Most research on US children’s and adolescents’ prejudice hasfocused on the extent to which European-American partici-pants hold prejudicial attitudes about minority individuals,particularly African-American and Latino children (Aboud &Amato, 2001; Killen, Margie, & Sinno, 2006). As the urbanareas around the globe become more heterogeneous, however,it is increasingly important to understand how individuals,both from majority and minority ethnic backgrounds, evaluateeveryday interracial interactions, particularly peer relationships(Graham & Juvonen, 2002). This approach extends pastresearch that focused on European-American children’sassignment of negative traits to others based on skin color toa focus on how children from a range of ethnic backgroundsevaluate intergroup relationships, specifically in peer contexts.The former approach provided a diagnostic test of whethermajority children associated negative terms with pictures ofchildren who represented minority groups (Aboud & Levy,2000). In contrast, the latter perspective has examined howchildren, majority and minority, evaluate social exchangesinvolving exclusion, and make attributions of intentions insocial encounters. Our approach stems from research on socialcognitive development (Smetana, 2006), which examines howchildren evaluate familiar everyday social experiences. Thefindings reveal that understanding children’s social interpret-ations and evaluations of their experiences provides infor-

mation about their behavior, both normative and clinical(Lemerise & Arsenio, 2000).

Recent social psychological research with adults in the areaof intergroup attitudes about race and ethnicity has beenaimed at understanding the “target’s perspective,” whichinvolves investigating how individuals who are typically thetarget of discriminatory and prejudicial behavior evaluateinterracial interactions (Swim & Stangor, 1998). Research withadolescents has indicated that experiences of discriminationlead to stress and anxiety (Fisher, Wallace, & Fenton, 2000),and that exclusion from groups as a function of ethnicitycontributes to depression and low motivation in adults(Baumeister, Twenge, & Nuss, 2002). Further, children’sperception of discrimination provides information regardingtheir behavior with one another (Brown & Bigler, 2005). Yetvery little research has been conducted on minority children’sviewpoints about racial exclusion in school and home contexts(Aboud, 2005). For example, what forms of reasoning dominority children and adolescents use to evaluate racialexclusion, and how often do they perceive themselves to be thetarget of racial exclusion by their peers? Further, when domajority children become aware that racial exclusion occurs,and how often do they perceive that it happens? Answers tothese questions will help to understand how children andadolescents from different ethnic backgrounds interpret

International Journal of Behavioral Development2007, 31 (5), 491–500

http://www.sagepublications.com

© 2007 The International Society for theStudy of Behavioural Development

DOI: 10.1177/0165025407081478

Evaluations of interracial peer encounters by majority and minority USchildren and adolescents

Melanie Killen, David Crystal Martin RuckAlexandra Henning, Georgetown University, USA Graduate Center, CUNY, USAMegan Clark Kelly

University of Maryland, USA

US majority (European-American) and minority (African-American, Latin-American, Asian-American) children were interviewed regarding race-based and non-race based reasons for exclusionin interracial peer contexts (N = 685), evenly divided by gender at 9, 12, and 15 years of age (4th,7th, and 10th grades) attending 13 US public schools. All children judged race-based exclusion aswrong using moral reasons. In contrast, non-race based reasons such as lack of shared interests,parental discomfort, and peer pressure were viewed as more wrong by minority children than bymajority children.This pattern decreased with age for all participants. Further minority children weremore likely to expect that racial exclusion occurs, indicating that ethnic background, social experi-ence, and age are significantly related to interpretations of interracial peer motives for exclusion.

Keywords: exclusion; intergroup attitudes; moral reasoning; peer relations; prejudice; racial attitudes;social cognition

Correspondence concerning this article should be addressed to:Melanie Killen, Department of Human Development, Universityof Maryland, College Park, MD, 20742–1131, USA; e-mail: mkillen@umd.edu

The research described in this paper was supported by a grant fromthe National Institute of Child Health and Human Development(1R01HD04121–01). We thank the research team at the University ofMaryland: Marguerite Adams, Holly Bozeman, Alaina Brenick,Michael J. Collins, Christina Edmonds, Julia Hadricky, Nancy Geyelin

Margie, Heidi McGlothlin, Alexander O’Connor, Christine Pitocchelli,and Stefanie Sinno, and the research team at Georgetown University:Nancy Gibbs, Lindsey Gansen, Dayna McGill, Elizabeth Kravec,Anne Blossom, Mia Shorteno-Fraser, and Meredith Mellody forassistance with instrument development, data collection, data coding,analyses, and reliability coding. In addition, we thank Nancy GeyelinMargie and Cameron Richardson for technical assistance with themanuscript, and the editors for their helpful comments.

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

interracial peer situations, which, in turn, will provide valuableinsight into peer conflicts that result from intergroupinteractions.

Recently social psychologists have provided a conceptualframework for examining exclusion, and for focusing on themultiple dimensions of decision-making around exclusion in arange of social relationships and contexts (Abrams, Hogg, &Marques, 2005). The findings indicate that the psychologicalconsequences of exclusion based on group membership, suchas ethnicity, can be quite negative, and that the motives andresponses to exclusion from individuals are complex. Littleresearch has been conducted on how children and adolescentsevaluate exclusion based on ethnicity. An exception is a recentdevelopmental study with minority and majority childrenwhich found that the vast majority of all children judged thatexplicit race-based exclusion from peer groups was wrong,using moral reasons such as unfair treatment (Killen, Lee-Kim, McGlothlin, & Stangor, 2002). This pattern reflected asimilar pattern of findings found in social psychology researchwith adults on explicit racism, which has shown that overtracist statements are very rare. Yet, while explicit racism hasdecreased dramatically over the past 50 years in the US,intergroup bias is still quite prevalent (Dovidio, Glick, &Rudman, 2005; Gaertner & Dovidio, 2005) and this has beendemonstrated extensively with adult samples.

Findings from studies using more indirect measures tocapture intergroup bias, however, have found significant differ-ences between majority and minority youth. For example,investigations of attributions of intention using ambiguouspictures have revealed that while majority and minoritychildren make similar judgments about attributions ofintention (with some qualifications as described below), theyhave different viewpoints about the potential for cross-racerelationships (Margie, Killen, Sinno, & McGlothlin, 2005;McGlothlin, Killen, & Edmonds, 2005). This is important ascross-race friendships have been shown to be one of the mostsignificant predictors for a reduction in prejudice (Pettigrew &Tropp, 2000). Further, in one study, European-Americanmajority children with little contact with minority children inschool settings demonstrated an in-group bias when makingattributions about intentions (McGlothlin & Killen, 2006), abias not found for European-American majority children whohad regular contact with minority children in school settings(McGlothlin et al., 2005). These findings tell us that racialbiases exist when children are making decisions about peerencounters, and that these biases may not be apparent whenusing more explicit measures.

