents appeared to be confusing factors 2 and 3. The aim here was to make modifications to the names until there was relatively little confusion (less than 20% false positives or false negatives) regarding the appropriate items being assigned to the correct factor label. This process was part

of the second stage of factor analysis recommended by Hattie (1981), which ensures face validity of the naming of these factors. In this stage, researchers are asked to agree that each factor consists of items that appropriately load on the factor.

After the factors had been named and modified twice, the percentage agreement was 70% or greater both for factor 1 (Pride in Ethnic Background and Language) and factor 4 (Sense of Belonging to “This Country”—“This Country” meaning Aus- tralia). However, the percentage agreement for factor 2 (Degree of Distance Required from Ethnic Group and Background) was approximately 50% and although percentage agreement for factor 3 (Degree of Liking for Traditional Activities of my Ethnic Group) was high (66–100%), a relatively high number of factor 3 items were assigned to factor 2. Thus, a further maximum likelihood analysis was run using items from factors 2 and 3, specifying two factors, using varimax rotation. Inspection of a plot of the items showed that items from factor 3 clustered at the opposite end of the dimension of factor 2 and that this polarity had apparently occurred due to the negative wording of most of the items in factor 2 as opposed to the positive wording of the items in factor 3. A plot of the negatively worded items from factor 2 against the positively worded items from factor 3 is shown in Figure 1.

HORIZONTAL FACTOR 1 VERTICAL FACTOR 2

= Items from Factor 2 = Items from Factor 3

Figure 1 A Plot of the Negatively Worded Items From Factor 2 Against the Positively Worded Items From Factor 3.

R Rooney et al. Universal Ethnic Identity Scale (UEIS)

241

Inspection of the plot reveals that the negatively and posi- tively worded items clustered at opposite ends of the same dimension suggesting that they measure the same thing. Con- sequently, items in factor 2 (negatively worded items) were dropped. The top five loading items from factor 2 where four of the five items are negatively worded are shown in Table 1.

The final version of the three dimensions of ethnic identity contained 22 items, the remaining 28 items being dropped from the questionnaire. Table 2 shows the top five loading items of factor 3, all of which are positively worded, in the factor analysis of items from factor 2 and factor 3.

As a result of the overlap between items from factor 2 and factor 3, all items from factor 2 were dropped and a final maximum likelihood analysis was run using the 22 items from the remaining three factors using varimax rotation. Three meaningful factors emerged accounting for 30% of the total variance. “Pride in Background and Language” accounted for 17.6% of the variance (alpha = 0.77). “Degree of Liking for Traditional and Social Activities of my Ethnic Group” accounted

for 8.8% of the variance (alpha = 0.77), and “Sense of Belong- ing to This Country” accounted for 4.4% of the variance (alpha = 0.75). Given this reduction in factors from 4 to 3, the percentage of agreement when assigning the items to the factors ranged from 75% for factor 1 to 65% for factor 3. The results of a varimax rotation for the ethnic identity measure are shown in Table 3.

Discussion

The results of the factor analyses and naming exercises of the items from the ethnic identity questionnaire showed that ethnic identity was represented by three independent dimensions which were named “Pride in Ethnic Background and Lan- guage,” “Liking for Traditional and Social Activities of my Ethnic Group” and “Sense of Belonging to This Country” (in this case, Australia). The two highest loading items on the “Pride in Ethnic Background and Language” factor primarily involved pride in being a member of an ethnic group making pride in membership an important element of this factor. Other items included the importance of using ethnic language, as well as use of ethnic names, and books and newspapers being written in the ethnic language. As well as several items relating to the impor- tance of ethnic background, the independent researchers who were given the task of assigning items to names interpreted all of the factors to relate to a general pride in ethnic background and language.

Table 1 Negative Wording of the Top Five Loading Items From Factor 2 in the Factor Analysis of Items From Factor 2 to factor 3

Item name Factor loadings

Factor 2 Factor 3

I don’t like being recognised as a member of my

ethnic group.

0.61 -0.10

I am not interested in visiting the country from

which my ethnic group came.

0.61 -0.02

I feel great pride in being a member of my ethnic

group

-0.56 -0.36

I often don’t know whether to tell people about

my ethnic background

0.50 0.05

I dislike the cultural activities of my ethnic group. 0.49 -0.23

Note. Factor 2: Degree of distance required from ethnic group to

background.

