Journal of Applied Developmental Psychology 32 (2011) 109–117

Contents lists available at ScienceDirect

Journal of Applied Developmental Psychology

American elementary school children's attitudes about immigrants, immigration, andbeing an American

Christia Spears Brown ⁎University of Kentucky, 125 Kastle Hall, Lexington, KY 40506, USA

⁎ Tel.: +1 859 257 6827.E-mail address: christia.brown@uky.edu.

0193-3973/$ – see front matter © 2011 Elsevier Inc. Aldoi:10.1016/j.appdev.2011.01.001

a b s t r a c t

a r t i c l e i n f o

Article history:Received 2 November 2009Received in revised form 18 January 2011Accepted 18 January 2011Available online 25 February 2011

Keywords:StereotypesNational identityAttitudes toward immigrantsElementary school children

The current study examined 5 to 11-year-old European American children's (N = 90) attitudes regardingimmigrants, immigration policy, and what it means to be an American. The majority of children in the sample(from a predominantly European American community) held strong American identities and had distinctideas about what it means to be an American (namely, one must love America, live by its rules, and beWhite).Children were in favor of legal immigration as a policy, and although they believed in allowing illegalimmigrants to stay if employed, many younger children believed they should go to jail. Many children in thesample were aware of Americans' anti-immigration sentiments, largely attributing it to ethnic/culturaldiscrimination. Finally, children held negative attitudes about immigrants, particularly Mexican immigrants.These negative attitudes weremost evident among children who held a strong, prototypical national in-groupidentity. In contrast, children did not hold differential attitudes about White and Black Americans.

l rights reserved.

© 2011 Elsevier Inc. All rights reserved.

The U.S. has recently seen rapidly changing demographics, largelyfueled by immigration. In 2008, a total of 1,107,126 immigrantsbecame legal permanent residents of the United States, mostcommonly immigrating from Mexico and People's Republic of China(Monger & Rytina, 2009). While the immigrant population isincreasing, anti-immigration sentiments are ubiquitous, whether itbe media coverage of the “Minutemen” who patrol the U.S.–Mexicanborder to “enforce immigration laws” (http://www.minutemanpro-ject.com); repeated Congressional “English as the National Language”proposals; or controversies surrounding nationwide pro-immigrationmarches (CNN, 2007). These controversies are often framed aroundconcerns about patriotism and a weakening of an American identity(Deaux 2008). American children are explicitly taught about theideals of a free country that welcome immigrants, yet are consistentlyexposed to these anti-immigration sentiments. As research hasshown, children's attitudes and reasoning are influenced by socialcontext and exposure (Brown, Mistry, & Bigler 2007; Killen,McGlothlin, & Henning 2008). Despite the presence of this ubiquitouscultural debate on immigration, and children's frequent interactionwith children of immigrants at school (Hernandez, Denton, &Macartney 2008), little to no research has examined children'sperceptions of and attitudes toward immigrants and immigration.

Because an immigrant represents someone from another nation(i.e., a national out-group) trying to become a member of one'snational in-group, individuals' attitudes towards immigration are

informed by their sense of their national identity (Deaux 2008).Research has previously examined the development of nationalidentity among children. Studies (typically with British children)have shown that children have developed a national in-group identityby age 5 (Barrett & Short 1992; Barrett, Wilson, & Lyons 2003;Bennett, Lyons, Sani, & Barrett 1998; Bennett et al. 2004). Children atthis age are able to label their nationality and categorize themselvesaccordingly. Children by age 5 to 6 also prefer their national in-groupto other nationalities (Bennett et al. 2004, 1998). There is a smallerbody of research specifically on attitudes toward immigrants. Namely,research has shown that 10- to 12-year-old Dutch early adolescentsare more negative toward immigrants (specifically, asylum seekers)than established minority groups (Verkuyten & Steenhuis 2005). Inaddition, in a large-scale international sample, female adolescentswere shown to have more positive attitudes toward immigrants'rights than males (Torney-Purta, Wilkenfeld, & Barber 2008).

The current study addresses four specific research questionsrelated both theoretically and practically to the immigration debate.First, the study assesses White European American children'sidentification and conceptualization of their national in-group (i.e.,Americans). For example, we examine how much children identifywith being an American, how they define being an American, andwhether they believe certain ethnic groups are more American thanother groups. Second, the study assesses White European Americanchildren's knowledge of and attitudes about immigration, both legaland illegal. Specifically, it examines children's endorsement ofattitudes that are most commonly expressed by adults (e.g., whetherimmigrants take jobs from other Americans). Third, the study assessesWhite European American children's stereotypes and attitudes about

110 C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

immigrants, and whether these stereotypes and attitudes differ basedon the country of origin of the immigrant. Finally, the study examineswhether the strength of children's national identity is related to theirattitudes and stereotypes about immigrants. Importantly, this studywas conducted in a city reflective of much of “middle America”: apredominantly (76%) White European American city, with 13%African Americans, and an increasing Mexican immigrant population.

The first research question addresses children's national identity,specifically what it means to be an American. One theoretical linkbetween national identity and attitudes toward immigrants is basedon the Common Ingroup Identity Model (Gaertner & Dovidio 2000,2009), which argues that the inclusion of out-group members into asuperordinate common group leads to more positive attitudes towardthat group. In other words, if individuals view immigrants as part ofthe national in-group of Americans, we would expect them to holdmore positive attitudes toward immigrants (Esses, Dovidio, Jackson,& Armstrong 2001). In contrast, if individuals have impermeable in-group boundaries, they may maintain their negative attitudes andstereotypes. Thus, it is important to examine the extent to whichindividuals hold tight restrictions about who can and cannot be anin-group member.

One way to assess the permeability of the national in-groupboundary is to ask participants to definewhatmakes a “true”American(Devos & Banaji 2005). When adults were asked this question, theyranked being born in America as a rather unimportant criterion.Instead, the most highly ranked criteria for being a true American wasthey must respect America and its laws, followed by they must speakEnglish, followed by they must live in America a long time (Devos &Banaji 2005). This suggests that adults support immigrants becomingAmericans in general, and they allow certain immigrants to enter theirnational in-group. However, despite seemingly open boundaries,adults have rather narrow definitions of prototypical Americans,often based on the prototype of aWhite American (Devos &Ma 2008).Specifically, although they were described as being born in America,Asian Americans and African Americans were rated by adults as “lessAmerican” than White Americans (Devos & Banaji 2005). Based onthese findings, we hypothesized that White children would also havepermeable in-group boundaries but would rate White Americans asmore American than other American-born groups (including AfricanAmericans). We predicted that children with more prototypicalrestrictions about their in-group would hold more negative attitudesabout immigrants.