This leaves us with several alternative interpretations ofthese mixed findings. On the one hand, direct measures ofjudgments reveal that majority and minority children viewracial exclusion in peer encounters as wrong, and on the otherhand, indirect measures reveal that majority children attributenegative intentions to minority children when evaluatingambiguous situations involving interracial peer encounters.What is not known is whether the racial biases that are revealedby indirect measures manifest when children and adolescentsare making explicit judgments about exclusion. In the Killenet al. (2002) study, children were explicitly asked about racialexclusion (e.g., “Is it all right or not all right for a group toexclude X?”). Further, the situations involved a friendship, apeer group, and an institutional setting (e.g., a school excludesgroup X).

The novel dimension of the present study was to extend the

findings from Killen et al. (2002) by using indirect (rather thandirect) measures of judgments about exclusion, and tomeasure children’s evaluations of interactions that reflected“intimate” relationships. Researchers and theorists haveproposed that one of the last areas for integrated interactionsand relationships is found in those involving intimate relation-ships, such as dating and marriage (Kennedy, 2003). Thus, inthe present study, complex exclusion situations involvingvarying levels of intimacy, such as friendship, a sleepover atone’s house, and dating, were described to participants.Further, the assessments reflected a mixture of direct andindirect assessments. The few studies conducted on implicitand intergroup bias with children have been conducted withvery young children (Rutland, Cameron, Milne, & McGeorge,2005). Thus, more research needs to be conducted with olderchildren and adolescents, particularly as much evidence indi-cates that exclusion based on group membership increasesduring the adolescent years (Horn, 2003).

In the present study, children and adolescents at 9, 12, and15 years of age (4th, 7th, and 10th grades) were interviewedregarding their interpretations and evaluations of three inter-racial peer encounters that provided a range of intimaterelationships. These three age groups were selected for severalreasons. First, interviewing children at these three age groupsenabled direct comparisons with previous studies using thesame age group with direct, explicit measures (Killen et al.,2002). Second, prior findings revealed a significant increasefrom 9 to 15 years of age regarding the legitimacy of exclusionbased on group functioning. Using more indirect methods, itwas expected that group functioning justifications wouldemerge earlier, even by 12 years of age, and that reasonsallowing exclusion would be revealed due to the more intimatenature of the interactions. Further, this age range was ofinterest due to the emergence of stereotypes based on race andethnicity during early and mid-adolescence (Killen et al.,2006). As the focus of this study was on relationships thatincreased in intimacy, early to mid-adolescence was appropri-ate given the increase in dating and romantic partners duringthis developmental period (Rubin, Bukowski, & Parker,2006).

Third, our sample reflected children enrolled in threedifferent types of school environments in the US: elementaryschool (4th grade), middle school (7th grade) and high school(10th grade).These school levels reflected an increasing contactwith a more diverse student body in this geographic region,and this increased heterogeneity was expected to contribute tochildren’s experience with diversity, and, in turn, with theirinclusion judgments. At the same time, middle-school and highschool are times when social cliques and peer networks peak,and become a salient influence on children’s inclusion andexclusion attitudes (Horn, 2003). It could be that the increas-ing diversity provides a positive source of influence oninclusion judgments, while the increasing peer group identifi-cation provides a negative source of influence on such judg-ments. While we were not able to directly test these alternativesources of influence in this study, determining the contexts inwhich we found age-related changes, and differences as afunction of ethnic status, was expected to provide additionalinformation about varying sources of influence for futureempirical investigations.

The non-race related exclusion factors described in thescenarios included non-shared interests (one child likes sportsand the other does not), lack of familiarity (the child is not

492 KILLEN ET AL. / EVALUATIONS OF INTERRACIAL PEER ENCOUNTERS

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

known to the parents), non-shared school affiliation (thepotential date is from a rival high school). We probed partici-pants’ reasons for why the exclusion occurred, and we askedparticipants to rate the wrongfulness of different reasons forexclusion (race-based, non-race-based), including a generalreason pertaining to group functioning (What if the groupwon’t work well with someone who does not “fit” in?). Weincluded the latter assessment based on previous findings inwhich group functioning was used as a basis for gender andrace-based exclusion decisions by majority adolescents (Killen& Stangor, 2001).

For the current study, we were interested in determiningwhether there were differences in terms of how children andadolescents from majority European-American and minorityethnic backgrounds evaluate peer exclusion.There are multiplebases for expecting that minority and majority children woulddiffer in their evaluations of interracial peer exclusionscenarios. Potentially these factors include different levels ofexperience with members of out-groups, different prior experi-ence with exclusion, and different messages about such inter-actions from family, school, and society. Social psychologistshave theorized that ethnic identity is related to intergroupattitudes, yet very little empirical research has been conductedon this dimension of children’s social experiences, with a fewexceptions (Rutland, Cameron, Bennett, & Ferrell, 2005).Moreover, in a recent analysis of intergroup contact research,Dixon and colleagues (Dixon, Durrheim, & Tredoux, 2005)called for research that investigates how participants evaluateeveryday interracial exchanges rather than research whichrelies solely on survey methods for documenting how thesetypes of exchanges are experienced and interpreted. Theformer approach was used in the present study. We conductedone-on-one interviews with individuals to assess how theyinterpret and evaluate interracial interactions involvingexclusion.

Thus, our measure was more direct than methodologiesemploying implicit bias (Baron & Banaji, 2006; Rutland et al.,2005), and yet, less direct than methodologies probing explicitdecisions to exclude solely on the basis of race (Killen et al.,2002). We investigated children’s explicit judgments usingsituations that were ambiguous in terms of the reason forexclusion (not whether the exclusion has occurred). Thisprovided information about decision-making in children’severyday lives regarding peer exchanges (unlike measures ofimplicit biases which assess associations between words andfaces such as the IAT), and, at the same time, allowed us toassess how interpretations of situations vary as a function ofone’s ethnicity.

Our three interracial contexts were Lunch (one student doesnot want to invite another student to join him/her with a friendat lunchtime), Sleepover (one student does not want to inviteanother student to a sleepover party), and Dance (one studentdoes not want to invite another student to a school dance). Wetested hypotheses about whether participants would be morewilling to exclude someone in a peer situation (Lunch, Dance)than in a situation involving parental expectations (Sleepover).In all three situations, the excluder was European-Americanand the target of exclusion was African-American.

Hypotheses about context, ethnicity, and gradeIn the two peer contexts (lunch and dating), we hypothesizedthat there would be no age or ethnic differences for evaluations

of race-based exclusion (everyone would judge it to be wrong)based on prior findings in which straightforward exclusionbased on race was viewed as wrong by children from a widerange of ethnic groups (Killen et al., 2002). We expected ageand ethnicity differences regarding whether it was all right toexclude an African-American child from an interracial peersituation in which non-race reasons, such as non-sharedinterests, unfamiliarity, and disruptions of group functioningwere espoused by the excluder.