Table 2 Positive Wording of the Top Five Loading Items From Factor 3

Item names Factor loadings

Factor 2 Factor 3

I like to socialise mainly with members of my

ethnic group.

0.06 0.73

I like to stick together with those of my own

ethnic group.

0.01 0.73

I prefer food to be prepared and served

according to the traditional ways of my ethnic

group.

-0.09 0.56

If I married someone from outside my ethnic

group I would insist hey would be raised

according to the traditions of my ethnic group.

-0.08 0.53

I like to carry on the traditional ways of my ethnic

group.

-0.24 0.49

Note. Factor 3: Degree of liking for traditional and social activities of my

ethnic group.

Table 3 Factor Loadings of Varimax Rotation for Three Dimensions of the Universal Ethnic Identity Questionnaire

Variable Factor 1 Factor 2 Factor 3

Membership pride 0.73 0.12 0.09

Pride in telling 0.62 0.01 0.16

Use ethnic language 0.58 0.22 0.02

Anger others ashamed 0.47 0.04 0.08

Anger don’t use names 0.47 0.32 -0.07 Books and newspapers 0.45 0.25 -0.16 Hide background 0.44 0.04 0.15

Others learn language 0.41 0.19 -0.04 Ethnic teach children 0.40 0.29 -0.11 Child ethnic language. 0.36 0.24 0.15

Confuse ethnic group 0.36 0.20 0.14

Not ethnic school 0.36 -0.03 0.16 Religious values 0.28 0.17 0.16

Socialise with ethnic 0.21 0.69 -0.10 Stick with ethnic 0.28 0.66 -0.09 Traditional food 0.15 0.57 -0.03 Outside raised ethnic 0.20 0.51 0.03

Business with ethnic 0.11 0.46 -0.13 Traditional ways 0.27 0.42 -0.00 Ethnic raise children 0.24 0.39 -0.01 Marry different ethnic 0.03 0.36 0.00

Enjoy ethnic food 0.05 -0.35 0.11 Tell members I belong 0.08 -0.14 0.74 Belong “this country” -0.06 -0.04 0.73 Clear I belong -0.04 0.07 0.73 Not sure I belong -0.02 0.23 -0.45 Proud history traditions 0.01 -0.05 0.40

Universal Ethnic Identity Scale (UEIS) R Rooney et al.

242

Items which loaded on “Liking for Traditional and Social Activities of my Ethnic Group” covered a range of topics tapping traditional and social issues. For example, social variables included preferences for socialising and doing business together with ethnic group members and a general feeling of sticking together. Traditional activities included eating ethnic food, following ethnic marriage, and child-rearing practices.

Items loading on “Sense of Belonging to This country” included making it clear to both ethnic group members and general members of the community that ethnic group members belong to the host country (Australia) as well as the importance of feeling a strong sense of belonging to the host country. There was also an item investigating attitudes towards the histories and traditions of the host country (Australia).

The three dimensions revealed represented three aspects of ethnic identity measured in a new Universal Ethnic Identity Scale. Further, the results of the factor analysis suggested that the dimensions of the new scale were independent—a partici- pant’s score may vary from high to low on each dimension. For example, a subject may report a strong sense of “Pride in Ethnic Background and Language,” while displaying only a moderate degree of “Liking for Traditional and Social Activities of my Ethnic Group,” while they may also have a strong “Sense of Belonging to This country” 0.30% of the variance was accounted for suggesting that there were other factors not being accounted for by the data. This result is reasonable given that the questionnaire was assessing for common variance across a large range of ethnic groups, and also suggests that validation of the factor structure in a different sample is important.

In line with Hattie’s (1981) recommendation, a second phase was carried out to determine if the factor structure would be replicated with a different sample of respondents. A further sample containing a range of ethnically diverse groups was chosen as this was similar in type to the participants in Phase 1 and was therefore deemed suitable to assess the factor structure of the new universal ethnic identity scale. In addition, Phase 2 allowed us to assess the psychometric properties of the universal ethnic identity scale. This assessment of the scale is important as it has been used as a measure of ethnic identity in published work to test a model of migrant distress where the influence of ethnic identity was investigated (reported elsewhere, Nesdale et al., 1997).