The second research question relates to children's knowledge ofspecific issues often raised by adults surrounding immigration. Adults(often college students) oftenholdnegative attitudes toward immigrantsbecause they represent a symbolic threat, which is a threat to in-groupmorals and standards, and a realistic threat, which is a threat to thematerial well-being of the in-group (Stephan, Ybarra, & Bachman 1999).These perceived threats are often framed in the language of patriotism,but the patriotism is focused on protection (e.g., “Wehave to protect ourcountry”) rather than simply positive attitudes toward the country (e.g.,“I love my country”). Many of these perceived threats are expressed inthemedia, and thus can influence children's attitudes about immigration.For example, a television commercial in 2010 by the Coalition for theFuture American Worker argued that immigrants are taking jobs awayfrom American workers. Adults also make a strong distinction betweenlegal versus illegal immigrants (Deaux 2008), holding considerablymoreunfavorable attitudes toward illegal or undocumented immigrants thanlegal immigrants (Lee & Fiske 2006; Short 2004). No research hasexaminedwhether children perceive immigrants to pose the same typesof symbolic or realistic threats as adults, and whether they distinguishbetween legal versus illegal immigrants. Further, no research hasexamined children's general understanding of why immigration occursand their rationale for why it should be legal versus illegal.

Children's attitudes are likely to be different than adults' attitudesabout immigration and immigration policy because of their develop-

ing cognitive abilities. For example, children in early rather than lateelementary school are particularly likely to focus on concrete ratherthan abstract attributes (Keil 2006). Research has shown that, prior toage seven, children regularly emphasize concrete external qualitieswhen describing people (Aboud 1988; Aboud & Levy 2000; Martin1989). This is consistent with Piagetian theory, which notes thatchildren in this age group tend to overemphasize observable qualities(see Inhelder & Piaget 1964). Thus, we hypothesized that earlyelementary school children, when asked to explain anti-immigrantattitudes, would attribute attitudes less to symbolic or realistic threatsand more to concrete reasons (such as immigrants looking different).In addition, we hypothesized that children would be more negativetoward illegal than legal immigrants. We also predicted that earlyrather than late elementary children would endorse legal ramifica-tions for illegal immigration (e.g., they broke the law, thus they shouldgo to jail) because of younger children's legalistic, rule-based moralreasoning in which it is rarely appropriate to break a law (Helwig &Jasiobedzka 2001; Tapp & Kohlberg 1971).

The third research question assessed whether children holdnegative attitudes toward or stereotypes about immigrants, andwhether those attitudes and stereotypes differ based on the country oforigin of the immigrant. Researchwith adults indicates that Americansholdmore negative stereotypes toward Hispanic/Mexican immigrantsthan Asian immigrants (Lee & Fiske 2006) or English immigrants(Short 2004). In general, Americans are more favorable to Whiteimmigrant groups than groups of color (Deaux 2006). Evidencesuggests childrenwill hold similarly differing stereotypes based on theimmigrant's country of origin (Enesco, Navarro, Paradela, & Guerrero2005). For example, by middle childhood, children are aware ofdifferent nationalities' stereotypes (Barrett & Short 1992). Though notregarding immigrants per se, Barrett and colleagues found Britishchildren, in line with cultural stereotypes, held more favorableattitudes toward Americans than Germans (Barrett et al. 2003).Children by age 5 also show preferences for children who speak theirsame language (Powlishta, Serbin, Doyle, &White 1994). Thismay be aresult of same-language speakers being more easily included into acommon in-group. Thus, in the current study, we hypothesized thatchildren would hold negative attitudes and stereotypes towardimmigrants, but these stereotypes would be more negative toward aMexican immigrant than a British immigrant (who is both White andspeaks English, and thus can more easily be included in a common in-group). As a point of comparison, children's stereotypes about BlackAmericans (who are included in White American children's nationalin-group, but not racial in-group) were also assessed. This comparisonallows for a test ofwhich bias ismore prominent: racial bias or nationalbias. We predicted that, because of the current climate of negativeimmigrant sentiments, children would be more negative towardsimmigrants than other Americans, regardless of race.

The degree of negative stereotypes children hold towardimmigrants may depend, however, on their own national identity.Based on social identity theory (Tajfel & Turner 2004), individualswho hold an important group identity strive for positive distinc-tiveness and may in turn hold more negative attitudes about theirout-group. In support of this theoretical rationale, research withadults has shown that a strong national identity is related tostronger biases against other nationalities (Reizábal, Valencia, &Barrett 2004) and more negative attitudes toward immigrant groups(Billiet, Maddens, & Beerten 2003). Verkuyten (2009) found similarresults with Dutch adolescents, in which a stronger national identitywas related to perceiving newcomer groups as more threatening,which in turn was related to less support for multiculturalism andimmigrant rights. In the current study, in line with social identitytheory, we hypothesized that White European American childrenwith strong national identities would hold more negative attitudesand stereotypes about immigrants than children with weakernational identities.

111C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

Methods

Participants

Participants consisted of 90 White European American children(50 girls, 40 boys), ranging in age from 5 to 11 years. The mean agewas 7 years, 10 months (SD = 1 year, 6 months). To examine agedifferences in attitudes about immigration, children were categorizedinto two age groups, based on a median split: younger children,ages 5 to 7 (n= 39), and older children, ages 8 to 11 (n= 51). Theseage groups also correspond with a qualitative shift in children'scognitive development that is related to their endorsement of racial/ethnic stereotyping and prejudice (Aboud 1988).

All children classified themselves as “White” and no children werethemselves first- or second-generation immigrants. Participants wererecruited from the after-school programs associated with threeelementary schools in a medium-sized city in the Upper South.Specifically, the city is 76% White European American, 13% AfricanAmerican, 5% Hispanic (specifically, 4.5% are from Mexico), and 3%Asian (specifically, 1% are Indian, .6% are Chinese, and .6% areJapanese). Of the total population, 8% are foreign born. The schooldistrict serves 35,429 students, of which 61% are European American,24% are African American, 8% are Hispanic, and 4% are Asian, with2137 students enrolled in English as a Second Language. The vastmajority of Hispanic students are first- and second-generationimmigrants from Mexico (whose parents are typically farm workers),and the majority of Asian students are first- and second-generationimmigrants from India, China, and Japan (whose parents are typicallymedical professionals or academics). The specific three schools thechildren attended have approximately 83–84% White EuropeanAmerican students, 5% African American students, 1–3% Hispanicstudents, and 3–11% Asian students. Thus, the children in the currentsample have little contact with immigrants at school (although theirdegree of contact was not directly measured). Fewer than 20% ofstudents qualified for free or reduced lunch. Analyses indicated nodifferences across school sites.

Only those children with signed parental consent, and whothemselves gave assent, participated in the study. Approximately75% of children who received consent forms (entitled “Living andWorking in America”) returned them signed.

Procedure and measures

OverviewAll measures were administered to the children individually in

their after-school program by a female experimenter. To control forreading ability, all questions were read aloud to each child, and theexperimenter wrote down the child's responses. The measures aredescribed in the order of the research questions, but in practice, theexplicit measures about immigration and anti-immigrant sentimentswere given last (the rest of the measures were counterbalanced).Upon completion of the measures, all children were fully debriefedand given a small toy from a “treasure box.”