Further, we predicted that, with age, majority childrenwould be more likely than minority children to view non-racebased reasons such as non-shared interests and group func-tioning as legitimate reasons to exclude someone. Prior studieshave shown that majority children refer spontaneously to groupfunctioning as a reason to justify exclusion (Killen & Stangor,2001). This study was different from prior research in thatparticipants were asked about different aspects of interactionsthat contribute towards group discomfort (lack of interests,unfamiliarity, and group functioning in three contexts, respec-tively). Given their prior experiences with discrimination(Fisher, Jackson, & Villarruel, 1998), we expected thatminority children would view reasons such as “unfamiliarity”and “she doesn’t fit the group” as a proxy for bias or preju-dice, and thus would be less likely to view group functioningas a legitimate reason to exclude. At the same time, we alsoexpected that all children would be less likely to rate non-racebased exclusion as wrong, with age, given that adolescentsbecome aware of multiple considerations involved in exclusiondecisions (Killen & Stangor, 2001). Further, we hypothesizedthat, in explaining their evaluations, minority children woulduse moral justifications that reflected empathy as well asfairness given prior experiences with exclusion.

In the home context, which involved parental concerns incontrast to messages from peers, we predicted that all childrenwould rate a parent’s decision to exclude a child’s friend at asleepover based on race as wrong.The reasons would be moral,that is, that it would be wrong due to prejudice. Yet, weexpected that exclusion in the home based on nonracialreasons would be evaluated differently, with minority childrenviewing such nonracial exclusion as more wrong than majoritychildren because they may be more likely to interpret non-racial reasons as an excuse for excluding someone because ofrace.

In previous studies on exclusion it has been shown thatmajority children, with age, challenged parental decisionswhich condoned racial/ethnic exclusion of peer relationships(Killen et al., 2002). However, majority children may be lesslikely to challenge parents’ expressions of discomfort oversomeone “unfamiliar” in the home context as this is a moreindirect reflection of parental viewpoints about racialexclusion. What is not known is whether minority childrendiffer from majority children regarding parental expressions ofdiscomfort, even when described for reasons other than “race”such as “unfamiliarity.” Minority children have experiencedexclusion from majority adults in a wide range of contexts inthe US (Fisher et al., 1998), and they may interpret majorityparents’ “discomfort” with someone different as reflective ofracial bias which would lead to their rejection of this type ofreason. In fact, studies have shown that European-Americanparents convey negative messages to children and adolescentsabout interracial friendships and dating (Killian, 2001; Yancey,1998), and extensive research has shown that interracialfriendships decrease with age (DuBois & Hirsch, 1990; Smith

INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2007, 31 (5), 491–500 493

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

& Schneider, 2000). These findings led us to hypothesize thatminority children, particularly adolescents, would be morecritical of a majority parent’s rejection of a friend for a sleep-over than would majority children.

Hypotheses about estimations of the frequency ofexclusionWe analyzed children’s estimations of the likelihood ofexclusion in order to determine whether minority and majoritychildren differed in their interpretations of how often racialexclusion occurs in various contexts. This was in line withDixon and colleagues’ (Dixon et al., 2005) recommendationthat in-depth analyses of how individuals reflect on interracialcontact is needed. Moreover, this analysis provided anothercheck on whether children in our sample, who attended thesame school district, reflected on interracial exclusion in thesame way. We predicted that perceptions of how often peerexclusion (race- and non-race-based) occurs would also differfor majority and minority children, particularly as a functionof the context of exclusion (school versus home). We expectedthat differences in perceptions of the frequency of exclusionwould be greater for the home context given that residentialsegregation is more pervasive than school segregation, particu-larly in the area that we collected our data. Specifically, weexpected that minority children would estimate that exclusionin interracial interactions (even when described as motivatedfor non-race based reasons) occurs more often than wouldestimate majority children given their prior experiences andknowledge of discrimination.

SummaryIn sum, we expected that (1) majority and minority childrenwould not differ regarding the wrongfulness of exclusion ininterracial peer contexts when the basis for exclusion wasexplicitly about race; (2) majority and minority children woulddiffer regarding empathy justifications for ratings of race-basedexclusion with minority children making more references toempathy; (3) minority children would rate non-race basedexclusion as more wrong than would majority children; and (4)that all children would rate non-race based exclusion as lesswrong with age. We also expected that (5) minority childrenwould estimate that non-race based exclusion occurs moreoften than would majority children.

Method

ParticipantsParticipants were 685 children and adolescents, ages 9, 12, and15 years of age (4th, 7th, and 10th grades in the US), attending13 public schools in mixed-ethnicity (range for school compo-sition was 20% to 45% minority) suburbs of a mid-size city inthe greater metropolitan Washington, DC, region of theMiddle-Atlantic region of the US. There were 94 girls and 70boys in 4th grade (M = 9.85 years, SD = .42), 167 girls and113 boys in 7th grade (M = 12.86 years, SD = .49), and 133girls and 108 boys in 10th grade (M = 15.89 years, SD = .52).The ethnic breakdown overall was 60% majority (EuropeanAmerican), and 40% minority (African-American, 14%;Asian-American, 12%; Biracial, 8%; Latin-American, 5%).

The ethnicity breakdown for each grade was: 4th grade, 115majority and 49 minority children; 7th grade, 172 majoritychildren and 108 minority children; and, 10th grade, 127majority and 113 minority children. The sample was relativelyevenly divided by gender for each ethnic group. All childrenwere from the same socioeconomic bracket, which was middle-income to working-income backgrounds.

Procedure and instrumentsWritten parental consent (average response rate = 80%) wasobtained for all children taking part in the study. Children wereindividually interviewed in a quiet room at their school by atrained research assistant who was matched with the partici-pants by race/ethnicity. Extensive pilot testing was conductedon the interview scenarios to ensure that children andadolescents were familiar with the situations, and that thelanguage was developmentally appropriate. Prior to beginningeach interview individual student assent was obtained andparticipants were assured of the voluntary, confidential andanonymous nature of the study. Interviewers administered theSocial Reasoning about Exclusion interview, which was audio-taped and later transcribed for coding purposes. Participantswere read three stories, each representing a different contextin which racial exclusion could occur. The three contexts were:Lunch (choice about cross-race friendship), Dance (cross-racedating in high school), and Sleepover (having a cross-racefriend in the home). The first two contexts were school-basedand the third context was home-based. The stories were readin the following order to all participants: Lunch, Sleepover,and Dance. As described earlier, all stories portrayed aEuropean American child excluding an African Americanchild.

After each story, participants were required to respond toeight assessments:

1 Wrongfulness of Racial Motives (the ratings of participants’judgments about the wrongfulness of exclusion if it werebased on race (such as “What if Michael thinks that theywon’t have much in common because Doug is Black?”);

2 Justification (and why?);3 Wrongfulness of Non-Racial Motives (such as “What if

Michael thinks that they won’t have much in commonbecause Doug doesn’t like sports?”);

4 Justification (and why?);5 Wrongfulness of Group Functioning Motives (such as “What if

Michael doesn’t invite Doug to lunch because he thinksDoug won’t fit with in with William and him?”);

6 Justification (and why?);7 Estimations of Race-Based Exclusion (estimates of the

frequency of exclusion among peers based on race, such as“How often do you think kids your age might not invitesomeone to lunch because they do not share the same inter-ests?”); and

8 Estimations of Non-Race Based Exclusion (such as “How oftendo you think kids your age might not invite someone tolunch because they are a different race?”).

