Social identities and intergroup bias in immigrant and nonimmigrant children.
ABSTRACT Ethnic and American identity, as well as positivity and negativity toward multiple social groups, were assessed in 392 children attending 2nd or 4th grade in various New York City neighborhoods. Children from 5 ethnic groups were recruited, including White and Black Americans, as well as recent immigrants from China, the Dominican Republic, and the former Soviet Union. For ethnic minority children, greater positivity bias (evaluating one's ingroup more positively than outgroups) was predicted by immigrant status and ethnic identity, whereas negativity bias (evaluating outgroups more negatively than one's ingroup) was associated with increased age, immigrant status, and (among 4th graders only) ethnic identity. In addition, a more central American identity was associated with less intergroup bias among ethnic minority children.
- [Show abstract] [Hide abstract]
ABSTRACT: Children's evaluations of classmates with learning difficulties tend to be less positive than their evaluations of classmates without learning difficulties; but it is not clear if these evaluations are associated with age, group norms, and group identification. These associations were examined within the context of inclusive elementary school classrooms. Participants (N = 192) were asked about their attitudes and their peers' attitudes towards children with or without learning difficulties. They were also asked to evaluate fictional target children who gave biased or unbiased opinions about children with or without learning difficulties. Results showed that group identification was predicted by age and group judgment. Participants with high group identification and judgment scores in favour of children without learning difficulties were inclined to believe other children would be similarly biased. However, most participants preferred the target child who gave unbiased opinions. Theoretical and educational implications are discussed.Review of Social Development 02/2012; 21(1). · 1.56 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: This study investigates among ethnic minority adolescents how friendships with ethnic minority and majority group peers are related to their attitudes towards the majority outgroup.International Journal of Intercultural Relations 01/2015; 44:88-99. · 1.14 Impact Factor
Article: Children's social identities[Show abstract] [Hide abstract]
ABSTRACT: This paper provides a brief overview of recent developmental research on themes related to children's social identities. Initially, consideration is given to the capacity for social categorization, following which attention is given to children's developing conceptions of social identities, their identification with social groups, and the consequences of identification upon various phenomena such as ingroup favouritism and well-being. Finally, some personal thoughts on the wider political implications of this research are offered. Copyright © 2011 John Wiley & Sons, Ltd.Infant and Child Development 07/2011; 20(4). · 1.20 Impact Factor
Social Identities and Intergroup Bias in Immigrant and Nonimmigrant
Jennifer H. Pfeifer
University of California, Los Angeles
Diane N. Ruble, Meredith A. Bachman,
Jeannette M. Alvarez, and Jessica A. Cameron
New York University
Andrew J. Fuligni
University of California, Los Angeles
Ethnic and American identity, as well as positivity and negativity toward multiple social groups, were
assessed in 392 children attending 2nd or 4th grade in various New York City neighborhoods. Children
from 5 ethnic groups were recruited, including White and Black Americans, as well as recent immigrants
from China, the Dominican Republic, and the former Soviet Union. For ethnic minority children, greater
positivity bias (evaluating one’s ingroup more positively than outgroups) was predicted by immigrant
status and ethnic identity, whereas negativity bias (evaluating outgroups more negatively than one’s
ingroup) was associated with increased age, immigrant status, and (among 4th graders only) ethnic
identity. In addition, a more central American identity was associated with less intergroup bias among
ethnic minority children.
Keywords: attitudes, development, ethnic identity, immigrants, social identity theory
Two important trends in current research on the development of
intergroup attitudes are the increased emphasis on the potential
role of social identity (Bar-Tal, 1996; Bennett, Sani, Lyons, &
Barrett, 1998; Nesdale, 2004) and the increased focus on improv-
ing methodologies of assessing intergroup bias (Aboud, 2003;
Cameron, Alvarez, Ruble, & Fuligni, 2001). These trends demon-
strate attempts by researchers to better understand the nature of
children’s intergroup attitudes and, in addition, to differentiate
between ingroup favoritism and outgroup derogation. In the cur-
rent study, we aimed to extend previous research on the develop-
ment of intergroup attitudes by examining social identities in
relatively understudied populations of ethnic minority and immi-
grant children, and by using a measure of intergroup attitudes that
allows positivity and negativity toward multiple social groups to
be assessed separately.
There are many ways researchers define and distinguish among
varying aspects of intergroup attitudes; therefore, we used the
following terminology throughout for clarity’s sake. Bias (or in-
tergroup bias) refers here to favoritism toward one’s own group
over other groups. On the other hand, prejudice is often considered
a more exclusive definition that specifically connotes outgroup
derogation, or negativity toward outgroups (see Aboud, 1988;
Fishbein, 2002). In the absence of evidence that outright negativity
toward other groups is exhibited, the more general term of bias or
intergroup bias is used, but when derogation is apparent, the more
specific term of prejudice is used.
Identity Development and Intergroup Bias
With increasing frequency, developmental psychologists rely on
social identity theory (Tajfel & Turner, 1979/2001, 1986) to ex-
plain children’s intergroup bias. This view proposes that identifi-
cation with a social group drives individuals to maintain a sense
that their group is positively distinct from other social groups,
which may be a source of increased self-worth. Indeed, member-
ship in novel groups has been shown to produce intergroup bias in
children, especially when the groups are functional and salient in
the child’s environment (e.g., Bigler, Brown, & Markell, 2001;
Bigler, Jones, & Lobliner, 1997; Yee & Brown, 1992). Although
social identity theory is not explicitly developmental, it implies
that children’s intergroup attitudes are likely to become more
biased with age as ethnic or other social identities become incor-
Jennifer H. Pfeifer, Department of Psychology, University of California,
Los Angeles; Diane N. Ruble, Meredith A. Bachman, Jeannette M. Al-
varez, and Jessica A. Cameron, Department of Psychology, New York
University; Andrew J. Fuligni, Department of Psychology and Center for
Culture and Health, University of California, Los Angeles.
This research was made possible in part by a MacArthur Foundation
grant to Diane N. Ruble and Andrew J. Fuligni, National Institute of
Mental Health Research Grant 37215 to Diane N. Ruble, a National
Science Foundation Graduate Research Fellowship to Jennifer H. Pfeifer,
and Russell Sage Foundation Support for a Working Group on Social
Identity. A portion of this work was presented at the 2005 biennial meeting
of the Society for Research in Child Development, held in Atlanta,
We thank the New York City principals, teachers, and children who
participated in this project. Despite their many other obligations, the
principals and teachers welcomed us to their schools, and the children were
willing to share their personal ideas and feelings with us so that we could
in turn learn from them.
Correspondence concerning this article should be addressed to Jennifer
H. Pfeifer, Department of Psychology, University of California, Los An-
geles, 1285 Franz Hall, Los Angeles, CA 90095-1563. E-mail:
2007, Vol. 43, No. 2, 496–507
Copyright 2007 by the American Psychological Association
porated into the self-concept. That is, as children increasingly
identify with a relevant social group, they become more likely to
make social comparisons that positively differentiate that group
from relevant outgroups.
On the basis of social identity theory, Nesdale (1999, 2004)
developed social identity development theory—a theory that un-
derscores the importance of social identification and social context
in intergroup processes. This theory also makes the distinction
between bias (i.e., a preference for one’s ingroup) and prejudice
(i.e., derogation of outgroups). According to this theory, children
are first unaware of racial or ethnic differences among people.
Later, they move from this undifferentiated phase to an awareness
of racial categories. Following this stage of ethnic awareness,
young children (between 4 and 5 years of age) tend to prefer their
ethnic or racial ingroup, but they only exhibit bias. A transition
from this phase of ethnic preference to one of ethnic prejudice,
which specifically includes negativity toward outgroups, is possi-
ble only when children identify with social groups (between 6 and
7 years of age). Prejudice may also emerge earlier if the social
context precipitates it, such as when relevant ingroups possess
norms supporting the expression of prejudice, or there is the
perception of threat from outgroups.