Conceptualization of national in-groupTo assess the strength of children's national identity, they were

asked, “How American do you feel?” Children responded on a 4-pointscale, ranging from not at all American (1) to very American (4). Asimilar one-item assessment of national identity has been usedpreviously (Pfeifer et al. 2007).

Children's definition of Americanswas assessed. Childrenwere askedabout their criteria for being a true American based on items adaptedfrom Devos and Banaji (2005). Of particular interest was whetherchildren held impermeable in-group boundaries (e.g., to be a trueAmerican onemust beborn inAmerica). Specifically, childrenwere told,“Think about what it means to be a true American. Some people say

there are some things thatmake a person a true American, while otherssay that there is nothing that makes one person more American thananother. What do you think it means to be an American?” The fourcriteria were: (a) “Be born in America,” (b) “Speak English,” (c) “Live inAmericamost of their lives,” and (d) “Love America and try to live by itsrules.” Children rated each criterion on a 4-point scale, ranging from notat all true (1) to very true (4).

Children's attitudes about the prototypicality of Americans werealso assessed. To measure the degree to which children differentiatedAmericans based on ethnicity, children were asked to think aboutpeople born in American (Devos & Banaji 2005). Specifically, theywere told, “Think about people who were born in America. In youropinion, how ‘American’ are people in these groups? Remember, youare thinking about people born in America” (original emphasis).Children rated how American the following groups were: AfricanAmericans, White Americans, Latino/Hispanic Americans, and AsianAmericans. Children used a 4-point scale ranging from not at allAmerican (1) to very American (4).

Conceptualization of immigrationTo examine children's knowledge of immigration and the immigra-

tion debate, childrenwere asked, “Some people are upset when peoplemove to the US from other countries. Have you ever heard aboutthis?” Children responded “yes” or “no”. If they answered “yes,”children were asked, “Where did you hear about this?” Theyresponded to 4 possible sources: TV, Newspapers, Parents, and School.To assess their knowledge of the reasons for immigration, childrenwere then asked, “Why do you think people move here from othercountries?” Their open-ended responses were recorded. There wereno a priori codes created; categories were derived from emergingthemes from the responses. Three primary themes emerged: to seekfreedom, to seek greater material/financial opportunities, and toescape hostile conditions (described in detail in results). Oncecategories were created, the author and a research assistantindependently coded every response. The interrater reliability washigh (Cohen's kappa was .97) and all discrepancies were discusseduntil agreement was reached on 100% of the responses.

To assess children's attitudes about legal and illegal immigration,they were asked, “Do you think people should be allowed to movehere from other countries? Why or why not?” Their open-endedresponses were recorded. As before, there were no a priori codescreated; categories were derived from emerging themes from theresponses. Three primary themes emerged: because of freedom,because America is desirable, and because diversity is good (describedin detail in results). Once categories were created, the author and aresearch assistant independently coded every response. The interraterreliability was high (Cohen's kappa was .86) and all discrepancieswere discussed until agreement was reached on 100% of theresponses.

To assess children's attitudes about the consequences of illegal (orundocumented) immigration, childrenwere asked, “Think about peoplewhomovehere fromother countries even though itwas against the law.What do you think should happen to them?” Children responded “yes,”“maybe,” or “no” to each counterbalanced option: (a) “They should besent back to the country they came from,” (b) “They should be allowedto stay in theU.S. if theyhave jobs,” and (c) “They should beput in jail forbreaking the law.” These items were not mutually exclusive (i.e.,although illogical, children could respond yes to all three) andwere notintended to represent all possible options (e.g., allowing immigrants tostay in the U.S. evenwithout jobs). Instead, because the study examinedattitudes that are likely influenced by social context and exposure, thequestions were designed to tap the options most frequently mentionedby adults in the ongoing cultural debates about immigration. Forexample, as of November 2010, Arizona has debated and passed a lawthat allows for the arrest of immigrantswithout legal documentation; inaddition, current debate has also focused on whether immigrants with

112 C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

jobs should be allowed to work toward legal status or whether theyshould be immediately deported (for a review of national pollingquestions asked of adults that taps each of these options, see http://www.pollingreport.com/immigration.htm).

To assess children's attributions for anti-immigration sentiments –

such as attributions to symbolic threat, realistic threat, or discrimi-nation – children were asked, “Why do you think American peoplemight not like it when other people move to America from othercountries?” There were six counterbalanced attributions that wereunique to the current study and designed to test both symbolic andrealistic threats frequently mentioned by adults and racial/ethnic/cultural discrimination: (a) “Because the people who move here arenot smart,” (b) “Because the people who move here take too muchmoney from the government,” (c) “Because the people who movehere are criminals (they break the law),” (d) “Because the peoplewho move here take too many good jobs from other Americans,”(e) “Because the people who move here like different things thanother Americans, like food and music,” and (f) “Because the peoplewho move here look very different from other Americans .” Childrenrated each criterion on a 4-point scale, ranging from not at all true(1) to very true (4).

Stereotypes of and attitudes toward immigrantsChildren were read 5 counterbalanced vignettes about different

people (all men) who moved to their city from different countries orregions of the U.S.: from (a) Mexico, (b) China, (c) England,(d) California, and (e) a small town in their state. Each vignette includeddetails about the person, specificallywhere the personmoved from, thathe moved for work, that he sent part of his money back home to help afamilymember, that hehad awife and oneor two children, andwhat hishobby was (e.g., watching baseball). A picture of a person from theparticular country accompanied each vignette. Of the two Americans,one was African American and the other was European American. Eachpicture was taken from a public source photography website, and eachphotographwaspilot testedwith college students by asking themwherethey thought the person immigrated from. There was 100% agreementon the photographs of the people ostensibly from Mexico and China.After hearingeach vignette and lookingat thephotograph, childrenwereasked (a) howmuchmoney the personmade, (b) howmuch school theperson went to, (c) how hard the person worked, (d) how much theirtown/neighborhood wanted the person to move there, (e) how muchthey wanted to go to school with the person's child, and (f) how smartthe person's child was. Children responded to each question usinga 4-point scale, ranging from not very much (1) to very much (4).

Results

Preliminary analyses and overview

Preliminary analyses were conducted to test for differences basedon gender. There were no gender differences (contrary to previous,larger-scale research, Torney-Purta et al. 2008). For all analyses aboutimmigration or attitudes toward immigrants, follow-up analyseswereconducted to examine whether previous knowledge of the immigra-tion debate was related to children's responses. Previous knowledgeof the debate was not related to responses. Therefore, the analysescollapse across gender and knowledge level, and examine agedifferences (based on a median split) between younger children,ages 5 to 7 (n = 39), and older children, ages 8 to 11 (n = 51).

Conceptualization of national in-group

American identityOverall, children endorsed a strong American identity (M = 3.71,

SD = .57), with 75% of children feeling “very American.” A one-wayanalysis of variance compared the strength of younger and older

children's American identity, F(1, 87) = .11, p N .50, η2 = .00. Therewere no differences across age groups (M [SD]: younger = 3.68 [.66];older = 3.73 [.49]).