Participants’ responses were audio-taped and later tran-scribed. Responses for all wrongfulness rating assessmentsranged from 1 (“very, very good”) to 8 (“very, very bad”).Responses for the two frequency estimation rating assessmentsranged from 1 (“never”) to 5 (“always”). Coding wasconducted on participants’ justifications for their ratings using

494 KILLEN ET AL. / EVALUATIONS OF INTERRACIAL PEER ENCOUNTERS

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

four social cognitive domain categories, based on several sub-categories identified in prior research (Killen et al., 2002;Smetana, 2006), which were collapsed for the purposes of thisstudy. The three overall categories were: Moral (Racial Preju-dice, Discrimination, and Empathy), Social-Conventional(Group functioning, Traditions, Customs), Stereotypes, andUncodable. There were three subcategories for Moral andSocial-Conventional which were analyzed for specific hypothe-sis testing, as described in the next section. For each partici-pant, justifications were scored dichotomously with a score of1 indicating that the category was used and a score of 0 indi-cating that the category was not used. Reliability coding wasconducted on 30% of the interviews by research assistantstrained on the coding system. Cohen’s kappas ranged from .81to 1.00. Uncertainties or discrepancies in the coding wereinitially resolved through discussion among the coders. Afterthis step, the coding categories were established, and a newdata set was used for reliability with coding discrepanciesdetermined by the primary coder.

Results

Plan for analysisAnalyses of Variance (ANOVAs) with repeated measures wereused to test hypotheses pertaining to responses to three assess-ments: Wrongfulness Ratings, Justifications, and Estimations ofExclusion. Follow-up analyses included univariate ANOVAs forbetween-subjects effects and t-tests for within-subjects inter-actions effects. In cases where sphericity was not met, correc-tions were made using the Huynh-Feldt method. Initialanalyses examining within-minority ethnicity effects on themajor variables were not significant, thus these participantgroups were collapsed into one “minority” category, which wascompared with the “majority” group comprised of European-Americans. In addition, gender was not a significant variablein preliminary analyses and was omitted from all subsequentanalyses. Ratings and Justifications were analyzed with ethnic-ity of participant and grade of participant as independent vari-ables. The repeated-measures factors were context (Lunch,Dance, and Sleepover).

Wrongfulness ratings for race-based peer exclusion ininterracial contexts at schoolA 3 (grade: 4th, 7th, 10th) ! 2 (ethnicity: majority, minority) !2 (context: lunch, dance) Analysis of Variance (ANOVA) withrepeated measures on the last factor was conducted on partici-pants’ ratings of the two peer contexts in which participantswere asked for their evaluations using race as a reason forexclusion. These means are displayed in Table 1. As predicted,the vast majority of all participants evaluated race-basedexclusion as wrong (M = 7.3). There were no differences forcontext or for ethnic status. There was an overall age effect,F(2,673) = 5.73, p < .001, with follow-up tests indicating thatwrongfulness ratings increased from the youngest group to thetwo older groups (ps < .05). Thus, when participants weredirectly asked about whether exclusion based on race waswrong, there was an increase from 4th to 10th grade in ratingsof wrongfulness. We did not predict age-related findings giventhat no prior research had documented this type of increasewith age in the wrongfulness of racial exclusion, and thus, thiswas a novel finding.

In addition, 2 separate 3 (grade: 4th, 7th, 10th) ! 2 (ethnicity:majority, minority) ! 3 (type of exclusion: race-based, non-race based, group functioning) ANOVA with repeatedmeasures on the last factor was conducted on participants’ratings of all three reasons for exclusion within each scenario.As expected, children viewed race-based reasons as morewrong than non-race based reasons or group functioningreasons: for Lunch: F(2,1350) = 458.99, p < .0001, for typeof exclusion, and for Dance, F(2,1346) = 365.14, p < .0001,for type of exclusion. Follow-up analyses indicated that allchildren rated race-based exclusion as more wrong than theother two reasons (non-race based and group functioning) andthere were no significant differences between the last tworeasons (the means are displayed in Table 1).

Participants’ reasons for their ratings of wrongfulness wereevaluated and analyses revealed that the vast majority ofparticipants used moral reasons to evaluate race-basedexclusion as wrong (M = .96). Analyses for sub-types of moralreasons used revealed ethnicity and grade differences, however,as well as differences for the context of exclusion. The vastmajority of all children cited reasons of racial prejudice inevaluating the wrongfulness of race-based exclusion (M = .83),and this type of reasoning increased with age, as indicated bythe analyses that majority children used more statements aboutprejudice (M = .87) than did minority children (M = .80),F(1,677) = 13.60, p < .0001. As hypothesized, minoritychildren used more empathy statements than did majoritychildren, F(1,677) = 22.11, p < .0001. Regarding age-relatedpatterns, 4th and 10th grade minority children used moreempathy statements (Ms = .14, .11, for 4th and 10th gradechildren) than did their majority counterparts who very rarelyreferred to empathy (Ms = .05, .04). Overall, all childrenexplained their evaluations of using race as a reason forexclusion with moral justifications (references to the wrongful-ness of racial prejudice and appeals to the feelings of others,such as empathy).

Wrongfulness ratings for non-race based reasons forpeer exclusion in interracial interaction at schoolTo test our hypotheses about whether majority and minoritychildren would differentially evaluate non-racial exclusion ininterracial peer contexts at school (referred to as “non-race”and “group functioning” in Table 1), we analyzed ratings ofwrongfulness. A 3 (grade) ! 2 (ethnic status) ! 4 (non-racebased exclusion decisions) ANOVA with the repeatedmeasures on the last factor revealed a main effect for grade,F(2,673) = 37.36, p < .0001, and for ethnic status, F(1,673),= 10.98, p < .001. As indicated in Figure 1, wrongfulnessratings decreased with age, p < .01, and, consonant withhypotheses, overall, minority children evaluated non-racebased exclusion in an interracial context as more wrong thandid majority children, p < .05. (See Table 1 for means.)

The explicit reasons given by participants for their ratingsdid not differ between majority and minority children,however. The majority of all children gave moral reasons fortheir ratings (M = .83) and less than a quarter of the childrengave social-conventional reasons (M = .13); there were nodifferences for ethnic status or context. There was a significantage-related decline in the use of moral reasons, however, toexplain their ratings for the exclusion between friends at lunch,with tenth grade children (M = .77) giving fewer moral reasonsin contrast to 4th (M = .87) and 7th (M = .87) grade children,

INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2007, 31 (5), 491–500 495

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

F(2,678) = 8.614, p < .0001. As predicted, participants’ratings revealed subtle but significant differences betweenmajority and minority children regarding the wrongfulness ofexclusion when the reasons given for exclusion were not explic-itly about race; minority children rated the reasons as more

wrong. The explicit justifications, however, for evaluations ofthe exclusion decision were the same for both groups. Thus,the form of reasoning does not differ by ethnic status, but thesalience of the considerations in the situation differ for childrenfrom different ethnic backgrounds.