An outstanding question raised by these theories is to what
extent the developmental literature implies a connection between
identity and bias. The few developmental studies that have exam-
ined whether identification is a necessary condition for biased
intergroup attitudes have yielded inconsistent results. As men-
tioned previously, under certain conditions, assigning children
social group memberships can facilitate intergroup bias (Bigler et
al., 1997, 2001; Yee & Brown, 1992). However, other develop-
mental research has suggested that social identity development
does not always precede intergroup bias (Aboud, 1988). For ex-
ample, children who were either unaware of or unable to report
their own nationality nonetheless made more positive evaluations
of their national ingroup than national outgroups (Bennett et al.,
1998). Such bias demonstrated by children prior to social identi-
fication may represent basic group identification processes,
wherein group awareness and socialization produce familiarity
with evaluative labels for different groups. Regardless of these
findings, this level of inquiry has failed to address social identity
predictions directly because individual variability in identification
was not assessed.
Similarly, those studies specifically examining whether in-
creases in identification are associated with increases in bias have
not yet reached a consensus. Some developmental research has
found a positive correlation between identification with relevant
social groups and intergroup bias (Bennett et al., 1998; Nesdale,
Durkin, Maass, & Griffiths, 2004, 2005; Nesdale, Maass, Durkin,
& Griffiths, 2005; Nesdale, Maass, Griffiths, & Durkin, 2003).
Other work looking at ethnic identity development in adolescents
and young adults has suggested that strong ethnic identities con-
stitute markers of maturity, consideration of intergroup differ-
ences, and are ultimately associated with less intergroup bias
(Berry, 1984; Berry, Kalin, & Taylor, 1977; Phinney, Ferguson, &
The inconclusive findings in both approaches—asking whether
identity precedes or is associated with increases in intergroup bias
during development—may have resulted because each study cap-
tured different aspects of ethnic identity, some of which may not
be sufficient to cause intergroup bias. Some researchers treated
ethnic identity almost like a categorical variable, essentially sep-
arating children by ethnic group membership as if all children felt
similarly about that aspect of their identity (e.g., conducting anal-
yses using categorical ethnicity as an independent variable indi-
cating ethnic identity). In other developmental studies, aspects of
awareness or self-identification of ethnicity—very early funda-
mentals of ethnic identity—were the focus (e.g., whether a child
can accurately report his or her membership in an actual or novel
social group). Very few investigators have measured the pertinent
novel or actual social identities in any detailed manner but, rather,
have assumed their existence and maturation with age. Therefore,
most work has underestimated what is implicated in children’s
identification with their ethnicity or other relevant social groups
(Ruble et al., 2004).
In contrast, current approaches to social identity are often mul-
tidimensional, highlighting different components of identity, in-
cluding centrality, private regard or evaluation, salience, and
knowledge (Ashmore, Deaux, & McLaughlin-Volpe, 2004; Ruble
et al., 2004; Sellers, Smith, Shelton, Rowley, & Chavous, 1998;
Shelton & Sellers, 2000; Waters, 1999). For example, in the case
of ethnic identity, these factors consist of the personal importance
of ethnic group membership (centrality) and personal feelings
about one’s own ethnic group (private regard). Our primary goal
in the present study was to examine some of these components of
identity as they emerge in middle childhood to further clarify the
nature of the connection between identification with ethnic or
other social groups and the development of intergroup bias. We
selected two age groups in middle childhood (second and fourth
grade) because these dimensions of social identity are most likely
to be first available for assessment during this period (Ruble et al.,
Alternative Social Identities and Bias Development
Although there appears to be some evidence that social identi-
ties at least play a role in the development of intergroup bias, prior
research has yet to fully assess a host of specific social identities,
such as those related to being a recent immigrant or a member of
an ethnic minority group. These social identities provide an op-
portunity to consider groups that are likely to experience variations
in context that may be quite important to the formation of bias.
Nesdale and colleagues (Nesdale & Flesser, 2001; Nesdale, Dur-
kin, et al., 2005; Nesdale, Maass, et al., 2005) have demonstrated
that when children belong to novel social groups in which preju-
dice expression is the norm, they are more likely to express
prejudice themselves. Furthermore, they have shown that this
effect is enhanced when the outgroup constitutes a threat to the
ingroup (Nesdale, Durkin, et al., 2005; a phenomenon also known
as realistic group conflict; see Sherif, 1966). Some additional
evidence suggesting that these factors are likely to precipitate
prejudice development includes Bar-Tal’s (1996) research exam-
ining the explicitly negative stereotypes of Arabs held by very
young (3-year-old) Jewish Israeli children toward Arabs (see also
Bar-Tal & Labin, 2001).
There is good reason to expect that ethnic minority and immi-
grant children may differ from each other and majority White
children in some, if not all, of these factors. First, research has
suggested that ethnic identity may be stronger or more salient for
SOCIAL IDENTITIES AND INTERGROUP BIAS
ethnic minority or immigrant children and adolescents compared
with White children and adolescents (Phinney, 1989, 1992; Phin-
ney & Alipuria, 1990; Phinney, Cantu, & Kurtz, 1997; Rumbaut,
1994); in addition, the strength of ethnic and national (i.e., Amer-
ican) identity has been shown to vary among adolescents from
different ethnic or immigrant groups (Phinney & Devich-Navarro,
1997; Phinney, Horenczyk, Liebkind, & Vedder, 2001). Second,
there is ethnographic and anecdotal evidence of norms in some
ethnic minority children’s families that supports the expression of
prejudice toward outgroups, such as the tendency of West Indian
or other immigrant families to make downward social comparisons
with Black Americans (Waters, 1999). Third, perceptions of threat
(including discrimination) are not uncommon among these groups,
and visible minorities might experience different levels of per-
ceived threat than nonvisible minorities (Dion & Kawakami,
Although group norms and perceptions of threat were not as-
sessed directly, we aimed to examine alternative social contexts
likely to be associated with variability in ethnic identity and
intergroup bias. However, we also considered the possibility that
White children may exhibit more intergroup bias than ethnic
minority and immigrant children. This perspective was based on
the extensive body of research documenting White children’s
biases against Black people, primarily using trait stereotyping
Trait stereotyping measures are ones in which children assign
positive or negative nonsynonymous traits to individuals repre-
senting ingroups and outgroups, for example, the Preschool Racial
Attitude Measure II (Williams, Best, Boswell, Mattson, & Graves,
1975) or the Multiresponse Racial Attitude measure (MRA; Doyle
& Aboud, 1995). Despite their names, the measures typically
assess global evaluative associations between different groups and
positive or negative affect and, to a lesser extent, cognitive com-
ponents of prejudice (such as the application of particular traits in
a stereotyped fashion to ingroups and outgroups). The trajectory of
White children’s attitudes based on these measures is thought to
consist of ingroup favoritism emerging as early as preschool and
then declining in middle childhood (after 8 years); furthermore,
these assessments are frequently correlated with cognitive ability
and socialization factors (for reviews, see Aboud, 1988; Fishbein,
2002; Williams & Morland, 1976). Much less is known about trait
stereotyping during adolescence or in children from ethnic minor-
National (Superordinate Group) Identity and Reductions
The social contexts surrounding studying children with immi-
grant and ethnic minority status also provide an opportunity to
examine alternative theories about the relationship between social
identities and intergroup bias. Previous research has suggested that
possessing a superordinate group identity can reduce biases at the
subgroup level (Gaertner & Dovidio, 2000; Gaertner, Dovidio,
Anastasio, Bachman, & Rust, 1993; Hornsey & Hogg, 2000). One
intervention found that enhancing a novel superordinate identity
reduced some forms of subgroup bias in children (Houlette et al.,
2004), and another study found that adolescents who possessed
more inclusive representations of their high school’s student body
had less intergroup bias (Gaertner, Rust, Dovidio, Bachman, &
Anastasio, 1994). However, this model has not been explicitly
tested in middle childhood using superordinate identities that are
part of children’s everyday lives.