Definition of AmericansToassess children's criteria for being a trueAmerican, a 2 (age group:

younger, older) × 4 (criteria: born in America, speak English, live inAmerica most of their lives, love America) mixed ANOVA wasconducted, in which the last variable was treated as a within subjectsfactor. Analyses revealed a significantmain effect for criteria, F(3, 85)=32.55, p b .001, η2 = .54. Bonferonni analyses (used throughout)revealed that children rated the criterion that true Americans mustsimply love America and try to live by its rules as more accurate (M =3.45, SD = .93) than all other criteria. They rated the criterion thatpeople must speak English to be a true American as less accurate (M=2.00, SD=1.15) than all other criteria. The criteria that people must beborn inAmerica (M=2.31, SD=1.29) andmust live inAmericamost oftheir lives (M=2.37, SD=1.19) did not differ from one another. Therewere no differences across age groups.

Prototypicality of AmericansTo assess whether children think some ethnic groups are more

prototypically American than other ethnic groups, a 2 (age group:younger, older) × 4 (ethnic group: White American, Black/AfricanAmerican, Latino American, Asian American) mixed ANOVA wasconducted, in which the last variable was treated as a within subjectsfactor. Analyses revealed a significant main effect for ethnic group,F(3, 86) = 29.24, p b .001, η2 = .51. Bonferonni analyses revealedthat children rated White Americans (M = 3.71, SD = .55) as moreAmerican thanAfricanAmericans (M=3.25, SD=.78),whowere ratedas more American than Asian Americans (M = 3.03, SD = .77), whowere rated as more American than Latino Americans (M = 2.84, SD =.91). There were no differences across age groups.

Conceptualization of immigration

Knowledge of immigration and immigration debateTo examine how many children were aware of anti-immigration

sentiments, and whether this differed by age, a chi-square test ofassociation was conducted. Results indicated a significant associationbetween awareness of immigration debate and age group, χ2 (1) =3.89, p b .05. Specifically, 36% of younger children had heard that somepeople were against immigration, whereas 57% of older children had.When askedwhere they had heard about it, 10% of younger and 18% ofolder children heard about it on television (no significant agedifference); 8% of younger and 16% of older children read about it innewspapers (no significant age difference); 10% of younger and 24% ofolder children heard about it from parents (no significant agedifference); and 21% of younger and 41% of older children heardabout it from school (significant age difference: χ2 [1] = 4.32, p b .05).

When asked “why” people move here from other countries,children gave three primary responses. Overall, 34% of children (43%of older and 23% of younger) gave responses referring to immigrantsseeking freedoms available in the U.S. (e.g., “To have freedom and tonot be bossed around,” “They want freedom. They don't want to beruled by a king. They want to have a choice on going to church ornot.”); 21% of children (26% of older and 15% of younger) gaveresponses referring to greater material and financial opportunities inthe U.S. (e.g., “Because the houses are bigger,” “They can get moremoney,” “Because there are better jobs and they can support theirfamilies better.”); and 10% of children (6% of older and 15% ofyounger) gave responses referring to immigrants escaping their homecountries because of war, disease, famine, or political reasons (e.g.,“Because like in Iraq there are many wars and price of rice and meatare going up,” “Some people in their country might be having aconflict.”). However, 19% of children (10% of older and 31% of

113C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

younger) were unable to give a reason. These responses differed byage, χ2 (6) = 12.83, p b .05, such that a greater percentage of youngerchildren compared to older children were unable to give a reason forimmigration (goodness of fit: χ2 [1] = 5.28, p b .05).

Attitudes about legal and illegal immigrationWhen asked if people should be allowed to move here from other

countries, 97% of children said yes (95% of younger and 98% of olderchildren; no significant age difference). When asked “why” peopleshould be allowed to move here from other countries, children gavethree primary responses. Overall, 58% of children (67% of older and46% of younger) gave responses referring to America being a freecountry or made an implicit reference to the ideals of freedom (e.g.,“America is a free country and everyone has the right to live here,”“Because I think people should have the right to move to othercountries,” “Because if you want freedom, you should be able to haveit,” “Because it is a free country and we welcome people.”); 21% ofchildren (14% of older and 31% of younger) gave responses referringto people's acceptable desire to live in the best possible country, inthis case the U.S. (e.g., “Because this is a really good place to live,”“Because I think America is the best,” “Because if they want to movehere, it is okay because it is better here.”); and 10% of children (14% ofolder and 6% of younger) gave responses referring to the benefits toAmerica of having a diverse population (e.g., “If we didn't allow otherpeople tomove here, people would not have pizza, pasta, quesadillas,”“So people can make different friends.”). In addition, 11% of children(6% of older and 18% of younger) were unable to give a reason. Theseresponses differed by age, χ2 (3) = 9.18, p b .05, such that a greaterpercentage of younger children compared to older children said weshould allow immigration because of the acceptable desire to live in agood country (goodness of fit: χ2 [1] = 3.15, p b .05).

When asked if people should be allowed to move here from othercountries even when it is against the law, 82% of children said no (80%of younger and 84% of older children; no significant age difference). Toassess what children thought should happen to immigrants whomove here illegally, a chi-square goodness of fit test was conducted onchildren's endorsement of three possible orthogonal consequences.Children were more likely than expected by chance to endorse thenotion that immigrants should be allowed to stay in the U.S. if theyhave a job, χ2 (2) = 34.52, p b .001 (63% endorsed this; i.e., said “yes”rather than “no” or “maybe”). Children's endorsement that immi-grants should be sent back to their home country (27% endorsed this)or be put in jail (35% endorsed this) did not differ from chance. Toassess whether these attitudes differed by age, a chi-square test ofassociation was conducted for each consequence. Results indicated asignificant association between “they should be sent to jail” and agegroup, χ2 (2) = 16.65, p b .01, with 58% of younger children ascompared to 16% of older children saying that they should be put injail for breaking the law. The other possible consequences did notdiffer by age, with 36% of younger and 20% of older children endorsingthat they should be sent back to their home country, and 59% ofyounger and 65% of older children endorsing that they should beallowed to stay in the U.S. if they have jobs, all χ2s b 3.20.

Attributions for anti-immigrant sentimentsTo assess children's attributions for why some Americans are

opposed to immigration, a 2 (agegroup: younger, older)×6 (attributiontype: not smart, take too much money from government, criminal andbreak the law, take too many jobs, like different things, look different)mixedANOVAwas conducted, inwhich the last variablewas treated as awithin subjects factor. Analyses revealed a significant main effect forattribution type, F(5, 81) = 22.17, p b .001, η2 = .58. As can be seen inTable 1, Bonferonni analyses revealed that children rated the two mostaccurate reasons for anti-immigration sentiments to be because(a) people who move here like different things than other Americansand (b)peoplewhomovehere look verydifferent fromotherAmericans

(these attributions did not differ from one another). Children rated thesecondmost accurate reasons to be because (a) people whomove heretake too many good jobs from other Americans and (b) people whomove here are criminals (these attributions did not differ from oneanother). Children rated the attribution that peoplewhomovehere taketoo much money from the government as the next most accuratereason, followed by the attribution that people who move here are notsmart.