Wrongfulness ratings for race-based peer exclusion ininterracial contexts at homeThe vast majority of all children rated a child’s decision toexclude a friend from a sleepover birthday party due toparents’ discomfort about race as wrong (M = 7.22); therewere no significant differences for grade or ethnic status. Thereasons were moral, that is, that it would be wrong due to prej-udice (M = .87).

Wrongfulness ratings for non race-based exclusion ininterracial contexts at homeOur hypotheses that minority children would view thedecision to not invite a friend of a different race to a sleep-over due to parents’ unfamiliarity with the person as morewrong than would majority children were confirmed. Analysesrevealed a main effect for grade, F(2,675) = 15.33, p < .0001,and for ethnic status, F(1,675) = 8.35, p < .004. The gradeeffect indicated that wrongfulness ratings decreased with age(Ms = 5.6, 5.3, 4.9 for 4th, 7th, and 10th grade, respectively),p < .0001. The ethnic status difference indicated that minority

496 KILLEN ET AL. / EVALUATIONS OF INTERRACIAL PEER ENCOUNTERS

Table 1Wrongfulness of exclusion ratings for three contexts by grade and ethnic group

Context by type of exclusion

Lunch Sleepover Dance

Group Group Group Group N Race Non-race functioning Race Non-race functioning Race Non-race functioning

4th gradeMajority 115 Mean 7.04 5.84 6.51 6.89 4.95 6.03 7.13 6.41 6.28

SD 0.98 1.22 0.96 1.04 1.30 1.23 0.91 1.05 1.18Minority 49 Mean 7.22 6.20 6.61 7.02 5.00 6.35 7.10 6.33 6.57

SD 1.07 1.24 1.20 0.95 1.72 1.39 1.61 1.21 1.21

7th gradeMajority 172 Mean 7.51 5.77 6.16 7.32 4.64 5.84 7.51 6.05 5.89

SD 0.69 1.19 1.08 0.76 1.36 1.34 0.63 0.99 1.13Minority 108 Mean 7.39 5.94 6.21 7.21 4.82 6.10 7.35 6.13 6.23

SD 0.81 0.98 1.20 0.99 1.31 1.06 0.79 1.07 1.31

10th gradeMajority 127 Mean 7.50 5.20 5.76 7.26 4.24 5.38 7.39 5.52 5.37

SD 0.71 1.14 1.14 0.85 1.32 1.30 0.67 1.12 1.24Minority 114 Mean 7.20 5.60 5.97 7.47 4.36 5.68 7.12 5.74 5.91

SD 1.11 1.20 1.18 3.92 1.38 1.21 1.36 1.12 1.22

TotalMajority 412 Mean 7.38 5.62 6.14 7.18 4.60 5.75 7.36 5.99 5.84

SD 0.81 1.21 1.10 0.89 1.36 1.32 0.75 1.10 1.23Minority 270 Mean 7.28 5.84 6.18 7.28 4.77 5.97 7.21 6.00 6.16

SD 0.99 1.15 1.21 2.65 1.42 1.21 1.10 1.14 1.27

Note. N = 685. Race = racial exclusion. Non-race = lack of shared interest (lunch); unfamiliarity (sleepover); rival school (dance). Group func-tioning = lack of fit with the group. 1 = very, very good; 8 = very, very bad.

4.8

5

5.2

5.4

5.6

5.8

6

6.2

6.4

6.6

4th 7th 10th

Grade

Wro

ng

to

exc

lud

e

MajorityMinority

Figure 1. Children’s ratings of whether it is wrong for peers toexclude a minority peer at school for reasons such as lack of sharedinterests or “won’t fit in”.

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

children viewed the non-race based reasons for exclusion asmore wrong than did the majority children as shown in Figure2. Thus, evaluating a child’s decisions not to invite a friendfrom a different racial background to a sleepover partybecause the parents were either unfamiliar with the child, ordid not believe that the child would “fit in” were viewed asmore wrong by minority than majority children at all threeages (for means, see Table 1). Further, a 3 (grade: 4th, 7th,10th) ! 2 (ethnicity: majority, minority) ! 3 (type ofexclusion: race-based, non-race based, group functioning)ANOVA with repeated measures on the last factor wasconducted on participants’ ratings for the three reasons forexclusion indicated that, as expected, children viewed race-based reasons as more wrong than non-race based reasons orgroup functioning reasons for Sleepover: F(2,1348) = 465.19,p < .0001, for type of exclusion. Follow-up analyses indicatedthat all children rated race-based exclusion as more wrongthan the other two reasons (non-race based and group func-tioning), ps < .05, and there were no significant differencesbetween the last two reasons.

Analyses of participants’ reasons for their ratings of achild’s decision not to invite a cross-race friend over for asleepover due to non-race based reasons from parentsrevealed no grade or ethnic status differences. All childrenused moral (M = .35) or social-conventional (M = .61)reasons to explain their rating of the parents’ lack of famil-iarity with a friend as the basis for exclusion. Moral reasonspertained to “empathy” (how the new friend would feel) andsocial-conventional referred to a parent’s need to know thechild better (authority jurisdiction). When asked aboutparents’ decision based on group functioning, that the newchild would not “fit in,” children used moral reasons (M =.75) and social-conventional reasons (M = .21) indicatingthat, for most children, this reason was viewed as unfair. Aswith the peer contexts, the forms of reasons did not vary byethnic status; instead, what varied had to do with the salience(how wrong) of the decision to exclude.

Estimations of the frequency of race-based exclusionA 3 (grade) ! 2 (ethnic status) ! 3 (context: Lunch, Sleep-over, Dance) ANOVA with repeated measures on the lastfactor for estimations of the frequency of race-based exclusionrevealed a significant grade effect, F(2,675) = 4.20, p < .01,and an ethnicity effect, F(2,675) = 3.87, p < .05. As shown inTable 2, with age, participants were more likely to state thatrace-based exclusion occurred. As predicted, minority childrenestimated that race-based exclusion occurred more often thandid majority children. Significant interaction effects emergedfor context and ethnic status, F(2,1350) = 8.18, p < .0001, andfollow-up tests, indicated that minority children estimatedrace-based exclusion occurred more often than did majoritychildren in both the Lunch and the Sleepover contexts, (ps <.05); there were no ethnic status differences for the dancecontext. For age-related changes, 10th grade children judgedthat friendship race-based exclusion and parental race-basedexclusion occurred more often than did 4th and 7th gradechildren (who were not different), ps < .0001). Thus, with age,children estimated a higher level of race-based exclusion andthis was also reported more often by minority than majoritychildren at all ages.

The pattern was reversed for children’ estimations of non-race based exclusion. Majority children estimated that non-race based exclusion occurred more often than did minoritychildren, F(1,676) = 9.55, p < .002. Further, there was anage-related increase, F(2,676) = 14.5, p < .0001, indicatingthat older children viewed exclusion as occurring more likelythan did younger children. A context by grade interaction,F(4,1352) = 12.70, p < .0001, showed that increases in esti-mations of the frequency of exclusion occurred for the Lunchand Sleepover scenarios but not for the Dance scenario.