Phase 2: A Cross-validation Study

The second study was carried out to cross validate the factor structure and items of the Universal Ethnic Identity Scale on a

different sample of ethnically diverse participants. This was the final stage in the four-stage procedure suggested by Hattie (1981) for use in factor analysing behavioural domains. The aim of this study was to examine whether the structure of ethnic identity that emerged as a result of the first study would be revealed with a different sample of participants.

Method

Participants

A convenience sample of 1,007 adults from culturally and linguistically diverse backgrounds participated in this study. Participants were recruited in various ways. For example, the Vietnamese participants were recruited through migrant resource centres and using a snowballing technique starting with the Vietnamese research assistant who worked on this project. The participants from the other ethnic groups were recruited similarly with the research assistant being the starting point for the snowballing technique. Their ethnicity, gender, and age are shown in Table 4.

Measures

A questionnaire was prepared that contained two scales, the UEIS and the MEIM. The MEIM was included to assess the validity of the UEIS via the correlations between the UEIS and the MEIM. The questionnaire also contained instructions for completing the scale, as well as several questions that addressed the participants’ demographic characteristics.

Universal Ethnic Identity Scale. The 22 items that were identified in study one were incorporated into a questionnaire. Partici- pants recorded the strength of their agreement to the statements using a Likert response format ranging from 1 (strongly agree) to 5 (strongly disagree).

Multigroup Ethnic Identity Measure (MEIM, Phinney, 1992). The questionnaire also included the 20-item MEIM. The MEIM was developed to measure ethnic identity in adolescents and young adults. In its original form, the 20-item MEIM comprised four subscales, namely Affirmation and Belonging (five items), Ethnic Identity Achievement (seven items), Ethnic Behaviours (two items), and Other Group Orientation (six items). Subse- quent factor analysis demonstrated that a two-factor solution was optimal in which items from the first three subscales loaded on a single factor, and the fourth subscale remained intact. Thus,

Table 4 Ethnic and Gender Representation of Participants

Ethnicity Frequency Per cent Men Women Mean age and (SD)

Italian 276 27.4 181 (65.6%) 95 (34.6%) 26 years (7.48)

Vietnamese 446 44.3 250 (56.1%) 196 (43.9%) 27 years (10.61)

Farsi 224 22.2 100 (44.2%) 124 (55.8%) 43 years (16.01)

Croatian 61 6.1 21 (34.4%) 40 (65.6%) 21 years (3.28)

Total 1,007 100.00 552 455

R Rooney et al. Universal Ethnic Identity Scale (UEIS)

243

the MEIM is most often used as a two-dimensional scale mea- suring Ethnic Identity and Other Group Orientation. The MEIM has been used with adult populations (e.g. Cuellar, Nyberg, Maldonado & Roberts, 1997; Lorenzo-Hernandez & Oullette, 1998), showing high internal consistency of the subscales. Phinney also demonstrated the construct validity of the MEIM by relating the scale significantly to variables with which it should theoretically correlate.

Procedure

Participants were invited to participate and were informed that the research project involved investigating ethnic attitudes and interests, involvement was voluntary, that their responses were confidential, and that there were no right or wrong answers. The completed questionnaires were posted back to the research- ers, collected personally, or placed in strategically placed deposit boxes.

Results

A maximum likelihood confirmatory factor analysis using LISREL (Version 8.7; Scientific Software International Inc., Lincoln, IL, USA; Jöreskog & Sörbom, 2004) was conducted in order to cross validate the factor structure found in Phase 1. This technique enables a model testing by providing a variety of goodness-of-fit measures. Because different fit indices evaluate model fit from slightly different perspectives, more than one fit index is generally reported. The present study reported the following fit statistics: The comparative fit index (CFI), the non- normed fit index (NNFI), the standardised root mean square residual (SRMR), and the root mean square error of approxi- mation (RMSEA). The CFI and the NNFI compare the chi- square value of the null model to the chi-square value of the hypothesised model, a value greater than or equal to 0.9 indi- cates a good fit (Benet-Martínez & Karakitapoglu-Aygun, 2003; Tabachnick & Fidell, 2007). The SRMR is considered to be one of the more meaningful fit indices to report in a CFA analysis (Hooper, Coughlan, & Mullen, 2008). It measures the square root difference between the residuals of the null model and the hypothesised model. An SRMR of less than or equal to 0.1 is required for the SRMR to be considered a good fit (Marsh & Hau, 2004). The RMSEA is considered an important fit index as it takes into account the number of parameters in the hypo- thesised models and selects the most parsimonious model to analyse (Hooper et al., 2008). A value of less than or equal to 0.08 on this index indicates a good fit (Benet-Martínez & Karakitapoglu-Aygun, 2003; Hu & Bentler, 1999).