Stereotypes of and attitudes toward immigrants

To assess children's stereotypes about immigrants, and whetherthese stereotypes differed based on immigrants' country of origin andchild's conceptualization of their national in-group, children'sresponses to the vignettes were assessed. First, to assess whetherchildren with a stronger versus weaker American identity differed intheir stereotypes about immigrants, a variable was created forchildren with a strong American identity (children who ratedthemselves as being “very American” [n = 67]) and children with aweaker American identity (all other children [n = 23]). As themajority of children rated themselves as “very American,” and nochildren felt “not at all American,” the only comparison group waswith children with a less extreme American identity. Second, to assesswhether (a) children who hold a prototypical ideal of an American(i.e., differentiate between Americans based on their ethnicity) and(b) children who do not hold a prototypical ideal of an American (i.e.,do not differentiate between Americans) differed in their stereotypesabout immigrants, a variable was created to compare the two groupsof children. Specifically, children's ratings of “how American” thethree non-white American groups were (i.e., African American, AsianAmerican, Latino American) were averaged. This average score wassubtracted from the rating of the White American. Children (n = 33)whose difference score was 0 (i.e., they did not differ in their ratingsbetween Americans) were compared to children (n = 57) who ratedWhite Americans as more prototypically American than non-WhiteAmericans.

Next, a series of 2 (age group: younger, older) × 2 (Americanidentity: weaker, stronger) × 2 (prototypical American: Whiteprototype is more American vs. no difference) × 5 (country: Mexicanimmigrant, Chinese immigrant, British immigrant, White American,Black American) mixed ANOVAs was conducted, in which the lastvariable was treated as a within subjects factor. Regarding thequestion about how much money the immigrant earns, analysesrevealed a significantmain effect for country of origin, F(4, 78)= 3.53,p b .01, η2 = .15. Means are presented in Table 2. Bonferonni post hoctests indicated that children rated the White and Black American asearning more money than the Mexican immigrant. The BlackAmerican was also rated as earning more money than the Chineseand British immigrant. There were no effects of age or national in-group.

Regarding the question about how smart the child of theimmigrant is, analyses revealed a significant interaction betweenAmerican identity, age, and country of origin, F(4, 78) = 2.32, p b .05,η2 = .11. Means are presented in Table 3. Tests of simple effectsrevealed that there were no differences across country of origin foryounger children or older children who had weaker Americanidentity. Older children with strong American identity did differen-tiate, however, based on country of origin, such that both Americanchildren (i.e., White American and Black American) were rated assignificantly smarter than all three immigrant children.

Regarding the question about how much the child would want togo to school with the child of the immigrant, analyses revealed a maineffect for country of origin, F(4, 78) = 2.17, p b .05, η2 = .10. Meansare presented in Table 4 (Note: This test did not meet thehomogeneity of variance assumption, but because this effect wasbased on a within-subjects analysis, the F test is robust without the

Table 1Attributions for anti-immigrant sentiments by age group.

Attribution

Age group Not smart Take money Criminals Take jobs Different culture Different appearance

Younger 1.34 (.94) 1.82 (1.16) 2.24 (1.34) 2.03 (1.26) 2.58 (1.27) 2.97 (1.30)Older 1.39 (.84) 1.73 (.88) 1.96 (1.06) 2.10 (1.14) 2.35 (1.16) 2.24 (1.18)Combined 1.37 (.88)a 1.77 (1.01)b 2.08 (1.19)c 2.07 (1.19)c 2.45 (1.21)d 2.56 (1.28)d

Note. Numbers represent means (standard deviations). Scores range from 1 (not at all true) to 4 (very true). Different superscripts indicate significant differences (p b .05) acrossattribution types. Numbers in bold indicate attribution types with significant differences across age groups.

114 C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

assumption; see Lindman 1974; Lomax 2007). Post hoc tests indicatedthat children rated wanting to go to school with White and BlackAmericans' children higher than the Mexican and Chinese immi-grants' children. The British child did not differ from any others. Therewas also significant interaction between age, country of origin, andwhether they endorsed a White prototypical American, F(4, 78) =2.85, p b .05, η2 = .13. Analysis of simple effects and post hoc testsindicated that older children who rated White Americans as theprototypical American rated wanting to go to school with bothAmerican children significantly more than the Mexican child. Ratingsfor the British and Chinese immigrants' children did not differ fromany others. There were no differences across country of origin foryounger children or older children who did not hold a prototypicalideal of an American.

There were no differences (see Table 2) in how importantimmigrants' jobs were, how hard they worked, or how much theirtown wanted them to live there.

Discussion

The current study addressed the general questions of whatchildren think it means to be an American and what it means to bean immigrant. The first research question addressed children'sconceptualization of their national in-group. Results revealed thatthe children in this sample held a strong American identity. Further,although children considered some Americans to be more Americanthan others, they believed that being an American is attainable byanyone and not based on nativity. Specifically, although all groupmembers were described as being born in the U.S., White EuropeanAmerican children in this sample considered White Americans to be“more American” than Black Americans, who were in turn “moreAmerican” than Asian Americans, who were in turn “more American”than Latino Americans. Despite these distinctions, children did nothold strict restrictions on what makes a true American. Theyconsidered the most important criterion to be that an individualloves America and lives by its rules; this was significantly moreimportant than speaking English, being born in America, or living inthe U.S. a long time. These conceptualizations of their national in-group did not differ across age groups, and suggests that children byearly elementary school have conceptualizations of Americans similarto that of adults (Devos & Banaji 2005).

Table 2Stereotypes about immigrants based on region of origin.

Stereotypedomain

WhiteAmerican

BlackAmerican

Mexicanimmigrant

Britishimmigrant

Chineseimmigrant

Money 3.31 (.65)ac 3.39 (.66)a 3.01 (.80)b 3.16 (.73)bc 3.14 (.87)bc

Importance 3.34 (.82) 3.46 (.72) 3.29 (.84) 3.38 (.77) 3.30 (.84)Hard work 3.63 (.62) 3.82 (.41) 3.75 (.45) 3.76 (.48) 3.78 (.46)Townattitudes

3.08 (.85) 3.02 (.90) 2.89 (.93) 2.88 (.90) 2.86 (.97)

Note. Numbers represent means (standard deviations). Scores range from 1 (not at all)to 4 (very much). Different superscripts indicate significant differences (p b .05) acrossimmigrants.

The second research question addressed children's knowledgeand attitudes regarding the immigration debate. Overall, many(although not the majority) of the children in this sample were awareof the immigration debate and almost all held positive attitudes aboutlegal immigration. Specifically, one-half of older children and one-thirdof younger children knew that some people are against immigration.When asked why people immigrate to the U.S., the largest percentageof children, particularly older children, referred to immigrants seekingthe freedoms found in America. Children also noted that immigrantsmove to America for more financial and material opportunities and toescape harsh conditions in their native country. A considerable numberof younger children, however, could not give a reason for immigration,reiterating that children learnmore about immigration over the courseof middle childhood. When asked why people should be allowed tomove here, the majority of children, again particularly older children,noted the ideals of freedom. Children seemed to have internalized theconcept of freedom as a human right — it is both as a reason peopleseek out America and a reason we should allow them to live wherethey choose (see Torney-Purta et al. 2008, for an elaboration onchildren's understanding of human rights). The finding that freedomwasmentionedmore often by older than younger children is consistentwith cognitive developmental trends in which older children betterunderstand abstract concepts and ideals (Helwig & Jasiobedzka 2001;Keil 2006). Younger children, in particular, also commonly mentionedthat America is the best possible country to live in and it isunderstandable and acceptable for people to want to live here.