Thus, minority children estimated that race-based exclusionoccurred more often than did majority children, and majoritychildren estimated that non-race based exclusion occurredmore often than did minority children, supporting our predic-tion that ethnicity status influences interpretations of inter-racial peer situations.

Discussion

The novel findings in this study were that minority childrenevaluated interracial peer exclusion as more wrong than didmajority children, and did so when non-race based reasonswere used by the excluder as the basis for the decision (e.g.,lack of shared interests, unfamiliarity, rival school, lack ofgroup functioning). While all children evaluated race-basedexclusion as more wrong than non-race based exclusion,minority children were more likely to rate non-race-basedexclusion as unacceptable (on a quantitative scale) than weremajority children. Majority and minority children did notdiffer, however, regarding their explicit reasons for whyexclusion was wrong. The majority used moral reasons (unfairtreatment) when evaluating peer exclusion (an exception tothis pattern had to do with children’s evaluations of parentaldecisions to reject a peer as discussed below). Minority andmajority children did not differ regarding their social-cognitivecapacities to articulate and verbalize moral and social-conventional categories or reasons (Turiel, 2002). All childrenused a range of social cognitive reasons when evaluatingexclusion decisions.

INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2007, 31 (5), 491–500 497

4.2

4.4

4.6

4.8

5

5.2

5.4

5.6

5.8W

ron

g t

o e

xclu

de

4th 7th 10th

Grade

MajorityMinority

Figure 2. Children’s ratings of whether it is wrong for peers toexclude a minority peer from a sleepover for reasons such as parentalunfamiliarity.

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

Thus, on an explicit level, there were few differencesbetween majority and minority children; social cognitivereasons for exclusion did not differ by ethnicity or age of theparticipants. On a more implicit level, however, in terms ofinterpreting how wrong it is for a majority peer to exclude aminority peer, contextual factors made a difference. Using arange of intimate contexts (friendship, dating, sleepover),minority children were more likely to view the decision toexclude a minority peer as wrong than were majority children.Similarly, age-related differences regarding reasons for whyexclusion was wrong revealed subtle, but significant findings,with minority participants more likely to use empathy reasonsthan were majority children. These findings differ from priorresults in which race and non-race based reasons for exclusionwere not described separately in the methodology (Killen et al.,2002). In the present study, participants significantly differen-tiated between race-based and non-race based reasons forexclusion, and increasingly so with age. Moreover, this is one ofthe very few studies on racial attitudes to include both minorityand majority children; the vast amount of research continues tofocus on majority children (and adults). Including minoritychildren in this sample revealed aspects of their interpretationsof exclusion decisions not previously documented.

In light of Pettigrew and Tropp’s (2005) recent findings onthe importance of cross-race friendships on the reduction ofprejudice, these findings provide a preliminary basis for under-standing why it is that cross-race friendships decrease with agein development (McGlothlin & Killen, 2005). One reason forthe different interpretations of exclusion in interracial peercontexts might be that minority children are more likely toexpect that majority children hold biases, explicit or implicit,that contribute to their decisions to exclude someone from a

different racial background in a peer situation. Recentresearch, in fact, has demonstrated that this is the case.European-American children at 6 and 9 years of age in homo-geneous schools were more likely to attribute negative inten-tions to minority children than to majority children inambiguous interracial peer situations (McGlothlin & Killen,2006). Even in heterogeneous schools in which race was notused to make attributions of intentions (Margie, Killen, Sinno,& McGlothlin, 2005; McGlothlin et al., 2005), European-American children were less likely to expect that cross-racepeer dyads could be friends than were minority children,suggesting that racial biases about friendships exist as early as9 years.

We theorized that the increasing diversity of school compo-sitions with age (reflective in the schools in the US region thatwas sampled) could lead to inclusive evaluations, and, at thesame time, the increasing salience of adolescent peer groups(Horn, 2003) could lead to exclusive evaluations. Bothperspectives were revealed in adolescents’ evaluations.Adolescents judged interracial exclusion to be wrong, and usedmoral reasons; at the same time, there was a significant age-related decline in the wrongfulness of exclusion whenconsiderations involving group functioning, unfamiliarity, andshared interests were considered.

The most striking context finding pertained to the homecontext in which a parent’s discomfort with a minority peer’svisit to the home for a sleepover party was viewed as legitimateby most majority children, and by half of the minoritychildren, in contrast to the peer contexts in which non-racebased reasons were viewed in a more negative light. Mostparticipants used social conventional reasons such as author-ity jurisdiction to justify the exclusion decision. These findings

498 KILLEN ET AL. / EVALUATIONS OF INTERRACIAL PEER ENCOUNTERS

Table 2Participants’ estimations of the frequency of exclusion for three contexts by grade and ethnic group

Lunch Sleepover Dance

Group N Non-race Race Non-race Race Non-race Race

4th gradeMajority 115 Mean 2.94 2.10 3.02 2.13 2.68 2.43

SD 0.73 0.89 0.93 0.83 0.89 0.83Minority 49 Mean 2.69 2.20 2.86 2.31 2.49 2.37

SD 0.74 1.12 0.91 0.90 0.96 0.95

7th gradeMajority 172 Mean 3.22 2.23 3.25 2.02 2.85 2.35

SD 0.80 0.90 0.78 0.73 0.91 0.90Minority 108 Mean 3.11 2.24 3.29 2.30 2.50 2.31

SD 0.81 0.94 0.86 0.78 0.94 0.92

10th gradeMajority 127 Mean 3.52 2.39 3.35 2.25 2.50 2.33

SD 0.63 0.81 0.68 0.64 0.94 0.79Minority 114 Mean 3.37 2.61 3.34 2.59 2.41 2.29

SD 0.77 1.00 0.81 0.85 0.90 0.90

TotalMajority 413 Mean 3.23 2.25 3.22 2.12 2.69 2.36

SD 0.76 0.84 0.81 0.74 0.93 0.85Minority 270 Mean 3.14 2.39 3.23 2.42 2.46 2.31

SD 0.81 1.02 0.86 0.84 0.92 0.91

Note. N = 685. “How often do you think kids your age might not invite someone tolunch/sleepover/dance for X reasons? 1 = never; 7 = always.

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

may reflect the unique parent–child context in which parentaljurisdiction has a significant influence on children’s judgments(Smetana, 1988). Issues of safety and protection by parents areviewed as legitimately within authority jurisdiction. Even so,majority participants were more likely to judge these reasonsas legitimate than were minority children, who were also morelikely to report that racial exclusion occurs in the home contextthan were majority children.

The findings that all children evaluated race-based exclusionas wrong, and that differences were revealed in the non-racebased exclusion situations, is consistent with social psychologi-cal research with adults which has shown that in straight-forward situations, adults support egalitarian views, and thatstereotypes are activated in situations that are ambiguous orcomplex (Dovidio, Kawakami, & Gaertner, 2002; Gaertner &Dovidio, 1986). Probing children about whether a non-race-based reason for exclusion was legitimate in interracial peercontexts provided a potentially ambiguous situation in thatthere were multiple reasons that could be used for exclusion(race-based and non-race based). The findings were subtle butpervasive and significant, which is what one would expectwhen evaluating a multifaceted situation.