The proposed factor structure from Phase 1 was labelled model 1 and involves the 22 items in three factors namely Pride in Ethnic Background and Language (Pride/Membership), Degree of Liking for Traditional and Social Activities of my Ethnic Group (Trad/Social), and Sense of Belonging to This Country (Belong/Australia). The major problem with this model was that the correlation between Pride/Membership and Trad/ Soc is 0.78—indicating that they are probably best combined. Thus, a two-factor model was tested, with one factor relating to Ethnic Identity and the second factor to Belonging to Australia. While one of the fit indices suggested a good fit

(SRMR = 0.078 < 0.1), the other three suggested a less than optimal fit (CFI = 0.886 < 0.9, NNFI = 0.874 < 0.9, RMSEA = 0.112 > 0.08). The modification indices indicated that the fit could be substantial improved by adding an error covariance between items p20 and p5, items p22 and p2; items t16 and t14; items t14 and t7, and items b21 and b8. Three of the four fit indices suggested a good fit (CFI = 0.926 > 0.9, NNFI = 0.915 > 0.9, SRMR = 0.068 < 0.1); however, the RMSEA fell slightly short of the recommended < 0.08 cut-off (RMSEA = 0.088). Given that the majority of the indices of fit were within recommended limits, the two-factor solution was concluded to be the best fit to the data. It is important that the decision to free error covariances is not simply a data-driven process, but can be justified on substantive grounds. It is clear from Table 5 that items within the specified pairs are asking virtually the same thing (e.g., item p20 “I feel great pride in

Table 5 Questions According to Variable Code in Figure 2

Factor 1: Ethnic identity

p2 It annoys me when members of my ethnic group hide their

ethnic background.

b3 I am not sure whether I belong to this country or not.

t4 I would not like to marry someone from a different ethnic

background to my own.

p5 I take pride in telling people that I am a member of my ethnic

group.

p6 If I had children I would like them to be taught at school by

teachers from my ethnic group.

t7 I prefer to do business with members of my ethnic group.

p10 I like members of my ethnic group to use the ethnic language as

much as possible.

p11 I would like members of my ethnic group to learn some of the

language of my ethnic group.

t12 If I married someone from outside my ethnic group I would

insist that my children be raised according to the traditions of

my ethnic group.

p13 It makes me angry when people from my ethnic group don’t use

their ethnic names.

t14 I would like to socialise mainly with members of my ethnic

group.

p15 I wish that there were more books and newspapers available in

my ethnic language.

t16 I like to stick together with those of my ethnic group.

t17 I like to carry on the traditional ways of my ethnic group.

t18 I like the way people of my ethnic group raise their children.

t19 I prefer food to be prepared and served according to the

traditional ways of my ethnic group.

p20 I feel great pride in being a member of my ethnic group.

p22 It makes me angry when kids from their ethnic group are

ashamed of their ethnic background.

Factor 2: Belong Australian

b21 I like to make it clear to people that I belong to Australia.

b8 I like to make it clear to members of my ethnic group that I

belong in Australia.

b9 I feel a strong sense of belonging to Australia.

b1 I am very proud of the history and traditions of Australia.

Universal Ethnic Identity Scale (UEIS) R Rooney et al.

244

being a member of my ethnic group.” and item p5 “I take pride in telling people that I am a member of my ethnic group.”) and therefore correlated errors are to be expected. The modified model is presented in Figure 2, and the items representing the variable codes for the two factors are shown in Table 5.

Using this two-factor structure, total scores for each subscale for each ethnic group can be calculated. The reader is reminded that lower scores reflect greater ethnic identity and belonging to Australia. The potential range of scores for Ethnic Identity is 18 to 90, while that of Belonging is 4 to 20. The means and standard deviations for each ethnic group are displayed in Table 6.