Although children in this sample are very positive about legalimmigration (with 97% favoring of it), children did not support illegalimmigration (with 82% opposed to it). However, despite being opposedto illegal immigration, the majority of children (63%) believedimmigrants should be allowed to stay in the U.S. if they hold a job.This is higher than the national polling data that indicate that less thanhalf of adults endorse this option as a means for illegal immigrants togain legal immigration status (see Polling Report, 2010). It is unclearfrom the current study, however, whether children believe immigrantsshould be allowed to stay in the U.S. regardless of employment. In otherwords, many children may simply be eschewing any punishment forillegal immigrants (in contrast to the majority of adults favoringdeporting or detaining illegal immigrants, Polling Report, 2010). Theonly finding in which children endorsed a punishment for illegalimmigrants is when the youngest children in this sample favored alegalistic response to breaking a law (i.e., being jailed, but not beingdeported). This finding is consistent with research showing thatyounger children believe that any legal violation is wrong and shouldbe punished (Tapp & Kohlberg 1971). In contrast, older children werevery unlikely to endorse placing illegal immigrants in jail, supportingprevious research showing that older children more so than youngerchildren weigh personal rights and freedoms when justifying legalviolations (Helwig & Jasiobedzka 2001). Thus, there may be adevelopmental trajectory in attitudes about illegal immigration inwhich young children favor a legal punishment (but only a legalpunishment) for breaking a law, and older children tend to focus onpersonal rights and reject any punishment for illegal immigrants(despite their disapproval). These attitudes appear to change again formany individuals as they reach adulthood, although the reasons for

Table 3Stereotypes about how smart the child of the immigrant is based on region of origin.

Age/national identity White American Black American Mexican immigrant British immigrant Chinese immigrant

OlderStrong American identity 3.53 (.60)a 3.47 (.51)a 3.24 (.59)b 3.21 (.62)b 3.05 (.65)b

Weaker American identity 3.23 (.59) 3.08 (.76) 3.31 (.63) 3.31 (.63) 3.54 (.52)Younger

Strong American identity 3.72 (.53) 3.79 (.41) 3.41 (.82) 3.66 (.72) 3.72 (.53)Weaker American identity 3.78 (.44) 3.44 (.73) 3.56 (.53) 3.33 (.50) 3.33 (.87)

Combined 3.57 (.58) 3.52 (.59) 3.34 (.67) 3.38 (.67) 3.37 (.68)

Note. Numbers represent means (standard deviations). Scores range from 1 (not at all) to 4 (very much). Different superscripts indicate significant differences (p b .05) acrossimmigrants.

115C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

adults' attitudes toward immigration are complex and multi-faceted.Future research should examine why children are opposed to illegalimmigration and why they think certain consequences are moreappropriate than others.

When askedwhy some people are against immigration, participantsattributed this sentiment to the notion that immigrants look differentand like different things from other Americans (as compared to takingtoomuchgovernmentmoney, being criminals, or taking toomany jobs).The finding that younger children, relative to older children, wereespecially likely to attribute anti-immigrant sentiments to the notionthat immigrants look different than other Americans is consistent withdevelopmental research. Considerable work has shown that childrenfocus on observable, external qualities in person perception (Aboud1988; Aboud & Levy 2000; Martin 1989) and often assume that skincolor differences reflect differences in the inherent essence of the group(known as racial essentialism, Hirschfeld 1996; Mahalingam 2007).Related, this developmental trend is similar tofindings in Brown (2006),in which younger children perceived teachers to be biased againststudents because of their skin colormore frequently thanolder children,most likely because of the salience of the physical differences betweenpeople. This findingmay also reflect the context inwhich these childrenlive, one in which the majority of the community is White EuropeanAmerican with an increasing Mexican immigrant population. To thechildren we tested, immigrants apparently do look different from theother Americans they see. Children in a community with a large Polishimmigrant groupmight be less likely to focus onphysical differences, forexample.

Regardless of the developmental differences, children overall moststrongly assumed that Americans are opposed to immigration becauseimmigrants look different and like different things from otherAmericans. These attributions do not parallel adults' beliefs (or atleast adults' explicitly stated beliefs) in symbolic or realistic threats(Stephan et al. 1999). In other words, children seem to endorse astrong positive type of patriotism (e.g., feeling “very American”)without also feeling the more chauvinistic type of patriotism (e.g.,“Outsiders threaten our country”). These attributions have potentiallyimportant implications. First, it indicates that children attribute anti-immigration sentiments to ethnic/cultural discrimination. Researchhas shown that children at this age are knowledgeable about ethnic

Table 4Stereotypes about how much they wanted to go to school with the child of the immigrant

Age/prototypical American White American Black American

OlderWhite prototypical American 3.19 (.83)a 3.03 (.80)a

No prototypical American 3.15 (.81) 3.30 (.86)Younger

White prototypical American 3.20 (1.0) 3.20 (1.2)No prototypical American 3.69 (.63) 3.54 (.66)

Combined 3.26 (.87)a 3.21 (.91)a

Note. Numbers represent means (standard deviations). Scores range from 1 (not at all) toimmigrants.

discrimination (Brown 2006; Brown & Bigler 2005), but no researchhas previously examined whether children perceive discriminationtoward immigrants. This study suggests that children in this sampleindeed perceive such discrimination. Second, this finding calls intoquestion adults' actual reasons for anti-immigration attitudes.Although purely speculative, perhaps children's responses reflectadults' actual (or even implicit) attitudes, minus the filter of socialdesirability that comes with age. Similarly, adults' endorsement ofsymbolic and realistic threats as reasons for anti-immigrationattitudes may be a form of symbolic racism (Tarman & Sears 2005),and children's responses may be a more explicit reflection of theracial/ethnic prejudices underlying anti-immigration sentiments.