Studies that have examined national (i.e., American) identity in
various ethnic majority or minority and immigrant groups have
shown that there is variability in the degree to which ethnic
minority and immigrant individuals feel “American” and report
being perceived as an “American” by White Americans (Barlow,
Taylor, & Lambert, 2000; Devos & Banaji, 2005; Waters, 1999).
Given the factors described above that may contribute to enhanced
intergroup bias in ethnic minority children and immigrant children,
we also aimed to examine whether individual differences in na-
tional (American) identity were negatively associated with ethnic
To summarize thus far, we had three primary goals in the current
study. The first goal was to investigate whether individual differ-
ences in ethnic identity during development were associated with
intergroup bias. The second goal was to focus specifically on this
relationship across groups of children who vary in immigrant and
ethnic minority status, as these children may find themselves in
social contexts supporting the development of intergroup bias.
Finally, the third goal was to assess whether a superordinate group
affiliation, specifically national (American) identity, would reduce
children’s intergroup bias. To accomplish these goals, we needed
to separately measure children’s attitudes toward multiple social
groups, in addition to aspects of their ethnic and national identity.
Many measures of childhood bias confound attitudes toward
ingroup(s) with attitudes toward outgroup(s) by assessing them
concurrently (Aboud, 2003; Cameron et al., 2001; Cross, 1991;
Lerner & Schroeder, 1975). Effectively separating ingroup favor-
itism and outgroup derogation, however, may be critical to a social
identity approach. Whereas there has been some debate about
whether social identity theory addresses ingroup favoritism only or
is useful for understanding outgroup derogation as well, social
identity development theory (Nesdale, 1999, 2004) makes specific
predictions about outgroup derogation and, thus, requires its sep-
We thus aimed to identify existing developmental trends in
attitudes toward multiple social groups at a stage during which
relevant social identities are developing. In accordance with pre-
viously reported developmental trends, we predicted that ingroup
favoritism would remain stable or increase with age but not de-
crease. However, if observed at all, we predicted that outgroup
derogation would occur only in children whose immediate social
contexts supported such a development. More important, we were
interested in the connection between individual variability in eth-
nic identity and intergroup bias. We predicted that children having
greater centrality and positive regard with respect to their ethnic
identity would demonstrate more intergroup bias. However, we
expected this relationship to apply more to older children, as
younger children may be less likely to have developed basic
aspects of ethnic identity.
Furthermore, we examined these developmental trends in inter-
group bias and ethnic identity in children who were first- or
second-generation Americans or members of an ethnic minority
group. We thought it was possible that recent immigrants and
PFEIFER ET AL.
ethnic minority children might show greater bias toward other
immigrant groups as a result of downward comparison processes
and other social contextual factors. However, it was also possible
that White American children might show greater intergroup bias,
particularly toward non-White immigrant groups, because they
may be socialized primarily against exhibiting bias toward Black
Americans. Finally, we predicted that children with relatively
more central national (American) identities would exhibit less
Given the goals of our study, it was important to include
children from a variety of ethnic and immigrant backgrounds. In
particular, we targeted first- or second-generation immigrant chil-
dren from the top three immigrant groups in New York City during
the 1990s: Chinese, Dominican, and Russian1(New York City
Department of City Planning, 1996). These groups were especially
beneficial to include, given their ethnic diversity and varying
experiences in the United States. For example, Russian immigrants
are not a racial minority, in contrast to Chinese and Dominican
immigrants. We also targeted third-generation or later White and
Black American children because these groups are most often
included in studies focused on the development of intergroup bias
and prejudice. To recruit children from these groups, we conducted
the study in several neighborhoods that had large populations of
Chinese, Dominican, Russian, or Black American families. Unlike
the other neighborhoods, Russian neighborhoods were much more
heterogeneous, and it is in these neighborhoods that we also
recruited our White sample. Consequently, Chinese, Dominican,
and Black children attended schools that were relatively ethnically
homogeneous, whereas Russian and White children attended more
We interviewed 600 children for the study; however, 164 did not
meet our ethnicity criteria, and an additional 8 did not meet our age
criteria, leaving 428 potential participants. Because of missing data
on key measures from 36 children, the final sample consisted of
392 children in the second and fourth grades. As shown in Table
1, the sample consisted of similar numbers of Chinese, Dominican,
Russian, and White American children across both the second and
fourth grades. Our Black American sample was smaller, primarily
because of time constraints imposed by these schools with regard
to interviewing. The sample tended to be equally divided by
gender (girls, n ? 221; boys, n ? 171) and grade level (second
grade, n ? 177; fourth grade, n ? 215). Schools in the sample
served children from lower to lower middle-class backgrounds.
The percentage of the population receiving free or reduced-cost
lunches at the schools ranged from 21% to 99%, with the students
from Black, Chinese, and Dominican American backgrounds be-
ing more likely to attend the schools with high percentages of free
The study consisted of three separate sessions that lasted ap-
proximately 40 min each, with an average of 2 weeks separating
each session (in part to reduce possible effects of repeated expo-
sure to similar measures). Each child was interviewed by a female
interviewer of the same racial background. It has been suggested
that, although disclosure to same- and different-ethnicity inter-
viewers may be equivalent when the topic is nonsensitive (i.e.,
topics of low social relevance), disclosure is higher to same-
ethnicity interviewers about sensitive topics (e.g., Weeks &
Moore, 1981). For these reasons, and for the children’s general
comfort, same-ethnicity individuals interviewed all children.
Because this study was part of a large study on social develop-
ment and academic engagement, children received a number of
measures in a set order per session. Of the measures presented,
only a subset was relevant to this particular study: (a) a group
attitudes measure and (b) an ethnic identity measure. In addition to
these measures, during the first interview, each child’s ethnic
background was carefully assessed to determine whether they fit
into one of our target groups. If they did not, they completed only
a subset of measures and were not included in the present study.
Group attitudes measure.
In the group attitudes measure, chil-
dren were asked to indicate how many people from four different
social groups (American, Black, Chinese, and Spanish) possessed
each of the following attributes: rich, bad, friendly, ugly, smart,
selfish, good at sports, lazy, honest, and shy. These characteristics
are a subset of those used in common trait stereotyping measures
(e.g., the MRA; Doyle & Aboud, 1995). The social groups used in
this measure were selected so that children could evaluate several
outgroups and an ingroup that almost all children identified with
(i.e., American). For the Chinese, Dominican, and Black American
children, this measure also allowed them to evaluate their own
ethnic group. Children were presented with a 5-point scale that
consisted of the following response items: none, few, some, most,
and almost all, which were presented both verbally and pictorially.
Children received these attitude measures in different sessions.
Children were asked to evaluate “Black” people in the first session
and “American” people as a whole in the second session. In the
third session, children evaluated “Chinese” and “Spanish” or “Do-
minican”2people, with several measures separating these attitude
1For simplicity’s sake, we use the labels Chinese and Russian. How-
ever, it is important to note that children from Taiwan and Hong Kong were
included in the Chinese sample, and children from countries that were part
of the former Soviet Union (e.g., the Ukraine) were included in the Russian
2It should be noted that this scale used social group labels that children
would understand. Therefore, the label Chinese was used in reference to
Asian Americans, and the label Spanish was used to refer to Latino
Americans. However, because of our interest in ingroup evaluations, Do-
minican children were specifically asked about Dominican people.