Figure 2 Restricted Factor Model Specifying Two Factors: Ethnic Identity and Belong/Australian.

Table 6 Means and standard deviations of ethnic identity and belonging to Australia subscales

Ethnic group n Ethnic identity Belonging to Australia

Mean SD Mean SD

Italian 276 50.89 10.40 9.54 3.18

Vietnamese 446 40.14 9.32 9.89 3.03

Farsi 224 47.24 10.94 9.91 3.25

Croatian 61 48.08 14.08 11.92 4.00

Total 1,007 45.27 11.26 9.92 3.23

R Rooney et al. Universal Ethnic Identity Scale (UEIS)

245

To examine the convergent validity of the UEIS subscales, Pearson’s correlations were computed with the subscales of the MEIM. In Table 7 the results are shown for each ethnic group for which we had data from both scales.

It can be seen that the correlations between the two ethnic identity subscales range from 0.48 to 0.82. The correlations between Other Group Orientation and Belonging to Australia are considerably smaller, ranging from -0.11 to 0.29, which is unsurprising given that they are measuring conceptually dis- tinct constructs. Of interest is the fact that the relationship between the UEIS Ethnic Identity and Belonging to Australia scales ranges from -0.36 to 0.42; for the Croatian and Italian groups, there is a negative correlation where the more they identify with their own group, the less they feel a sense of belonging to Australia. For the Vietnamese and Farsi groups, the direction is reversed where a stronger identification with their own group is associated with a stronger sense of belonging to Australia.

Discussion

The most parsimonious factor structure that emerged in Phase 2 was of two factors, rather than the three factors found in Phase 1. The two-factor structure was made up of “Sense of Belonging to this Country” and the second factor was a combination of items from both “Pride in Ethnic Background and Language” and “Liking for Traditional and Social Activities of my Ethnic Group.” The items from “Pride in Ethnic Background and Lan- guage” and “Liking for Traditional and Social Activities of my Ethnic Group” were combined to form a single ethnic identity factor. Four out of five items from “Sense of Belonging to this Country” were the same for the sample in Phase 2 as they were for Phase 1. The item “I am not sure whether I belong to this country or not” no longer loaded on “Sense of Belonging to this Country” and now loaded on the single “Ethnic Identity” factor. Convergent validity between the ethnic identity factors of the UEIS and the MEIM has been shown across a range of groups with correlations of between 0.48 and 0.82. The correlations between “Other Group Orientation” and “Belonging to Austra- lia” were smaller, which may be because “Belonging to Austra- lia” was a more specific measure of other group orientation with identification with the host country, in this case, Australia being nominated in the items.

The most likely reason for the difference in factor structure between Phase 1 and Phase 2 is that the samples differed in

terms of diversity of ethnic groups. It is possible that for the more ethnically diverse participants in Phase 1 there was more likely to be a wider diversity of factors, thus the three rather than the two-factor solution. The same explanation, that is the differences in cultural diversity between the phases, can be given for why the National Identity item found in Phase 1 did not load on the same factor on Phase 2. Further investigation of the factor structure in different ethnic groups will help to clarify this issue.

The two dimensions of the new scale have a different number of items, both of which are responded to on a scale ranging from 1 (strongly agree) to 5 (strongly disagree). Thus, for “Ethnic Iden- tity,” which has 18 items, the minimum score on the dimension is 18, and the maximum score is 90, indicating a high level of identification in ethnic background. Finally, there are four items in the dimension, “Sense of Belonging to This Country,” so the minimum score is 4 and the maximum score is 20, indicating a high degree of sense of belonging to “This Country,” in this case, Australia.

When comparing the two factors found in Phase 2 with pre- vious studies of different ethnic groups, there is a similarity in the depiction of Phinney’s (1992) two-factor structure. Specifi- cally, “Pride in Ethnic Background and Language” and “Liking for Traditional Activities of my Ethnic Group” have formed a combined factor, which is similar to a general social identity or affirmation factor. It appears that the combined factor contains both internal psychological and affective processes including pride in membership as well as external social behavioural items. “Sense of Belonging to This Country” is a relatively new dimension and its appearance can be attributed to the intro- duction of items in the brainstorming stage. “Other Group Orientation” in Phinney’s MEIM scale is similar in that it assesses identification with other groups but this factor is less specific than “Sense of Belonging to This Country,” which directly identifies the new country of location for migrants. The result is that ethnic identity for migrants can be said to involve not only the extent of identification with one’s own ethnic group but also the extent of identification with the new country in which an individual has settled and so provides an advantage over the MEIM. In addition, the UEIS has been developed on adults and includes participants from South-East Asian cultures in an Australian context, which was lacking in the development of the MEIM which did not have South-East Asian participants and was developed with adolescents as participants. A previous exploratory study has identified four components of ethnic