As expected, minority children estimated that race-basedexclusion occurred more often than non-race based exclusion,and this difference was greater for exclusion in the homecontext than in the school context. These analyses providefurther support for social psychological arguments for analyz-ing individuals’ justifications for intergroup behavior (Jost &Banaji, 1994). Jost and colleagues have asserted that adults usestereotypes as a form of justification for maintaining the statusquo. In this study, we did not find that children used explicitstereotypes to justify exclusion. Instead, children relied onsocial-conventional explanations such as nonshared interests,lack of familiarity, or disruptions to the social group function-ing. Social conventional messages often incorporate implicitcultural messages that perpetuate racial stereotypes, however,and further investigation of the multifaceted aspects ofconventional norms needs to be examined (Levy, Chiu, &Hong, 2006). Assessing attributions of non-race based motivesmay allow for further information on when stereotypesunderlie justifications regarding nonshared interests.

Coupled with the findings that minority children estimatedthat exclusion based on race occurs more often than didmajority children these findings indicate that minority childrenmay attribute racial motives to peers’ and parents’ reasons forexclusion even when non-race based motives are given for thebasis for the decision. In this study, exclusion decisions weremade by majority (European-American) individuals regardinga minority (African-American) peer. Most likely, minorityparticipants identified with the excluded child in the scenariosdescribed in this study, who was a member of an ethnicminority group. Individuals’ social identities are derived frommembership in various groups (Brown & Zagefka, 2005) andsocial identification with the in-group is related to judgmentsof bias and prejudice (Bennett & Sani, 2004). A larger samplereflecting diverse school environments could provide infor-mation regarding diversity within the minority population, andpoint to how social and ethnic identity may serve as a signifi-cant mediator for judgments about exclusion. For futurestudies, individual assessments of children’ social and ethnicidentity would shed further light on why it is that majority andminority children’ evaluations of non-race based exclusiondiffered in this study.

For a future study, it would be fruitful to provide both direc-tions of exclusion (majority of minority and vice versa) todisentangle identification with the in-group and ratings ofwrongfulness of exclusion. Given that minority childrenexperience more exclusion than do majority children in theUS, reversing the direction of exclusion raises different issues,such as the ecological validity of the experience (e.g.,European-American children are less likely to experience beingrejected for a sleepover at an African-American household thanthe reverse based on numeric proportions of ethnic diversityand status hierarchies for most US communities). Furtherresearch is required, however, to determine what accounts fordifferences between majority and minority children’ interpret-ations of exclusion decisions.

Recently, Graham (2006) has discussed the importance ofstudying ethnicity in context, and has asserted that it is essen-tial for developmental researchers to examine how childrenfrom different ethnic backgrounds evaluate peer rejection,and how this is reflected in their social experiences in schools.Intergroup contact theory (Dixon et al., 2005; Pettigrew &Tropp, 2005) asserts that individuals’ social experiences withothers from different backgrounds has the potential tocontribute to a reduction in prejudice. Future studies exam-ining the relationship between intergroup contact andchildren’ evaluations of race-based and non-race basedexclusion would further illuminate the findings in the presentstudy.

Another area for future research would be to investigate theintragroup and intergroup dynamics of interracial interactions(Abrams, Rutland, Cameron, & Marques, 2003). Whathappens during interracial interactions that contribute todifferent interpretations of motives and reasons for inclusionor exclusion? What messages are conveyed to children from themajority and minority members of the group about how thegroup functions and what are the goals of the interactions?Prior studies on group dynamics have shown that children aresensitive to in-group as well as out-group norms; most of thesestudies have used nationality or minimal groups to studychildren’s interpretations of deviant norms (Abrams, Rutland,Cameron, & Ferrell, 2007). Applying this model to the contextof racial exclusion would help to reveal the intragroupdynamics that contribute to intergroup exclusion.

This project, which drew on social psychological constructsfor understanding age-related patterns of social cognition andsocial interpretations of situations, reflects a recent burgeon-ing area of research on developmental intergroup attitudeswhich holds promise for understanding the developmentaltrajectories of prejudice, racism, and discrimination (Levy &Killen, in press). By adulthood, stereotypes and biases aredeeply entrenched; close interracial relationships remain theexception rather than the norm (Kennedy, 2003). Understand-ing the subtleties of interracial communication, messages, anddynamics of interactions in childhood provides a basis to facili-tate positive interracial relationships, which has been shown tobe significantly related to positive intergroup attitudes.

References

Aboud, F.E. (2005). The development of prejudice in childhood and adoles-cence. In J.F. Dovidio, P. Glick, & L. Rudman (Eds.), On the nature ofprejudice: Fifty years after Allport (pp. 310–326). Malden, MA: Blackwell.

Aboud, F.E., & Amato, M. (2001). Developmental and socialization influenceson intergroup bias. In R. Brown & S. Gaertner (Eds.), Blackwell handbook of

INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2007, 31 (5), 491–500 499

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from

social psychology: Intergroup relations (pp. 65–85). Oxford, England: BlackwellPublishers.

Aboud, F.E., & Levy, S.R. (2000). Interventions to reduce prejudice anddiscrimination in children and adolescents. In S. Oskamp (Ed.), Reducingprejudice and discrimination (pp. 269–293). Mahwah, NJ: Lawrence ErlbaumAssociates.

Abrams, D., Hogg, M.A., & Marques, J.M. (2005). A social psychological frame-work for understanding social inclusion and exclusion. In D. Abrams et al.(Eds.), The social psychology of inclusion (pp. 1–24). New York: PsychologyPress.

Abrams, D., Rutland, A., Cameron, L., & Ferrell, J. (2007). Older but wilier:In-group accountability and the development of subjective group dynamics.Development Psychology, 43, 134–148.

Abrams, D., Rutland, A., Cameron, L., & Marques, J.M. (2003). The develop-ment of subjective group dynamics: When in-group bias gets specific. BritishJournal of Developmental Psychology, 21, 155–177.

Baron, A.S., & Banaji, M.R. (2006). The development of implicit attitudes:Evidence of race evaluations from ages 6, 10, and adulthood. PsychologicalScience, 17, 53–58.

Baumeister, R., Twenge, J., & Nuss, C. (2002). Effects of social exclusion oncognitive processes: Anticipated aloneness reduces intelligent thought. Journalof Personality & Social Psychology, 83, 817–827.

Bennett, M., & Sani, F. (Eds.) (2004). The development of the social self. New York:Psychology Press.

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

Brown, R., & Zagefka, H. (2005). Ingroup affiliation and prejudice. InJ.F. Dovidio, P. Glick, & L. Rudman (Eds.), On the nature of prejudice: Fiftyyears after Allport (pp. 54–70). New York City: Blackwell Publishers.

Dixon, J., Durrheim, K., & Tredoux, C. (2005). Beyond the optimal contactstrategy. American Psychologist, 60, 697–711.