Number of Children in Each Group
Total2nd grade4th grade2nd grade 4th grade
SOCIAL IDENTITIES AND INTERGROUP BIAS
Eight intergroup attitude subscales were created by averaging
the responses to each of the four groups separately for the five
positive (rich, friendly, smart, good at sports, and honest) and the
five negative (bad, ugly, selfish, lazy, and shy) items. These eight
subscales demonstrated moderate internal consistency. The “shy”
item in the negative scale and the “rich” item in the positive scale
did not fit with the rest of the items, so they were removed. The
alphas ranged from .61 to .74, with a mean of .69, as calculated on
the full sample.
Ethnic identity measure.
The ethnic identity measure was de-
veloped specifically for this study, in a format similar to Harter’s
(1985) Self-Perception Profile for Children. It was presented in
Session 2 and included six items intended to tap ethnic centrality
(Items 2, 5, and 6 from the list that follows) and private regard
(Items 1, 3, and 4). These items asked children to report whether
they are happy to be [ethnicity], believe being [ethnicity] is an
important part of self, are proud to be [ethnicity], like being
[ethnicity], see self as more [ethnicity] than American, and believe
being [ethnicity] is a big part of who one is. The dimensions of
centrality and private regard came directly from the social identity
literature, and specific items were based on those used with adults
but revised to be suitable for young children (Ashmore et al.,
Children were asked about their own ethnicity (Chinese, Do-
minican, Russian, Black, and White), which we categorized as
their “basic” ethnicity. (It is important to note that, prior to this
measure, we established that all children categorized themselves as
members of their basic ethnic group.) In this measure, participants
were presented with two groups of children; for example, children
in the first group felt that their ethnic identity was important (or
felt good about their ethnic identity), and children in the second
group felt that their ethnic identity was not important (or did not
feel good about their ethnic identity). Participants were asked to
indicate which group of children they were most like, and then
specify if that was really true for them or just sort of true for them.
Like the group attitude measures, this scale demonstrated moderate
internal consistency. One item in the scale fit particularly poorly:
whether the child saw him- or herself as more ethnic than Amer-
ican. After removing this item, the alpha improved (? ? .65, as
calculated on the full sample).
However, because of the intragroup variability of this item, we
used it in one analysis that examined the link between national
(American) identity and group evaluations. This item was used to
conduct a split that allowed us to compare average intergroup
attitudes between ethnic minority children who reported feeling
more American than ethnic (n ? 135) and children who felt more
ethnic than American (n ? 107).
To clarify the results and discussion sections, we use the fol-
lowing terms throughout: positivity bias, negativity bias, and in-
tergroup bias. Positivity bias refers to the extent to which children
reported that positive traits applied to more of their own ethnic or
national group members than to outgroup members. Negativity
bias refers to the extent to which children reported that negative
traits applied to fewer of their own ethnic or national group
members than outgroup members; it was used in this case because
there was no strong evidence that the bias was so severe as to
constitute outgroup derogation or prejudice (i.e., means on nega-
tive trait items were far below the midpoint of the scale, indicating
that children applied negative traits to few people in any social
group). The subscales used to measure negativity bias toward
different social groups were recoded such that higher values indi-
cated less negativity toward a given group to mirror the direction
of the positive subscales. As a result, positivity bias scales and
recoded negativity bias scales were positively correlated. Finally,
intergroup bias refers to the sum of reported positivity bias and
Patterns of Intergroup Attitudes
The initial step was to examine differences in attitudes toward
multiple social groups on the basis of age, gender, or ethnic group
membership for evidence of intergroup bias. We predicted that
children would generally be more positive and less negative to-
ward their own ethnic or national group members than outgroup
members. A repeated measures analysis of variance (ANOVA)
was conducted with three between-subjects factors—ethnicity
(Chinese, Dominican, Russian, White, and Black), gender (male
and female), and grade level (second and fourth)—and two within-
subject factors—target group (American, Black, Chinese, and
Spanish/Dominican3) and valence of attribute (positive and nega-
tive [recoded] subscales) on the intergroup attitude subscales. We
were specifically looking for evidence of intergroup bias, which
would be indicated by a two-way interaction between target group
and child’s ethnicity. To the extent that intergroup biases varied by
the valence of attributes children were ascribing, we expected to
find a three-way interaction between target group, ethnicity, and
valence. When sphericity assumptions were not met, we applied
the Greenhouse–Geisser correction. Post hoc multiple comparison
procedures used Bonferroni correction to assess the significance of
differences between pairs of cell means.
As predicted, there was a significant interaction between target
group and child’s ethnicity, F(11, 333) ? 6.9, p ? .001, ?p
The means in Table 2 suggest that children from some ethnic
groups demonstrated more intergroup bias than others. Specifi-
cally, post hoc multiple comparisons indicated that immigrant
(Chinese, Dominican, and Russian) children demonstrated some
degree of intergroup bias, whereas nonimmigrant (White and
Black) children did not. Immigrant children were significantly
more positive and less negative toward their own ethnic or national
group members than toward Black people (see Table 2). Because
there was a relatively smaller sample of Black children, an absence
of significant differences in attitudes toward target groups might
also have been due to a larger standard error in this subsample.
This two-way interaction qualified a main effect of target group,
F(3, 333) ? 7.4, p ? .001, ?p
were significantly more positive and less negative toward Amer-
icans and Spanish/Dominican people than toward Black people,
whereas there were no significant differences in attitudes toward
Chinese people relative to other target groups (see Table 2).
2? .02, that suggested that children
3To reiterate we used the label Spanish/Dominican in this section to
refer to the target group. Although the label Dominican was used with
Dominican children, the label Spanish was used with all other children to
ensure they understood to which target group we were referring (Latino
PFEIFER ET AL.
These general patterns of intergroup bias were qualified by a
three-way interaction between target group, valence of attribute,
and child’s ethnicity, F(11, 349) ? 1.9, p ? .05, ?p
hoc multiple comparisons indicated that each ethnic group had a
slightly different pattern of positivity bias (Figure 1) and negativity
bias (Figure 2; see also Table 2). Although both White and Black
children showed no indication of intergroup bias at the more
general level, some differences emerged when distinguishing be-
tween positivity and negativity bias. Specifically, White children
were more positive toward American people than toward Black,
Chinese, or Spanish/Dominican people, but they exhibited no
significant differences in negativity toward any group. However,
Black children still did not exhibit any significant differences in
positivity or negativity toward any group; again, this may have
been due to lower power in this group.
These higher order interactions qualified a two-way interaction
between target group and valence, F(3, 349) ? 21.4, p ? .001, ?p
? .05. The means in Table 2 suggest that children were least
positive toward Chinese people and most negative toward Black
people. This general pattern may indicate that the traits captured
both evaluative and stereotypic biases to some extent, and that
children felt positive traits were least stereotypical of Chinese
people and negative traits were most stereotypical of Black people.
Several additional main effects and lower order interactions
were found, which were unexpected or irrelevant to the primary
goals of the study. These included main effects of valence, F(1,
372) ? 86.4, p ? .001, ?p
11.4, p ? .001, ?p
and child’s ethnicity, F(4, 372) ? 4.3, p ? .01, ?p
as an uninterpretable four-way interaction between target group,
valence of attribute, child’s ethnicity, and grade level, F(11,
349) ? 2.3, p ? .01, ?p
that children were more positive than negative toward groups
(Ms ? 3.8 and 3.5, for positivity and negativity, respectively),
indicating, as expected, the presence of a well-documented bias
toward excessively positive self- or social evaluations (Schuster,
Ruble, & Weinert, 1998). Similarly, the main effect of grade level
indicated that older children were less positive and more negative
toward target groups overall, in comparison to younger children
(Ms ? 3.8 and 3.6, for second and fourth graders, respectively), as
would be predicted from previous research on the decline of this
tendency toward overly positive evaluations (Schuster et al.,
1998). The two-way interaction between valence and child’s eth-
nicity suggested that Dominican participants tended to use the
endpoints of the scales more frequently. They were more positive
toward groups overall than other children except for White partic-
ipants, and they were also more negative toward groups overall
than other children except for Black participants.
