Content uploaded by Jacopo Grisolaghi
Author content
All content in this area was uploaded by Jacopo Grisolaghi on Mar 21, 2018
Content may be subject to copyright.
Telling tales in school: extended contact interventions
in the classroom
Krista Maywalt Aronson
1
, Cristina Stefanile
2
, Camilla Matera
2
, Amanda Nerini
2
, Jacopo Grisolaghi
2
,
Gianmarco Romani
2
, Federica Massai
2
, Paolo Antonelli
2
, Laura Ferraresi
3
, Rupert Brown
4
1
Bates College
2
University of Florence
3
Pedagogist and Childhood Writer Firenze
4
Sussex University
Correspondence concerning this article should
be addressed to Krista Maywalt Aronson,
Bates College, Psychology, 4 Andrews Road,
Lewiston, Maine 04240, USA.
E-mail: karonson@bates.edu
doi: 10.1111/jasp.12358
Abstract
Extended contact has demonstrated its effectiveness in promoting more positive
intergroup attitudes among children in previous experimental research. The current
project evaluates the use of extended contact interventions involving story-telling in
two novel school settings: the United States (N5213) and Italy (N5222) among
children aged 5–12 years. Findings indicate support for the effectiveness of the inter-
ventions in these two locations among younger minority and majority children but
call into question the durability of positive outcomes. Applications and implications
are discussed.
Telling tales in school: moving
extended contact interventions one
step closer to the classroom
The pervasiveness of different forms of prejudice in children
under the age of ten years is by now well established (Aboud,
1988; Brown, 2010; Killen & Rutland, 2011). There is now
incontrovertible evidence for the existence of ingroup favorit-
ism and, in many cases, outgroup derogation among children
in the domains of ethnicity, gender, body size, language,
nationality and even artificial groups (Asher & Allen, 1969;
Castelli, Zogmaister, & Tomelleri, 2009; Cramer & Steinwert,
1998; Dunham, Chen, & Benaji, 2013; LaFreniere, Strayer, &
Gauthier, 1984; Maccoby & Jacklin, 1987; Nesdale, Durkin,
Maass, & Griffiths, 2005; Yee & Brown, 1994). Given this
apparent ubiquity of prejudice in young children, it is not
surprising that researchers have turned their attention to
finding effective ways of reducing it (Aboud et al., 2012). The
work reported here is a contribution to that research effort
and explores the utility—and also the limitations—of story-
telling interventions in elementary schools.
Intergroup contact as the basis
for interventions
The year 2014 marks the 60th anniversary of the publication
of Allport’s (1954) TheNatureofPrejudice.Overthose60
years, this book has come to be regarded as the standard point
of departure for those interested in ways to reduce prejudice,
primarily because of its inclusion of the Contact Hypothesis.
As is well known, Allport (1954) proposed that sustained
equal status, goal interdependent, cooperative contact between
members of different groups with the support of institutional
authorities would bring about favorable attitude and behav-
ioral change in both groups. Research has largely supported
the Contact Hypothesis, even if effect sizes are typically mod-
est (Brown & Hewstone, 2005; Pettigrew & Tropp, 2006).
Direct contact is thought to diminish prejudice by enhanc-
ing knowledge about the outgroup, increasing empathy and
perspective taking and reducing anxiety about intergroup
contact (Stephan & Stephan, 1985; Pettigrew & Tropp, 2008).
Higher levels of intergroup anxiety, for instance, have been
associated with diminished willingness, interest and motiva-
tion to interact with members of an outgroup (Halperin,
Crisp, Trzesniewski, Dweck, & Gross, 2012), potentially lead-
ing to the polarization of evaluations of outgroup members
(Stephan & Stephan, 1985). Direct contact is thought to
reduce this anxiety, thus facilitating cross-group interactions.
Similarly, intergroup anxiety has been found to mediate the
relationship between direct contact attitudinal and behavioral
outcomes (Pettigrew & Tropp, 2008).