Dovidio, J.F., Glick, P., & Rudman, L. (Eds.) (2005). Reflecting on the nature ofprejudice: Fifty years after Allport. Malden, MA: Blackwell.

Dovidio, J., Kawakami, K., & Gaertner, S.L. (2002). Implicit and explicit prej-udice and interracial interaction. Journal of Personality and Social Psychology,82(1), 62–68.

DuBois, D.L., & Hirsch, B.J. (1990). School and neighborhood patterns ofblacks and whites in early adolescence. Child Development, 61, 524–536.

Fisher, C.B., Jackson, J.F., & Villarruel, F.A. (1998). The study of African-American and Latin-American children and youth. In W. Damon & R. Lerner(Eds.), Handbook of child psychology (5th ed., Vol. 1: Theoretical models ofhuman development, pp. 1145–1207). New York: Wiley.

Fisher, C.B., Wallace, S.A., & Fenton, R.E. (2000). Discrimination distressduring adolescence. Journal of Youth & Adolescence, 29(6), 679–695.

Gaertner, S.L., & Dovidio, J.F. (1986). The aversive form of racism. InJ.F. Davidio & S.L. Gaertner (Eds.), Prejudice, discrimination, and racism.Orlando, FL: Academic Press.

Gaertner, S.L., & Dovidio, J.F. (2005). Categorization, recategorization andintergroup bias. In J.F. Dovidio, P. Glick, & L. Rudman (Eds.), Reflecting onthe nature of prejudice: Fifty years after Allport (pp. 71–88). Malden, MA: Black-well.

Graham, S. (2006, March). Ethnicity and peer harassment: Implications for studyingthe psychosocial benefits of ethnic diversity in urban schools. Paper presented atthe Society for Research in Adolescence, San Francisco, CA.

Graham, S., & Juvonen, J. (2002). Ethnicity, peer harassment, and adjustmentin middle school: An exploratory study. Journal of Early Adolescence, 22,173–199.

Horn, S. (2003). Adolescents’ reasoning about exclusion from social groups.Developmental Psychology, 39, 11–84.

Jost, J.T., & Banaji, M.R. (1994). The role of stereotyping in system justificationand the production of false consciousness. British Journal of Social Psychology,33, 1–27.

Kennedy, R. (2003). Interracial intimacies: Sex, marriage, identity, and adoptions.New York: Pantheon Books.

Killen, M., Lee-Kim, J., McGlothlin, H., & Stangor, C. (2002). How childrenand adolescents evaluate gender and racial exclusion. Monographs of the Societyfor Research in Child Development, 67(4). Oxford: Blackwell.

Killen, M., Margie, N.G., & Sinno, S. (2006). Morality in the context of inter-group relationships. In M. Killen & J.G. Smetana (Eds.), Handbook of moraldevelopment (pp. 155–183). Mahwah, NJ: Lawrence Erlbaum Associates.

Killen, M., & Stangor, C. (2001). Children’s social reasoning about inclusionand exclusion in gender and race peer group contexts. Child Development, 72,174–186.

Killian, K.D. (2001). Crossing borders: Race, gender, and their intersections ininterracial couples. Journal of Feminist Family Therapy, 13, 1–19.

Lemerise, E., & Arsenio, W.F. (2000). An integrated model of emotion processesand cognition in social information processing. Child Development, 71,107–118.

Levy, S.R., Chiu, C.Y., & Hong, Y.Y. (2006). Lay theories and intergrouprelations. Group Processes and Intergroup Relations, 9, 5–24.

Levy, S.R., & Killen, M. (in press). Intergroup relations: An integrative develop-mental and social psychological perspective. New York: Oxford University Press.

Margie, N.G., Killen, M., Sinno, S., & McGlothlin, H. (2005). Minoritychildren’s intergroup attitudes about peer relationships. British Journal ofDevelopmental Psychology, 23, 251–269.

McGlothlin, H., & Killen, M. (2005). Children’s perceptions of intergroup andintragroup similarity and the role of social experience. Applied DevelopmentalPsychology, 26, 680–698.

McGlothlin, H., & Killen, M. (2006). Intergroup attitudes of EuropeanAmerican children attending ethnically homogeneous schools. Child Develop-ment, 77, 1375–1386.

McGlothlin, H., Killen, M., & Edmonds, C. (2005). European-Americanchildren’s intergroup attitudes about peer relationships. British Journal ofDevelopmental Psychology, 23, 227–249.

Pettigrew, T.F., & Tropp, L.R. (2000). Does intergroup contact reduce preju-dice?: Recent meta-analytic findings. In S. Oskamp (Ed.), Reducing prejudiceand discrimination (pp. 93–114). Mahwah, NJ: Lawrence Erlbaum Associates.

Pettigrew, T.F., & Tropp, L.R. (2005). Allport’s Intergroup Contact Hypothesis:Its history and influence. In J.F. Dovidio, P. Glick, & L. Rudman (Eds.),Reflecting on the nature of prejudice: Fifty years after Allport (pp. 262–277).Malden, MA: Blackwell.

Rubin, K.H., Bukowski, W., & Parker, J. (2006). Peer interactions, relationships,and groups. In W. Damon & R.M.Lerner (Series Eds.), Handbook of childpsychology: Vol. 3: Social, emotional and personality development (6th ed.,pp. 571–645) (N. Eisenberg Vol. Ed.). New York: Wiley.

Rutland, A., Cameron, L., Bennett, L., & Ferrell, J. (2005). Interracial contactand racial constancy: A multi-site study of racial intergroup bias in 3–5 yearold Anglo-British children. Journal of Applied Developmental Psychology, 26,699–713.

Rutland, A., Cameron, L., Milne, A., & McGeorge, P. (2005). Social norms andself-presentation: Children’s implicit and explicit intergroup attitudes. ChildDevelopment, 76, 451–466.

Smetana, J.G. (1988). Adolescents’ and parents’ conceptions of parental author-ity. Child Development, 59, 321–335.

Smetana, J.G. (2006). Social-cognitive domain theory: Consistencies and vari-ations in children’s moral and social judgments. In M. Killen & J.G. Smetana(Eds.), Handbook of moral development (pp. 119–153). Mahwah, NJ: LawrenceErlbaum Associates.

Smith, A., & Schneider, B. (2000). The interethnic friendship of adolescentstudents: A Canadian study. International Journal of Intercultural Relations, 24,247–258.

Swim, J.K., & Stangor, C. (1998). Prejudice: the target’s perspective. San Diego,CA, US: Academic Press, Inc.

Turiel, E. (2002). The culture of morality. Cambridge, England: CambridgeUniversity Press.

Yancey, G. (1998). Who interracially dates: An examination of the character-istics of those who have interracially dated. Journal of Comparative FamilyStudies, 33, 179–190.

500 KILLEN ET AL. / EVALUATIONS OF INTERRACIAL PEER ENCOUNTERS

© 2007 International Society for the Study of Behavioral Development. All rights reserved. Not for commercial use or unauthorized distribution. at Mina Rees Library/CUNY Graduate Center on May 10, 2008 http://jbd.sagepub.comDownloaded from