2? .02). Post
2? .19, and grade level, F(1, 372) ?
2? .03, a two-way interaction between valence
2? .04, as well
2? .02. The main effect of valence showed
Relation Between Ethnic Identity and Intergroup Attitudes
Although the repeated measures ANOVA indicated some evi-
dence of intergroup bias, our primary goal was to examine the
relationship between ethnic identity and intergroup bias. For the-
oretical reasons (which were also supported by the significant
three-way interaction between target group, valence of attribute,
and child’s ethnicity), we created new variables to separately
capture the degree of positivity bias and negativity bias for each
child. That is, ascription of positive traits to ethnic outgroups was
Table 2 Positivity and Negativity Toward Multiple Social Groups as a Function of Children’s Ethnic Group Membership
Potential scores on both scales ranged from 1 to 5. Negativity scales have been reverse coded such that higher scores indicate less negativity toward a group.
aAmerican was used to refer to American people as a whole.bAlthough the label Dominican was used with Dominican children, the label Spanish was used with all other children to ensure they
understood to which target group we were referring (Latino Americans).
SOCIAL IDENTITIES AND INTERGROUP BIAS
averaged and subtracted from positivity toward children’s ethnic
ingroup. Similarly, ascription of negative traits to ethnic outgroups
was averaged and subtracted from negativity toward children’s
ethnic ingroup. This latter calculation also used the recoded neg-
ative subscales, so that, in both cases, a greater degree of inter-
group bias was indicated by higher scores on these new variables
of positivity bias and negativity bias. Because nonminority (Rus-
sian and White American) children ascribed traits only to their
national ingroup and ethnic outgroups, not their ethnic ingroup,
this subset of children was excluded from this particular analysis.
We used multiple regression analysis to examine whether pos-
sessing a more central and positive ethnic identity, in addition to
other predictor variables, was associated with children’s positivity
bias and negativity bias. To examine how alternative social iden-
tities influenced the development of intergroup biases, we included
immigrant status as a predictor variable (all children in this anal-
ysis were members of an ethnic minority group). Gender was also
included as a predictor because research with adults has suggested
that men tend to display more prejudice than women (e.g., Sida-
nius & Pratto, 1999). Finally, because we hypothesized that ethnic
identity may be more predictive of intergroup attitudes in older
children as compared with younger children, we included variables
accounting for grade level and the interaction between ethnic
identity and grade level. The interaction term was entered in a
second block, after entering the main effects of ethnic identity
(group mean centered), immigrant status, grade level, and gender.
Two separate regressions were conducted, one for each dependent
variable of positivity bias and negativity bias.
Results showed that ethnic identity and immigrant status each
accounted for significant and unique proportions of variance in
ethnic minority children’s positivity bias, and the overall model
was significant, F(4, 285) ? 5.2, R2? .07, p ? .001. The
regression coefficients (see Table 3) illustrated that possessing a
more positive and central ethnic identity was associated with
greater positivity bias, and children with recent immigrant status
demonstrated greater positivity bias than nonimmigrant children.
Grade level and gender, as well as the interaction between ethnic
identity and grade level, were not significant predictors of posi-
tivity bias. Therefore, the addition of this interaction term did not
significantly improve the model, F(1, 284) ? .82, R2? .003, ns.
Results also showed that ethnic identity, immigrant status, gen-
der, and grade level accounted for significant and unique propor-
tions of variance in minority children’s negativity bias, and the
overall model was significant, F(4, 284) ? 6.3, R2? .08, p ?
.001. Furthermore, with the addition of the interaction term, the
model accounted for a significantly greater proportion of variance,
F(1, 283) ? 9.0, R2? .03, p ? .003, and the overall model was
significant, F(5, 283) ? 6.9, R2? .11, p ? .001. However, with
the addition of the interaction term between grade level and ethnic
identity, the main effect of ethnic identity was no longer a signif-
icant and unique predictor of outgroup negativity. The regression
coefficients (see Table 3) illustrated that children with stronger
ethnic identities and those with recent immigrant status demon-
strated greater negativity bias. In addition, girls exhibited less
negativity bias than boys, and fourth graders showed less negativ-
ity bias than second graders.
Regressions conducted separately for second and fourth graders
demonstrated that ethnic identity was a significant predictor of
negativity bias for fourth graders but not for second graders (see
Table 4). For second graders, immigrant status and gender were
both significant and unique predictors of negativity bias, and the
model was significant, F(3, 132) ? 6.3, R2? .13, p ? .001. For
fourth graders, only ethnic identity was a significant predictor of
negativity bias, and the model was significant, F(3, 149) ? 5.4,
R2? .10, p ? .001. These findings were consistent with our
developmental predictions regarding the relationship between eth-
marked bar and the other bars with which it is grouped.
Degree of positivity by child’s ethnicity. Asterisks indicate significant differences between the
PFEIFER ET AL.
nic identity and negativity bias, namely that associations between
ethnic identity and negativity bias may present themselves only in
Relation Between National Identity and Intergroup
Although the regression analyses indicated that both ethnic
minority children with stronger ethnic identities and immigrant
children evidenced greater intergroup bias, our third goal was to
examine whether identification with a superordinate group (i.e.,
nationality [American]) was associated with less intergroup bias,
as suggested by previous research. Specifically, we predicted that
children who felt they were more American than ethnic would
demonstrate lower levels of bias. Again, because Russian children
did not evaluate their ethnic ingroup, they were excluded from this
analysis, along with White children. Our hypothesis was sup-
ported, as minority children who reported feeling more Amer-
ican than ethnic demonstrated significantly less intergroup bias
than children who felt more ethnic than American, t(1, 240) ?
2.1, p ? .05, Ms ? .36 and .20. This pattern was confirmed
separately in each of the Chinese, Dominican, and Black Amer-
ican participant groups, which indicated that the effect was not
driven by the Black children feeling more American than the
immigrant children. For further confirmation, we broke down
intergroup bias into its separate components—positivity bias
and negativity bias—and examined the patterns of means for
each. Importantly, this indicated that children who felt more
American than ethnic simply did not feel less positive and more
negative toward their own ethnic group, which would decrease
the difference scores between ingroup and outgroup evalua-
tions, but that they also felt more positive and less negative
toward ethnic outgroups.
The overarching purpose of this study was to address whether
individual differences in social identities during middle childhood
between the marked bar and the other bars with which it is grouped.
Degree of negativity (recoded) by child’s ethnicity. Asterisks indicate significant differences
Regression Coefficients for Positivity and Negativity Bias
Positivity biasNegativity bias
B SEB SE
Ethnic Identity ? Grade
*p ? .05.
Dashes indicate data not reported due to nonsignificance.
**p ? .01.
Regression Coefficients for Negativity Bias as a Function of
2nd grade4th grade
*p ? .05.
Dashes indicate data not reported due to nonsignificance.
***p ? .001.
SOCIAL IDENTITIES AND INTERGROUP BIAS
are significantly associated with variability in children’s inter-
group bias. Our primary findings suggest this was indeed the case
across a number of social identities. First, we found that ethnic
identity was associated with greater positivity bias in both younger
and older ethnic minority children. However, only among older
ethnic minority children was having a stronger and more central
ethnic identity associated with greater negativity bias. Second, we
demonstrated that immigrant status was associated with greater
positivity and negativity bias. Finally, we observed that national
identity was associated with less intergroup bias among ethnic
minority and immigrant children.