Recent decades have seen theoretical elaborations of the
Contact Hypothesis that promise to widen its sphere of
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology 2016, 46, pp. 229–241
Journal of Applied Social Psychology 2016, 46, pp. 229–241
applicability and shed light on the processes that underlie the
contact–prejudice relationship. Of most relevance here are
Gaertner and Dovidio’s (2000) Common Ingroup Identity
model and Brown and Hewstone’s (2005) Intergroup Contact
Theory. According to Gaertner and Dovidio, contact will be
more effective if it permits and encourages a recategorization
of the intergroup situation from one involving two groups to
one with a single superordinate group, also called a common
identity group. They have adduced much evidence in support
of this idea (Gaertner & Dovidio, 2012). While not disagree-
ing with the importance of recategorization, Brown and Hew-
stone have argued that it is important to retain some category
salience in contact situations to allow the generalization of
the subsequent attitude and behavioral change. They also
proposed a variety of mechanisms which are likely to mediate
the contact-prejudice relationship, of which reduced inter-
group anxiety seems to be prominent (Brown & Hewstone,
2005). A combination of the two models is the Dual Identity
Model in which some subgroup distinctiveness is maintained
under the umbrella of a superordinate category. There is some
evidence that direct contact under such dual identity condi-
tions can have positive consequences (Gonz
alez & Brown,
2006; Gaertner, Rust, Dovidio, Bachman, & Anastasio, 1994;
Hornsey & Hogg, 2000).
As influential as the Contact Hypothesis has been in
informing policies aiming to reduce prejudice (Pettigrew,
2004), direct contact interventions suffer from a number of
challenges. They can be practically difficult and often expen-
sive to implement. Arranging for neighborhoods, schools,
and work-places to be more ethnically or religiously diverse
can sometimes pose problems, especially in areas where there
is already a degree of de facto residential segregation. Counter-
acting that segregation can be logistically and politically chal-
lenging, as the history of school bussing in the United States
reveals (Cascio, Gordon, Lewis, & Reber, 2008; Reardon,
Greenwall, Kalogrides, & Greendberg, 2012). Even in post-
Apartheid South Africa, the opportunities for direct inter-
group contact remain limited due to economic and other
factors (Dixon & Durrheim, 2003). Finally, there are indica-
tions that there is asymmetry in the contact-prejudice rela-
tionship between majority and minority groups; the effects of
direct contact are generally stronger for majorities than for
minorities (Binder et al., 2009; Tropp & Pettigrew, 2005).
Extended contact hypothesis
The extended (or indirect) contact hypothesis offers some
promiseinaddressingtheseconcerns(Wright,Aron,
Mclaughlin-Volpe, & Ropp, 1997). According to this hypoth-
esis, it is not necessary for direct contact to occur to facilitate
prejudice reduction; it may be enough to experience contact
indirectly by knowing (of) ingroup members who have posi-
tive relationships with outgroup members. Such indirect con-
tact is thought to work by any or all of the following four
mechanisms (Wright et al., 1997): reduced intergroup anxi-
ety, altered perceptions of ingroup or outgroup norms about
the appropriateness of cross-group friendships, and inclusion
of the other in the self. Research from a wide range of inter-
group contexts has found support for these ideas (De Tezanos
Pinto, Bratt, & Brown, 2010; Paolini, Hewstone, Cairns, &
Voci, 2004; Pettigrew, Christ, Wagner, & Stellmacher, 2007;
Turner, Hewstone, Voci, Paolini, & Christ, 2007; Turner et al.,
2007; Wright, Aron, McLaughlin-Volpe, & Ropp, 1997).
Of course, direct and indirect contact are not mutually
exclusive; they can, and often do, co-occur. When they do,
some research suggests that direct contact is the more potent
determinant of prejudice reduction since, given the presence
of many opportunities for direct contact, indirect contact has
weak or negligible effects (Awad, 2010; Christ et al., 2010).
The durability of indirect contact effects over time remains to
be established (Feddes, Noack, & Rutland, 2009), whereas the
longitudinal effects of direct contact are now well attested
(Binder et al., 2009; Feddes et al., 2009; Levin, van Laar, &
Sidanius, 2003). Nevertheless, in cross-sectional studies,
where both direct and indirect contact are measured, both are
found to be associated with lessened prejudice, albeit via dif-
ferent mediational routes (De Tezanos Pinto, Bratt, & Brown,
2010; Turner et al., 2007; Turner, Tam, Hewstone, Kenworthy,
& Cairns, 2013).