Table 7 Pearson’s Correlations Between Subscales of the UEIS and the MEIM

Ethnic group Subscale n MEIM EI MEIM OGO UEIS EI UEIS belong

Italian UEIS EI 276 0.69 0.12 1.00 -0.20 UEIS belong 276 -0.16 0.17 -0.20 1.0

Vietnamese UEIS EI 167 0.48 -0.10 1.00 0.42 UEIS belong 167 0.23 -0.02 0.42 1.00

Farsi UEIS EI 224 0.63 -0.38 1.00 0.22 UEIS belong 224 0.16 -0.11 0.22 1.00

Croatian UEIS EI 61 0.82 -0.34 1.00 -0.36 UEIS belong 61 -0.28 0.29 -0.36 1.00

Note. MEIM = Multigroup Ethnic Identity Measure; OGO = other group orientation; UEIM = Universal Ethnic Identity Scale.

Universal Ethnic Identity Scale (UEIS) R Rooney et al.

246

identity (Luhtanen & Cocker, 1992) but this was not of first- generation migrants and was less specific about the social groups that were being referred to. Occasionally, measures of the ethnic identity of a range of groups have found multiple factors (e.g., Driedger, 1976). However, it appears that for scales across a range of cultural groups, which addresses individual’s ethnic identification with their own and other groups, a two-factor model has been supported most often in studies. Given that the range of groups in the second phase was less diverse than in the first phase, further investigation of the factor structure is required across further samples of culturally diverse popula- tions. In addition, there is evidence that the way in which the two factors of the UEIS are related may vary, for example a strong ethnic identity for Croatians was associated with a weaker sense of belonging to Australia where the reverse was the case for Italian participants in Phase 2. Consequently, further investigation of the factor structure and relationship between the variables in different cultural groups is warranted.

In summary, the final stage of the four-stage factor analytic process suggested by Hattie (1981) was carried out, and involved cross validating the proposed ethnic identity question- naire on a different sample of participants from culturally diverse backgrounds. The cross validation supported a two, rather than three-factor structure found in Phase 1 and further investigation of the factor structure in a diverse range of ethnic groups is needed.

Conclusions

The dimensional structure of ethnic identity for participants from diverse culturally and linguistically diverse (CALD) back- grounds in Phase 2 was different to that obtained in Phase 1. The final factor structure contained two factors, a general Ethnic Identity factor and a factor assessing identification with the more recent country of location, i.e., “Sense of Belonging to This (host) country.”

Overall, the two factors “Ethnic Identity” and “Sense of Belonging to This Country,” appeared to be the most parsimo- nious solution. It now needs to be further investigated as a potentially general measure of ethnic identity across a range of groups, certainly including European and Asian adults. When comparing the factor structures, the overlap found between “Pride in Ethnic Background and Language” and “Liking for Traditional Activities of my Ethnic Group” in the cross- validation stage suggests that the structure of these two factors was so similar in meaning that the most parsimonious solution was to combine the two factors into one general Ethnic Identity factor. “Sense of Belonging to This Country” is a relatively new dimension, and consistent with Berry’s (Berry, 1990, 1997) fourfold classification of acculturation strategies, suggests that individuals may choose to identify with their own ethnic group but also with the new country in which the individual has settled. The UEIS allows this identification with the host country to be assessed in terms of degree of identification.

Consistent with Tajfel and Turner (1986), items pertaining to cognitive, behavioural, and affective components were found to be contained within the two dimensions that appeared to measure ethnic identity. However, rather than forming separate components, they were mixed throughout the dimensions.