Positive Associations Between Ethnic Identity and
The discovery of a positive association in middle childhood
between ethnic identity and intergroup bias provides strong sup-
port for social identity theory, thereby supporting its applicability
to developmental populations. Importantly, this was demonstrated
in a context outside that of novel social groups, from which the
majority of evidence for both social identity and social identity
development theories has been obtained. The association between
positivity bias and ethnic identity in ethnic minority children from
both age groups may represent the construct of ethnic preference
specified by social identity development theory (Nesdale, 1999,
2004). During this phase, children are aware of their ethnic group
membership and are concerned primarily with continuing mem-
bership in this group and positive distinctiveness from outgroups
(see also Cameron et al., 2001). On the other hand, the increased
level of negativity bias seen in fourth graders with more positive
and central ethnic identities may mark the beginnings of ethnic
prejudice. This is noted as a possibility because, all else being
equal, fourth graders tended to display less negativity bias (not to
be confused with the greater overall use of negative traits exhibited
by fourth graders in the repeated measures ANOVA). The signif-
icant interaction between grade level and ethnic identity also
highlights the increasingly important role of ethnic identity in
older children. In contrast, the basic group membership variables
that predicted young children’s level of negativity bias (i.e., im-
migrant status and gender) were not predictive for older children,
suggesting that relatively simple categorization processes may be
informing children’s judgments before a more complex sense of
self has developed.
Further research should continue to explore this association
between increases in ethnic identity strength and intergroup bias in
middle childhood and early adolescence to confirm this finding, as
well as determine whether it is correlated with more positive or
detrimental outcomes. This is important because these results,
along with social identity and social identity development theories,
seem to contradict multicultural theories based on research con-
ducted with older adolescents and young adults. A multicultural-
ism perspective suggests that achieving a strong ethnic identity
reduces intergroup biases and enhances well-being (Berry, 1984;
Berry et al., 1977; Phinney, Ferguson, & Tate, 1997). One possible
way to integrate these disparate sets of findings is to propose that,
at the earliest stages of ethnic identity development, intergroup
bias may be expected to increase with ethnic identity strength
because of classic group identity effects. Only after going through
all the stages is a more complex understanding of ethnic identity
achieved, which then results in its association with less intergroup
bias. Alternatively, different aspects of ethnic identity strength
may be relevant during different developmental periods (for ex-
ample, adolescents may emphasize group belonging and affiliation
more than children).
Another potential integration of our results with these varying
theoretical perspectives may be achieved given that the nature of
the intergroup bias we observed appears to be primarily ingroup
favoritism, rather than outgroup derogation. Although we created
separate subscales for positive and negative items, the means for
negative subscales were substantially lower than average (prior to
recoding), indicating that children tended to ascribe negative traits
to only a small proportion of each group. Because of the consensus
among immigrant children in reporting the most negativity and
least positivity toward Black people, it seemed possible that Black
people as social targets were being derogated. However, these
children may have known significantly more about Black people
than the other target groups they were evaluating (or formed
evaluations of them earlier), or there may have been an order
effect, because Black people were evaluated in the first interview
session of the study. Thus, in this sample and pertaining to these
social targets, both positive and negative subscales appeared to be
indicators of varying degrees of favoritism, rather than outright
prejudice. These results provide further evidence supporting the
account that at this age and using explicit measures of trait ste-
reotyping, outgroup derogation is exceedingly rare and may be
seen only in extreme social structural conditions (Bar-Tal, 1996;
Cameron et al., 2001). As a result, we chose to refer to children’s
attitudes as positivity bias and negativity bias to indicate their
difference from outgroup derogation. The results of this study also
demonstrate that positivity bias and negativity bias have both
common and unique predictors, thus suggesting that they should
continue to be assessed independently of each other.
Positive Associations Between Immigrant Status and
The second goal of this study was to examine intergroup bias
among not only ethnic minority children but among those from
immigrant families. Our results demonstrate that immigrant status
was a unique predictor of both positivity bias and negativity bias,
controlling for other significant factors such as grade level, gender,
and ethnic identity. Immigrant status may indicate greater expo-
sure to, or acceptance of, familial or cultural norms supporting the
explicit expression of bias against outgroups (Waters, 1999). Fur-
thermore, recent immigrant families may face greater threats to the
ingroup, including competing for resources, maintaining model
minority status, and perceiving discrimination from outgroups
(Dion & Kawakami, 1996; Waters, 1999). Because the immigrant
ethnic minority (Chinese and Dominican) children lived in rela-
tively homogeneous neighborhoods and attended relatively segre-
gated schools, one might expect these children to be least likely to
directly experience or perceive such threats from outgroups. Per-
haps perceptions of threat and real intergroup conflict primarily
affect parents or other socializing agents, who subsequently com-
municate these threats to children or model intergroup biases to
children that were derived from their own intergroup experiences.
However, these children are likely to be exposed to the potential
for discrimination even if they attend homogeneous schools. Other
PFEIFER ET AL.
factors, such as linguistic differences, may highlight their status as
“outsiders” in this country (Phinney, Romero, Nava, & Huang,
2001). The absence of data detailing the social contexts experi-
enced by these immigrant children remains a limitation of our
study and an important area for future research.
Downward social comparisons may thus result from immigrant
children’s efforts to enhance self-esteem. However, the basic anal-
ysis of intergroup attitudes suggests that all three groups of immi-
grant (Chinese, Dominican, and Russian) children may target
Black people for this process of self-enhancement via intergroup
bias, given that they were significantly less positive and more
negative toward this group than any other. This particular pattern
of downward social comparison between immigrants and Black
Americans has been documented previously, lending some cre-
dence to this idea (Waters, 1999). In particular, Dominicans and
other immigrants from the West Indies are often mistaken for
being Black. This results in efforts to differentiate themselves from
Black Americans (including, but not limited to, downward social
comparison) because of perceptions that this group is often nega-
tively evaluated by society. As mentioned above, however, alter-
native explanations are that children evaluated Black people first
and may have simply known more about Black people than other
target outgroups. This may have been especially true in the case of
Chinese people, given that children seemed to feel that the traits
(both positive and negative) applied least well to that target group.
Unfortunately, we were limited in our ability to compare the
intergroup attitudes of immigrant and nonimmigrant children. The
lack of bias demonstrated by White children directly contrasts
previous research in which White children report bias against
Black people (Aboud, 1988; Fishbein, 2002; Williams & Morland,
1976), but this may have resulted because we assessed White
children’s attitudes toward a national ingroup only, not an ethnic or
racial ingroup. Because we did not explicitly assess White or
Russian children’s attitudes toward White or Russian people, re-
spectively, we were unable to ascertain whether White children
with stronger ethnic identities were more biased, or whether being
White made Russian children less biased. In addition, our conclu-
sions about Black children’s relative lack of bias may be due to a
lack of power to detect significant differences resulting from a
smaller sample. Therefore, future research should specifically
compare attitudes of immigrant and nonimmigrant children from
ethnic majority and minority groups to disentangle the effects of
immigrant and minority status.
Negative Associations Between National Identity and
The third and final goal of the current study was to examine
whether a stronger superordinate group identity (specifically, na-
tional [American] identity) would be associated with less inter-
group bias toward its constituent subgroups. We observed that
ethnic minority and immigrant children who felt more “American”
than ethnic reported less intergroup bias. Identifying as an “Amer-
ican” may be associated with children opposing familial or cultural
norms favoring the expression of bias toward ethnic outgroups
such as African Americans, as has been reported in the case of
West Indian immigrants (Waters, 1999). Alternatively, possessing
a more central national identity may be associated with children’s
greater exposure to, or acceptance of, “American” (i.e., White
majority) norms prohibiting the expression of explicit prejudice
against outgroups (Brown & Bigler, 2004; Rutland, 1999, 2004;
Rutland, Cameron, Milne, & McGeorge, 2005). Finally, a strong
national identity also may have reduced children’s intergroup bias
via ethnic outgroup members’ inclusion in this superordinate
group. This particular finding implies that there are positive inter-
group outcomes associated with ethnic minority children possess-
ing a strong national identity, but the exact process underlying this
association remains to be examined in future research.