A significant potential benefit offered by indirect contact is
in the design of prejudice reduction interventions in schools.
Among the first to exploit this were Liebkind and McAllister
(1999) in their use of story-telling in Finnish schools. The
stories depicted Finnish adolescents enjoying close friend-
ships with immigrants and were accompanied by class-room
discussions of the benefits of such cross-group relationships.
Children in the intervention schools showed improved inter-
group attitudes toward foreigners compared to the control
schools where no stories were told.
Cameron and her colleagues developed this idea further by
adapting existing children’s story books which portrayed
friendships between young majority and minority group chil-
dren (Cameron, Rutland, Brown, & Douch, 2006; Cameron,
Rutland, Hossain, & Petley, 2011a) to examine the effects of
extended contact among majority (White) children aged
between 5 and 12 years. Different books with broadly similar
content were read to participating children, and, depending
on experimental condition, discussions following book read-
ings were designed to highlight either “common ingroup
identity” (e.g., noting how the children all belong to the same
community) “dual identity” (noting how the children belong
to the same community and have their own unique cultures),
or “decategorizaton” (noting individual character qualities
and de-emphasizing category memberships) aspects of the
relationship between the protagonists in the stories; “control”
children received no stories (Cameron et al., 2006). At post-
230 Telling tales in school
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
test (1 week after the intervention), all three intervention con-
ditions showed more favorable attitudes toward the target
group (asylum-seekers) than controls, with the dual identity
condition showing the most positive attitudes of all. The gen-
eral effect of condition was partially mediated by enhanced
inclusion of other in the self (Cameron et al., 2006).
In a subsequent study, using a similar paradigm, the inter-
vention was simplified by dropping the “interpersonal” con-
dition, and the moderating effects of prior intergroup
contact and age were examined (Cameron, Rutland, Hossain
et al., 2011a). This time, no difference was observed between
“dual” and “common ingroup” conditions, though both
proved superior to the no story “controls.” These experimen-
tal effects were partially mediated by perceived outgroup
norms and were qualified by age of child and amount of
prior contact with minority group children: the intervention
was only effective for younger children (6–7 years), and for
those who had had little prior contact.
Behavioral journalism has also been used to enact
extended contact with older youth. Using this approach,
researchers expose participants to newsletters or stories
depicting cross-group friendships followed by guided discus-
sions. Using this approach with Finnish youth grades 7–9,
Liebkind, M€
ah€
onen, Solares, Solheim and Jasinskaja Lahti
(2014) also found some support for the benefits of extended
contact among their younger participants, although their
intervention seemed to increase intergroup anxiety among
older participants.
Moving extended contact
interventions one step closer
to the classroom
In sum, findings show promise of extended contact interven-
tions, and offer a practical mechanism for affecting inter-
group attitudes and relationships among elementary and
adolescent aged youth. But, several issues need further explo-
ration before such work can be handed off to classrooms and
schools with confidence. For instance, although the longitu-
dinal effects of direct contact are now well attested (Binder
et al., 2009; Feddes et al., 2009; Levin, van Laar, & Sidanius,
2003; Swart, Hewstone, Christ, & Voci, 2011), the durability
of indirect contact effects over time remains to be established
(Feddes et al., 2009; Beelman & Heinemann, 2014) which is
an important component as schools will want to evaluate
long term effects prior to utilizing story book techniques.
Additionally, it is not clear whether indirect contact effects
are equally strong among majority and minority children.
Previous research indicates that the developmental course of
intergroup attitudes for minority children is different than
for majority children, with minority attitudes becoming
more negative with age, unlike majority attitudes, which
seem to decrease between ages 5–7 and then increase
(Cameron et al., 2006; Cameron, Rutland, Hossain et al.,
2011a; Liebkind & McAllister, 1999). This minority trajectory
is likely attributable to increased awareness of social attitudes
and encounters with racism and discrimination both at
school and within the broader society (Raabe & Beelman,
2011). In correlational work, Feddes et al. (2009) found that
TurkishchildreninGermanydidnotshowindirectcontact
effects (cf., Turner et al., 2007). Although Liebkind et al.
(2014) found support for the benefits of indirect contact
among younger minority and majority adolescents using a
behavioral journal based intervention, previous story book
based indirect contact studies have included White children
exclusively (Cameron et al., 2006; Liebkind & McAllister,
1999). As many school led interventions would involve inte-
grated classrooms and reading groups, it is important to con-
sider the effect books will have on both minority and
majority children.