“Ethnic Identity” included items pertaining to cognition through items related to being a “member of the ethnic group” For example, “pride in being a member” and “pride in telling people about being a member” both involve cognitive identifi- cation with being a member of the ethnic group, in addition to affect which is present in the feeling of pride. Cognition was also involved in “Sense of Belonging to This Country” where items involved identifying the degree of “sense of belonging to this country,” i.e., Australia. Affect was present throughout the two dimensions with feelings such as “pride,” “anger,” and “shame” being part of many items; for example, “I feel angry when members of my ethnic group are ashamed of their ethnic back- ground.” Finally, behaviour was mainly involved in “Ethnic Identity” with items including socialising, eating traditional food, and child rearing.

In summary, Phase 1 provided evidence of a new scale to assess ethnic identity in adults from a range of ethnic groups, including members from South-East Asian groups such as Viet- namese. The four-stage analysis (Hattie, 1981) which was fol- lowed ensured that choice of items was guided by a relevant theoretical platform, in this case, the SIT (Tajfel & Turner, 1986), as well as the extant literature and pilot testing. Exploratory analysis aided in the assessment of the number of factors; factor names were based on theoretical arguments and assignment of items to factors agreed among researchers. Goodness of fit was assessed using confirmatory factor analysis, and finally, the instrument was cross validated in Phase 2 on a different multi- cultural sample. The resultant questionnaire needs to be further investigated in ethnically diverse ranges of participants before it can be used as an assessment of a migrant’s identification with the host culture as well as an investigation of identification with their home country.

References

Ashmore, R., Deaux, K., & McLaughlin-Volpe, T. (2004). An organising framework for collective identity: Articulation and significance of multidimensionality. Psychological Bulletin, 130, 80–114.

Benet-Martínez, V., & Karakitapoglu-Aygun, Z. (2003). The interplay of cultural values and personality in predicting life-satisfaction: Comparing Asian- and European-Americans. Journal of Cross-Cultural Psychology, 34, 38–61.

Berry, J. W. (1990). Psychology of acculturation. In J. Berman (Ed.), Cross-cultural perspectives: Nebraska Symposium on Motivation (pp. 201–234). Lincoln: University of Nebraska Press.

Berry, J. W. (1997). Immigration, acculturation and adaptation. Applied Psychology: An International Review, 46, 5–34.

Berry, J. W., Phinney, J. S., Sam, D. L., & Vedder, P. (2006). Immigrant youth: Acculturation, identity and adaptation. Applied Psychology: An International Review, 55, 303–332.

Berry, J. W., Poortinga, Y. H., Segall, M. S., & Dasen, P. R. (1992). Cross-cultural psychology. New York: Cambridge University Press.

Cuellar, I., Nyberg, B., Maldonado, R., & Roberts, R. (1997). Ethnic identity and acculturation in a young adult Mexican-origin population. Journal of Community Psychology, 25, 535–549.

Driedger, L. (1976). Ethnic self-identity: A comparison of in-group relations. Sociometry, 39, 131–141.

Hattie, J. (1981). A four-stage factor analytic approach to studying behavioural domains. Applied Psychological Measurement, 5, 77–88.

R Rooney et al. Universal Ethnic Identity Scale (UEIS)

247

Hogg, M. A., Abrams, D., & Patel, Y. (1987). Ethnic identity, self-esteem, and occupational aspirations of Indian and Anglo-Saxon British adolescents. Genetic, Social, and General Psychology Monographs, 113, 487–508.

Hogg, M. A., Turner, J. C., Nascimento-Schulze, C., & Spriggs, D. (1986). Social categorisation, intergroup behaviour and self-esteem: Two experiments. Revista du Psicologica Social, 1, 23–37.

Hooper, D., Coughlan, J., & Mullen, M. (2008). Structural equation modelling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–60.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indices in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.

Jöreskog, K. G., & Sörbom, D. (2004). LISREL 8.7 for Windows [computer software]. Lincoln, IL: Scientific Software International Inc.

Kwan, K.-L. K. (2000). The internal-external ethnic identity measure: Factor analytic structures based on a sample of Chinese Americans. Educational and Psychological Measurement, 60, 142–152.

Leclezio, M. K., Louw-Potgieter, J., & Souchon, M. B. S. (1986). The social identity of Mauritian immigrants in South Africa. Journal of Social Psychology, 126, 61–69.

Lemyre, L., & Smith, P. M. (1985). Intergroup discrimination and self-esteem in the minimal group paradigm. Journal of Personality and Social Psychology, 49, 660–670.