The results obtained from the repeated measures ANOVA also
suggest the possibility that national identity is associated with less
intergroup bias. Specifically, nonimmigrant (i.e., White and Black
American) children showed relatively little indication of inter-
group bias; the only significant finding was that White children
were significantly more positive toward Americans than ethnic
outgroups. This may have resulted in part from stronger national
(superordinate group) identities in these children as compared with
immigrant children, although it is difficult to distinguish superor-
dinate from ingroup identities (and associated consequences) for
these children. However, there are a variety of other reasons that
nonimmigrant children may have demonstrated less intergroup
bias. As mentioned previously, this also may have resulted from a
small sample of Black children or a lack of knowledge about some
of the target groups. In addition, these children probably receive a
significant amount of socialization about acceptable treatment of
ethnic outgroup members, both at home and at school. Again,
future research should look directly at intergroup biases of White
children given that the only ingroup for these children measured in
the current study was a national one.
One major caveat is that, although our independent variables of
ethnic identity, immigrant status, grade level, and gender signifi-
cantly predicted positivity and negativity biases among ethnic
minority children, the amount of variance captured by these vari-
ables was relatively low. Our estimates of R2ranged from .07 to
.13, suggesting that, on average, these particular variables may
account for about only 10% of the variance in ethnic minority
children’s intergroup attitudes. This implies that other untapped
factors also have an impact on young children’s attitudes, and
future research should attempt to capture the effect of these factors.
For example, educational variables addressing the child’s class-
room composition or exposure to antibias curricula may play an
important role in attitudes toward particular groups. Alternatively,
it is possible that our independent variables would capture a
greater percentage of variance on other measures of intergroup
attitudes, such as ones that are less susceptible to social desirability
factors or other forms of systematic bias that may have captured a
substantial amount of variance on our measures. Similarly, our
estimates of ?p
of interest in our repeated measures ANOVA (specifically, the
two-way interaction between target group and participant ethnicity
and the three-way interaction between target group, participant
ethnicity, and valence of attribute). Although the interactions sug-
gested many children possessed some degree of intergroup bias,
including both positivity and negativity biases, overall children
were unlikely to attribute negative traits to very many people from
any social group.
2were in the small to moderate range for the factors
SOCIAL IDENTITIES AND INTERGROUP BIAS
In addition, a limitation of the analysis relating national identity
centrality and intergroup bias among ethnic minority children is
that a single item was used to split the sample into two groups:
children who felt more ethnic than American and vice versa. It
should be noted that immigrant children and adolescents often feel
that they are both ethnic and American (e.g., Portes & Rumbaut,
2001). However, there was additional evidence suggesting that the
forced-choice item captured national identity centrality in a mean-
ingful way for these children. On a separate item to which children
simply indicated how American they felt (independent of how they
felt about any other social identities), those who selected more
American than ethnic on the forced-choice item had significantly
higher scores than the remaining children, t(240) ? 2.6, p ? .01.
The more concrete forced-choice item may have been necessary at
this early developmental period to test the hypothesis that national
identity centrality would be associated with less intergroup bias. In
the future, it would be ideal to develop more refined scales of
national identity centrality that would allow comparisons among
children with a variety of patterns of national and ethnic identity.
In conclusion, the analyses reported here contribute in an im-
portant way to the closer investigation of the roles social identities
play in the development of intergroup attitudes. By demonstrating
an association between ethnic identity strength and intergroup bias,
this study provides strong support for social identity theory (Tajfel
& Turner, 1979/2001, 1986) across a variety of authentic inter-
group contexts, including ethnic minority children and those from
recently immigrated families. It similarly supports its counterpart,
social identity development theory (Nesdale, 1999, 2004), al-
though we contend that no convincing evidence of outgroup der-
ogation was observed in this diverse sample. This study also
demonstrates that greater centrality of ethnic minority and immi-
grant children’s national identity is associated with more positive
attitudes toward other ethnic outgroups, supporting the superordi-
nate ingroup identity model (Gaertner & Dovidio, 2000; Gaertner
et al., 1993; Hornsey & Hogg, 2000). Furthermore, it underscores
the assertion that measures of children’s intergroup attitudes
should no longer confound measures of positivity and negativity
(Cameron et al., 2001). The separate assessment of each construct
demonstrated that positivity bias and negativity bias possessed
different, unique predictors. However, ethnic identity strength and
immigrant status were significantly associated with both kinds of
bias, suggesting that a continued exploration of the relationship
between identities or contexts and intergroup attitudes during this
period of development will prove fruitful. In particular, as previous
research on adolescents and young adults has proposed that ethnic
identity development decreases intergroup bias (Berry, 1984;
Berry et al., 1977; Phinney, Ferguson, & Tate, 1997), an important
next step is to determine why this relationship is reversed in
children—an endeavor that may require measures of ethnic iden-
tity that differentiate between regard and centrality, as well as
other components of a multidimensional identity.
Aboud, F. E. (1988). Children and prejudice. New York: Blackwell Pub-
Aboud, F. E. (2003). The formation of in-group favoritism and out-group
prejudice in young children: Are they distinct attitudes? Developmental
Psychology, 39, 48–60.
Ashmore, R. D., Deaux, K., & McLaughlin-Volpe, T. (2004). An organiz-
ing framework for collective identity: Articulation and significance of
multidimensionality. Psychological Bulletin, 130, 80–114.
Barlow, K. M., Taylor, D. M., & Lambert, W. E. (2000). Ethnicity in
America and feeling “American.” Journal of Psychology, 134, 581–600.
Bar-Tal, D. (1996). Development of social categories and stereotypes in
early childhood: The case of “the Arab” concept formation, stereotype,
and attitudes by Jewish children in Israel. International Journal of
Intercultural Relations, 20, 341–370.
Bar-Tal, D., & Labin, D. (2001). The effect of a major event on stereo-
typing: Terrorist attacks in Israel and Israeli adolescents’ perceptions of
Palestinians, Jordanians, and Arabs. European Journal of Social Psy-
chology, 31, 265–280.
Bennett, M., Sani, F., Lyons, E., & Barrett, M. (1998). Children’s subjec-
tive identification with the group and in-group favoritism. Developmen-
tal Psychology, 34, 902–909.
Berry, J. (1984). Cultural relations in plural societies: Alternatives to
segregation and their sociopsychological implications. In N. Miller & M.
Brewer (Eds.), Groups in contact: The psychology of desegregation (pp.
11–29). Orlando, FL: Academic Press.
Berry, J., Kalin, R., & Taylor, D. (1977). Multiculturalism and ethnic
attitudes in Canada. Ottawa: Supply and Services Canada.
Bigler, R. S., Brown, C. S., & Markell, M. (2001). When groups are not
created equal: Effects of group status on the formation of intergroup
attitudes in children. Child Development, 72, 1151–1162.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization
and the formation of intergroup attitudes in children. Child Develop-
ment, 68, 530–543.
Brown, C. S., & Bigler, R. S. (2004). Children’s perceptions of gender
discrimination. Developmental Psychology, 40, 714–726.
Cameron, J. A., Alvarez, J. M., Ruble, D. N., & Fuligni, A. J. (2001).
Children’s lay theories about ingroups and outgroups: Reconceptualiz-
ing research on prejudice. Personality and Social Psychology Review, 5,
Cross, W. (1991). Shades of black: Diversity in African-American identity.
Philadelphia: Temple University Press.
Devos, T., & Banaji, M. R. (2005). American ? White? Journal of
Personality and Social Psychology, 88, 447–466.
Dion, K. L., & Kawakami, K. (1996). Ethnicity and perceived discrimi-
nation in Toronto: Another look at the personal/group discrimination
discrepancy. Canadian Journal of Behavioural Science. Special Ethnic
Relations in a Multicultural Society, 28, 203–213.
Doyle, A.-B., & Aboud, F. E. (1995). A longitudinal study of White
children’s racial prejudice as a social cognitive development. Merrill-
Palmer Quarterly, 41, 209–228.