The third research question assessed children's stereotypes andattitudes about immigrants based on their country of origin. Resultsindicated that children held differential attitudes about immigrantsbased on country of origin, with Mexican immigrants being viewedmost negatively (although, no attitudes were entirely negative, butrather less positive). For example, the Mexican immigrant wasperceived to make less money than either the Black or WhiteAmerican, his child was perceived to be less smart than either theBlack orWhite American, and childrenwanted to go to school with hischild less than the child of the Black orWhite American. This tendencyto view the Mexican immigrant most harshly is consistent withprevious research with adults (Lee & Fiske 2006; Short 2004), andreflects the most negative stereotypes in America being associatedwith Mexican immigrants. Although children reported not wanting togo to school with the Chinese immigrant relative to the two Americanchildren, children were less biased against the Chinese immigrantthan the Mexican immigrant. The slightly more favorable attitudestoward Chinese immigrants may be their associationwith the positive“model minority” stereotype about Asian Americans (e.g., Park 2008).The difference between attitudes about Mexican versus Chineseimmigrants also reflects the context in which the children in thecurrent sample live. They live in a community with a predominantlylow-income Mexican immigrant population who often hold manuallabor jobs, and a relatively high-income Chinese immigrant popula-tion who are often medical professionals or academics. Thus, becausethe community make-up parallels stereotypes supported in previousresearch, the pattern of children's attitudes is not surprising.

based on region of origin.

Mexican immigrant British immigrant Chinese immigrant

2.61 (.84)b 2.84 (.82)ab 2.84 (.89)ab

2.95 (.69) 3.15 (.75) 3.00 (.86)

3.20 (.91) 3.32 (.95) 2.84 (1.1)3.23 (1.0) 2.92 (1.2) 3.38 (.96)2.94 (.88)b 3.05 (.92)ab 2.96 (.96)b

4 (very much). Different superscripts indicate significant differences (p b .05) across

116 C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

In addition, as predicted, children did not hold many negativeattitudes or stereotypes about the British immigrant (i.e., they did notmind going to school with the British immigrant). British immigrantscan bemore easily incorporated into a common in-group identity (seeGaertner & Dovidio 2000) because of similarities in skin color andlanguage, and this has been associated with more positive attitudes inchildren as young as 5 (Powlishta et al. 1994). Thus, it is easier to viewBritish immigrants as similar to the prototypical American, andsubsequently it is more likely that Americans can apply their positivein-group attitudes to the British immigrant.

Importantly, the study indicates that the White EuropeanAmerican children in this sample did not show biases favoring theWhite American over the Black American. Decades of research haveindicated that European American children hold racial biases againstAfrican Americans, and hundreds of studies have attempted to reducesuch biases (for a brief review, see Pfeifer, Brown, & Juvonen 2007).The current findings suggest that children also hold biases (perhapseven stronger biases) against Mexican immigrants. Although childrenno doubt continue to hold racial biases not captured by this study,these findings demonstrate that immigration status is an importantdistinction among elementary school-aged children and should not beignored. The implication of this finding is that anti-bias interventionprograms in the U.S. – programs that rarely, if ever, address immigrantgroups – are overly restricted in focus and miss the groups towardwhich many children hold biases.

The final research question assessed the relationship betweennational identity and attitudes toward immigrants. Findings indicatethat older children with a strong American identity perceived the twoAmerican children to be smarter than all three immigrant children.This bias was not evident in childrenwith aweaker American identity.This finding supports both theory, namely social identity theory(Tajfel & Turner 2004) in which individuals who hold an importantgroup identity hold more negative attitudes about their out-groups,and previous research (Billiet et al. 2003; Reizábal et al. 2004;Verkuyten 2009). In addition, the degree to which children believedthat a prototypical American is White moderated some of theirattitudes about immigrants. Specifically, older children who believedthat a White American is the most prototypical American wanted togo to school with both American children more than the Mexicanimmigrant's child. This bias was not evident in children who did notdifferentiate across Americans. This suggests that children who holdtight restrictions around who is a true member of their national in-group hold more biases than children without such restrictions,suggesting that they are less likely to allowMexican immigrants (or atleast their children) into a common in-group. Both of thesemoderating variables were only relevant for older children. Previousresearch has found that national identity becomes more importantacross middle childhood (see Barrett, Lyons, & del Valle 2004), andperhaps the current finding reflects older children's more sophisti-cated integration of their in-group conceptualizations and their out-group attitudes.

As with all research, there are limitations to the current study. Forexample, the current sample was ethnically homogenous. Althoughethnicity has not been a significant predictor of attitudes towardimmigrants in previous work (Stephan et al. 1999), more currentresearch should explore ethnic group differences in children'sattitudes toward immigrants. Unlike the current study, futureresearch should address attitudes toward immigration amongchildren of immigrants. In addition, the current study examineschildren's attitudes about the immigration threats perceived byadults. Future research should examine whether children perceiveunique threats from immigrants that differ from those of adults andexplore the ways that children view important issues in immigration.Further, the current study did not assess parents' role in shapingchildren's attitudes about immigrants, and studies on the role ofsocialization in developing attitudes about immigration would be

valuable. Finally, future research should tease apart the numerousconfounds associated with stereotypes about immigrants. Immigrantsdiffer in terms of race, language, generational status, and socioeco-nomic status. All of these variables likely affect children's attitudes ofimmigrants, and thus should be examined independently.

In conclusion, this study suggests that elementary school agedchildren in this sample hold strong American identities and haveflexible ideas about what it means to be an American. Further, by age10, children seem to make important distinctions between bothAmericans and immigrants. White Americans were often consideredmore prototypical of Americans than other ethnic group members,legal immigrants were favored over illegal immigrants, and Mexicanimmigrants were viewed more negatively than immigrants fromother areas (perhaps a reflection of the sample community).Furthermore, children's own attitudes and American identities seemto influence their attitudes towards immigrants. Taken together, thisstudy suggests that researchers who are concerned with theintergroup attitudes of children (and likewise, interventions to reducenegative intergroup attitudes) should more frequently examinechildren's attitudes toward immigrants and immigration. Althoughthis is a heated debate among adults, and often concerns abstractlegal, financial, and logistical concepts, children are attending to thedebate, forming their own attitudes about the issues, and endorsingstereotypes related to their attitudes. As immigration shows few signsof appreciably slowing and is increasingly debated in the news,understanding children's conceptions of these issues is of growingimportance.

References

Aboud, F. E. (1988). Children and prejudice. Oxford: Basil Blackwell.Aboud, S., & Levy, S. (2000). Interventions to reduce prejudice and discrimination in

children and adolescents. In S. Oskamp (Ed.), Reducing prejudice and discrimination(pp. 269−293). Mahwah, NJ: Lawrence Erlbaum.

Barrett, M., Lyons, E., & del Valle, A. (2004). The development of national identity andsocial identity processes: Do social identity theory and self-categorisation theoryprovide useful heuristic frameworks for developmental research? The developmentof the social self (pp. 159−188). New York, NY: Psychology Press.

Barrett, M., & Short, J. (1992). Images of European people in a group of 5–10-year-oldEnglish schoolchildren. British Journal of Developmental Psychology, 10(4),339−363.

Barrett, M., Wilson, H., & Lyons, E. (2003). The development of national in-group bias:English children's attributions of characteristics to English, American and Germanpeople. British Journal of Developmental Psychology, 21(2), 193−220.

Bennett, M., Barrett, M., Karakozov, R., Kipiani, G., Lyons, E., Pavlenko, V., et al. (2004).Young children's evaluations of the ingroup and of outgroups: A multi-nationalstudy. Social Development, 13, 124−141.