Lewin, K. (1948). Resolving social conflicts. New York: Harper. Liebkind, K. (1996). Acculturation, identity and stress. Journal of

Cross-Cultural Psychology, 27, 161–180. Liebkind, K., Jasinskaja-Lahti, I., & Solheim, E. (2004). Cultural identity,

perceived discrimination, and parental support as determinants of immigrant’s school adjustments: Vietnamese youth in Finland. Journal of Adolescent Research, 19, 635–656.

Lorenzo-Hernandez, J., & Oullette, S. C. (1998). Ethnic identity, self-esteem, and values in Dominicans, Puerto Ricans, and African Americans. Journal of Applied Social Psychology, 28, 2007–2024.

Luhtanen, R., & Cocker, J. (1992). A collective self-esteem scale: Self-evaluation of one’s social identity. Personality and Social Psychology Bulletin, 18, 302–318.

Marsh, H. W., & Hau, K.-T. (2004). In search of golden rules: Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing H (1999) findings. Structural Equation Modeling: A Multidisciplinary Journal, 11, 320–341.

Nesdale, D., Rooney, R., & Smith, L. (1997). Migrant ethnic identity and psychological distress. Journal of Cross-Cultural Psychology, 28, 569–588.

van Oudenhoven, J. P., & Eisses, A. M. (1995). Integration and assimilation of Moroccan immigrants in Israel and The Netherlands. Paper presented at IV European Congress of Psychology, Athens, July, 1995.

Phinney, J. S. (1990). Ethnic identity in adolescents and adults: A review of research. Psychological Bulletin, 108, 499–514.

Phinney, J. S. (1992). The Multigroup Ethnic Identity Measure: A new scale for use with diverse groups. Journal of Adolescent Research, 7, 156–176.

Phinney, J., & Chavira, V. (1995). Parental ethnic socialisation and adolescents coping with problems related to ethnicity. Journal of Research on Adolescents, 5, 31–53.

Phinney, J. S., Chavira, V., & Williamson, L. (1992). Acculturation attitudes and self-esteem among high school and college students. Youth and Society, 23, 299–312.

Phinney, J. S., Horenczyk, G., Liebkind, K., & Vedder, P. (2001). Ethnic identity, immigration, and well-being: An interactional perspective. Journal of Social Issues, 57, 493–510.

Phinney, J. S., & Ong, A. D. (2007). Conceptualisation and measurement of ethnic identity: Current status and future directions. The Journal of Counseling Psychology, 54, 271–281.

Ponterotto, J. G., Gretchen, D., Utsey, S. O., Stracuzzi, T., & Saya, R. (2003). The Multigroup Ethnic Identity Measure (MEIM): Psychometric review and further validity testing. Educational and Psychological measurement, 63, 502–515.

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.

Sam, D. L., & Berry, J. W. (1993). Acculturative stress among young immigrants in Norway. Scandinavian Journal of Psychology, 34, 276–288.

Sands, E. A., & Berry, J. W. (1993). Acculturation and mental health in Greek-Canadians in Toronto. Canadian Journal of Community Mental Health, 12, 117–124.

Schmitz, P. (1995). National and cultural identity problems of immigrants. Paper presented at IV European Congress of Psychology, Athens, Greece.

Spencer, M., Icard, L. D., Harachi, T. W., Catalano, R. F., & Oxford, M. (2000). Ethnic identity among monoracial and multiracial early adolescents. Journal of Early Adolescence, 20, 365–387.

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics. Boston, MA: Allyn & Bacon.

Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behaviour. In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7–24). Chicago, IL: Nelson-Hall.

Ullah, P. (1985). Second generation Irish youth: Identity and ethnicity. New Community—Journal of the Commission for Racial Equity, XII, 310–320.

Ullah, P. (1987). Self-definition and psychological group formation in an ethnic minority. British Journal of Social Psychology, 26, 17–23.

Whittler, T. E., Calantone, R. J., & Young, M. R. (1991). Strength of ethnic affiliation: Examining black identification with black culture. The Journal of Social Psychology, 131, 461–467.

Universal Ethnic Identity Scale (UEIS) R Rooney et al.

248

emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.

However, users may print, download, or email articles for individual use.