Fishbein, H. D. (2002). Peer prejudice and discrimination: The origins of
prejudice (2nd ed.). Mahwah, NJ: Erlbaum.
Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: The
common ingroup identity model. New York: Psychology Press.
Gaertner, S. L., Dovidio, J. F., Anastasio, P. A., Bachman, B. A., & Rust,
M. C. (1993). The common ingroup identity model: Recategorization
and the reduction of intergroup bias. European Review of Social Psy-
chology, 4, 1–26.
Gaertner, S. L., Rust, M. C., Dovidio, J. F., Bachman, B. A., & Anastasio,
P. A. (1994). The contact hypothesis—The role of a common ingroup
identity on reducing intergroup bias. Small Group Research, 25, 224–
Harter, S. (1985). The self-perception profile for children. Denver, CO:
University of Denver.
Hornsey, M. J., & Hogg, M. A. (2000). Assimilation and diversity: An
PFEIFER ET AL.
integrative model of subgroup relations. Personality and Social Psychol-
ogy Review, 4, 143–156.
Houlette, M. A., Gaertner, S. L., Johnson, K. M., Banker, B. S., Riek,
B. M., & Dovidio, J. F. (2004). Developing a more inclusive social
identity: An elementary school intervention. Journal of Social Issues, 60,
Lerner, R. M., & Schroeder, C. (1975). Racial attitudes in young White
children: A methodological analysis. Journal of Genetic Psychology,
Nesdale, D. (1999). Developmental changes in children’s ethnic prefer-
ences and social cognitions. Journal of Applied Developmental Psychol-
ogy, 20, 501–519.
Nesdale, D. (2004). Social identity processes and children’s ethnic preju-
dice. In M. Bennett & F. Sani (Eds.), The development of the social self
(pp. 219–246). East Sussex, England: Psychology Press.
Nesdale, D., Durkin, K., Maass, A., & Griffiths, J. (2004). Group status,
out-group ethnicity and children’s ethnic attitudes. Journal of Applied
Developmental Psychology, 25, 237–251.
Nesdale, D., Durkin, K., Maass, A., & Griffiths, J. (2005). Threat, group
identification and children’s ethnic prejudice. Social Development, 14,
Nesdale, D., & Flesser, D. (2001). Social identity and development of
children’s group attitudes. Child Development, 72, 506–517.
Nesdale, D., Maass, A., Durkin, K., & Griffiths, J. (2005). Group norms,
threat, and children’s racial prejudice. Child Development, 76, 652–663.
Nesdale, D., Maass, A., Griffiths, J., & Durkin, K. (2003). Effects of
in-group and out-group ethnicity on children’s attitudes toward members
of the in-group and out-group. British Journal of Developmental Psy-
chology, 21, 177–192.
New York City Department of City Planning. (1996). The newest New
Yorkers 1990–1994: An analysis of immigration to NYC in the early
1990s (DCP No. 96–19). New York: Author.
Phinney, J. (1989). Stages of ethnic identity development in minority group
adolescents. Journal of Early Adolescence, 9, 34–49.
Phinney, J. (1992). The Multigroup Ethnic Identity Measure: A new scale
for use with diverse groups. Journal of Adolescent Research, 7, 156–
Phinney, J., & Alipuria, L. (1990). Ethnic identity in college students from
four ethnic groups. Journal of Adolescence, 13, 171–183.
Phinney, J., Cantu, C., & Kurtz, D. (1997). Ethnic and American identity
as predictors of self-esteem among African American, Latino, and White
adolescents. Journal of Youth and Adolescence, 26, 165–185.
Phinney, J., & Devich-Navarro, M. (1997). Variations in bicultural iden-
tification among African American and Mexican American adolescents.
Journal of Research on Adolescence, 7, 3–32.
Phinney, J., Ferguson, D. L., & Tate, J. D. (1997). Intergroup attitudes
among ethnic minority adolescents: A causal model. Child Development,
Phinney, J., Horenczyk, G., Liebkind, K., & Vedder, P. (2001). Ethnic
identity, immigration, and well-being: An interactional perspective.
Journal of Social Issues, 57, 493–510.
Phinney, J., Romero, I., Nava, M., & Huang, D. (2001). The role of
language, parents, and peers in ethnic identity in immigrant families.
Journal of Youth and Adolescence, 30, 135–153.
Portes, A., & Rumbaut, R. G. (2001). Legacies: The story of the immigrant
second generation. Berkeley: University of California Press.
Ruble, D. N., Alvarez, J. M., Bachman, M. A., Cameron, J. A., Fuligni,
A. J., Garcı ´a Coll, C., et al. (2004). The development of a sense of “we”:
The emergence and implications of children’s collective identity. In M.
Bennett & F. Sani (Eds.), The development of the social self (pp. 29–76).
East Sussex, England: Psychology Press.
Rumbaut, R. (1994). The crucible within: Ethnic identity, self-esteem, and
segmented assimilation among children of immigrants. International
Migration Review, 28, 748–794.
Rutland, A. (1999). The development of national prejudice, in-group
favoritism and self-stereotypes in British children. British Journal of
Social Psychology, 38, 55–70.
Rutland, A. (2004). The development and self-regulation of intergroup
attitudes in children. In M. Bennett & F. Sani (Eds.), The development
of the social self (pp. 247–265). East Sussex, England: Psychology
Rutland, A., Cameron, L., Milne, A., & McGeorge, P. (2005). Social norms
and self-presentation: Children’s implicit and explicit intergroup atti-
tudes. Child Development, 76, 451–466.
Schuster, B., Ruble, D. N., & Weinert, F. E. (1998). Causal inferences and
the positivity bias in children: The role of the covariation principle.
Child Development, 69, 1577–1596.
Sellers, R. M., Smith, M. A., Shelton, J. N., Rowley, S. J., & Chavous,
T. M. (1998). Multidimensional model of racial identity: A reconceptu-
alization of African American racial identity. Personality and Social
Psychology Review, 2, 18–39.
Shelton, J. N., & Sellers, R. M. (2000). Situational stability and variability
in African American racial identity. Journal of Black Psychology, 26,
Sherif, M. (1966). The psychology of social norms. Oxford, England:
Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory
of social hierarchy and oppression. New York: Cambridge University
Tajfel, H., & Turner, J. (1986). The social identity theory of intergroup
behavior. In S. Worchel & W. G. Austin (Eds.), The psychology of
intergroup relations (pp. 7–24). Chicago: Nelson-Hall.
Tajfel, H., & Turner, J. (2001). An integrative theory of intergroup conflict.
In M. A. Hogg & D. Abrams (Eds.), Intergroup relations: Essential
readings (pp. 94–109). Philadelphia: Psychology Press. (Reprinted from
The social psychology of intergroup relations, pp. 33–47, by W.G.
Austen & S. Worchel, Eds., 1979, Monterey, CA: Brooks/Cole)
Waters, M. (1999). Black identities: West Indian immigrant dreams and
American realities. New York: Sage.
Weeks, M. F., & Moore, R. P. (1981). Ethnicity-of-interviewer effects on
ethnic respondents. Public Opinion Quarterly, 45, 245–249.
Williams, J. E., Best, D. L., Boswell, D. A., Mattson, L. A., & Graves, D. J.
(1975). Preschool Racial Attitude Measure II. Educational and Psycho-
logical Measurement, 35, 3–18.
Williams, J. E., & Morland, J. K. (1976). Race, color, and the young child.
Chapel Hill: University of North Carolina Press.
Yee, M. D., & Brown, R. (1992). Self-evaluations and intergroup attitudes
in children aged three to nine. Child Development, 63, 619–629.
Received September 19, 2005
Revision received March 22, 2006
Accepted March 24, 2006 ?
SOCIAL IDENTITIES AND INTERGROUP BIAS