Bennett,M., Lyons, E., Sani, F., &Barrett,M. (1998). Children's subjective identificationwiththe group and in-group favoritism. Developmental Psychology, 34(5), 902−909.

Billiet, J., Maddens, B., & Beerten, R. (2003). National identity and attitude towardforeigners in amultinational state: A replication. Political Psychology, 24(2), 241−257.

Brown, C. S. (2006). Bias at school: Perceptions of racial/ethnic discrimination amongLatino and European American children. Cognitive Development, 21(4), 401−419.

Brown, C. S., & Bigler, R. (2005). Children's perceptions of discrimination: Adevelopmental model. Child Development, 76(3), 533−553.

Brown, C. S., Mistry, R. S., & Bigler, R. S. (2007). Hurricane Katrina: Children'sperceptions of race, class, and government involvement amid a national crisis.Analyses of Social Issues and Public Policy, 7, 191−208.

CNN (2007). Thousands of immigrationmarchers rally across U.S. Retrieved onOctober 26,2009 from. http://www.cnn.com/2007/US/05/01/immigration.protests/index.html

Deaux, K. (2006). A nation of immigrants: Living our legacy. Journal of Social Issues, 62(3),633−651.

Deaux, K. (2008). To be an American: Immigration, hyphenation, and incorporation.Journal of Social Issues, 64, 925−943.

Devos, T., & Banaji, M. (2005). American = White? Journal of Personality and SocialPsychology, 88(3), 447−466.

Devos, T., & Ma, D. S. (2008). Is Kate Winslet more American than Lucy Liu? The impactof construal processes on the implicit ascription of a national identity. The BritishJournal of Social Psychology, 47, 191−215.

Enesco, I., Navarro, A., Paradela, I., & Guerrero, S. (2005). Stereotypes and beliefs aboutdifferent ethnic groups in Spain. A study with Spanish and Latin American childrenliving in Madrid. Journal of Applied Developmental Psychology, 26(6), 638−659.

Esses, V., Dovidio, J., Jackson, L., & Armstrong, T. (2001). The immigration dilemma: Therole of perceived group competition, ethnic prejudice, and national identity. Journalof Social Issues, 57(3), 389−412.

Gaertner, S., & Dovidio, J. (2000). Reducing intergroup bias: The common ingroup identitymodel. New York, NY: Psychology Press.

117C.S. Brown / Journal of Applied Developmental Psychology 32 (2011) 109–117

Gaertner, S., & Dovidio, J. (2009). A common ingroup identity: A categorization-basedapproach for reducing intergroup bias. Handbook of prejudice, stereotyping, anddiscrimination (pp. 489−505). New York, NY: Psychology Press.

Helwig, C., & Jasiobedzka, U. (2001). The relation between law and morality: Children'sreasoning about socially beneficial and unjust laws. Child Development, 72(5),1382−1393.

Hernandez, D. J., Denton, N. A., & Macartney, S. E. (2008). Children in immigrantfamilies: Looking to America's future. Social Policy Report, 22(3).

Hirschfeld, L. (1996). Race in the making. Cambridge, MA: MIT Press.Inhelder, B., & Piaget, J. (1964). The early growth of logic in the child.New York, NY: Norton.Keil, F. (2006). Cognitive science and cognitive development. Handbook of child

psychology: Vol 2, Cognition, perception, and language (6th ed., pp. 609−635).Hoboken, NJ: John Wiley & Sons.

Killen, M., McGlothlin, H., & Henning, A. (2008). Explicit judgments and implicit bias. InS. R. Levy & M. Killen (Eds.), Intergroup attitudes and relations in childhood throughadulthood (pp. 126−145). New York, NY: Oxford University Press.

Lee, T., & Fiske, S. (2006).Not anoutgroup,not yet an ingroup: Immigrants in the stereotypecontent model. International Journal of Intercultural Relations, 30(6), 751−768.

Lindman, H. R. (1974). Analysis of variance in complex experimental designs. SanFrancisco, CA: W. H. Freeman & Co.

Lomax, R. G. (2007). An introduction to statistical concepts for education and behavioralsciences (2nd ed.). New York, NY: Routledge.

Mahalingam, R. (2007). Essentialism, power, and the representation of socialcategories: A folk sociology perspective. Human Development, 50, 300−319.

Martin, C. (1989). Children's use of gender-related information in making socialjudgments. Developmental Psychology, 25, 80−88.

Monger, R., & Rytina, N. (2009). Annual flow report. U.S. legal permanent residents:2008. Homeland security. Retrieved September 21, 2009 from. http://www.dhs.gov/xlibrary/assets/statistics/publications/lpr_fr_2008.pdf

Park, J. (2008). Second-generation Asian American pan-ethnic identity: Pluralizedmeanings of a racial label. Sociological Perspectives, 51(3), 541−561.

Pfeifer, J. H., Brown, C. S., & Juvonen, J. (2007). Fifty years since Brown v. Board ofEducation: Lessons learned about the development and reduction of children'sprejudice. Social Policy Report, 21(2).

Pfeifer, J., Rubble, D., Bachman, M., Alvarez, J., Cameron, J., & Fuligni, A. (2007). Socialidentities and intergroupbias in immigrant andnonimmigrantchildren.DevelopmentalPsychology, 43(2), 496−507.

Polling Report (2010). Immigration. Polling Report, Inc. Retrieved December 7, 2010from. http://www.pollingreport.com/immigration.htm

Powlishta, K., Serbin, L., Doyle, A., & White, D. (1994). Gender, ethnic, and body typebiases: The generality of prejudice in childhood. Developmental Psychology, 30(4),526−536.

Reizábal, L., Valencia, J., & Barrett, M. (2004). National identifications and attitudes tonational ingroups and outgroups amongst children living in the Basque country.Infant and Child Development, 13(1), 1−20.

Short, R. (2004). Justice, politics, and prejudice regarding immigration attitudes.Current Research in Social Psychology, 9(14).

Stephan, W., Ybarra, O., & Bachman, G. (1999). Prejudice toward immigrants. Journal ofApplied Social Psychology, 29(11), 2221−2237.

Tajfel, H., & Turner, J. (2004). The social identity theory of intergroup behavior. Politicalpsychology: Key readings (pp. 276−293). New York, NY: Psychology Press.

Tapp, J., & Kohlberg, L. (1971). Developing senses of law and legal justice. Journal ofSocial Issues, 27(2), 65−91.

Tarman, C., & Sears, D. (2005). The conceptualization and measurement of symbolicracism. The Journal of Politics, 67(3), 731−761.

Torney-Purta, J., Wilkenfeld, B., & Barber, C. (2008). How adolescents in 27 countriesunderstand, support, and practice human rights. Journal of Social Issues, 64, 857−880.

Verkuyten, M. (2009). Support for multiculturalism and minority rights: The role ofnational identification and out-group threat. Social Justice Research, 22(1), 31−52.

Verkuyten, M., & Steenhuis, A. (2005). Preadolescents' understanding and reasoningabout asylum seeker peers and friendships. Journal of Applied DevelopmentalPsychology, 26(6), 660−679.