Several methodological improvements are also necessary
to move story book programs closer to school based, com-
munity led interventions. Within previous story book based
interventions researchers read children in all experimental
conditions a similar story with minor variations, relying on
discussions following each reading to convey an experimental
message (e.g., common group or dual identity; Cameron
et al., 2006; Cameron, Rutland, Hossain, et al. 2011). This
approach posed several methodological challenges, including
the fact that there were varieties in story content across con-
ditions and character gender imbalance (the majority of
book characters were male). To insure that children will iden-
tify with the story book characters, systematize message
delivery, control more precisely the experimental variations
one seeks to implement, and ensure similarity across multiple
readers within the community, specific messages should be
embedded within the stories themselves. Likewise, although
the efficacy of story book interventions has been established
in two countries, England and Finland, the extension of story
book interventions to other contexts with different immi-
grant populations may enhance our ability to anticipate how
well the technique will generalize to other locations. In doing
so, we are assuming no equivalence between contexts, but
asserttheimportanceoftestingindiverseenvironments
(e.g., different educational systems and immigrant groups) is
important to understanding how well extended contact inter-
ventions generalize to differing environments. Finally, previ-
ous story book interventions used ‘no story’ control
conditions, thus leaving open the possibility that the experi-
mental effects observed were due to story-telling per se,
rather than to the indirect contact portrayed in the stories.
Finally, evaluative measures have often been administered
by researchers knowledgeable of participant condition, a sub-
optimal methodology (Cameron et al., 2006; Cameron,
Rutland, Hossain, et al., 2011a).
Aronson et al. 231
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
The current study
In present research, we extend previous research in the fol-
lowing ways. First, we used purpose-written stories in which
the content of the narratives was carefully controlled, accord-
ing to experimental condition. Second, we extended the use
of this story-telling technique to two new immigration
contexts–American-Somali relations in the USA, Italian-
immigrant (mainly North African & Chinese) relations in
Italy. These national contexts and immigrant groups are
unique to one another and to previously researched areas,
thus allowing for the exploration of generalizability of this
type of intervention. Third, children in our control groups
also received stories with identical narratives to the other
conditions, but with outgroup protagonists replaced by
ingroup protagonists. Fourth, we examined the effects of the
intervention with minority as well as majority group chil-
dren. Fifth, we examined the durability of the story-telling
interventions by including a second post-test 9 weeks after
the intervention. Finally, although findings indicate inter-
group anxiety plays a role in direct and indirect contact proc-
esses (Liebkind et al., 2014; Pettigrew & Tropp, 2008; Stephan
& Stephan, 1985), previous research using a story book
approach has not yet examined the role of intergroup anxi-
ety. Thus, we included intergroup anxiety as a potential
mediator of the effects of the intervention on intergroup atti-
tudes to expand our understanding of the role anxiety may
play in extended contact effects among children.
Our hypotheses were: (1) Story-telling interventions
depicting positive intergroup relationships between majority
and minority children would improve intergroup attitudes
and behavioral outcomes, especially among younger age chil-
dren (following Cameron, Rutland, Hossain, et al., 2011a;
Cameron, Rutland, Turner, Holman-Nicolas, & Powell, 2011;
Liebkind et al., 2014). (2) Intergroup anxiety would mediate
the effects of story-telling interventions on intergroup atti-
tudes (following Liebkind et al., 2014; Pettigrew & Tropp,
2008; Stephan & Stephan, 1985). (3) The effects of extended
contact would be similar for majority and minority children
(following Liebkind et al., 2014). Given the lack of research
on stability of extended contact effects, no additional hypoth-
eses were made a priori.
Study 1
Method
Participants and settings
Participants were 213 elementary school children (54% male;
106 second grade, 107 third grade). The self-reported mean
age was 7.80 years (range 7–9 years). Birth date information
was not made available by participating schools. The State
requirement for Kindergarten admission is age 5 by October
15th.Thus,weusegradeasamoreaccurateproxyforagein
this sample. The majority of participants (177) were non-
Somali (including, 166 Caucasian, 11 African American or
mixed race descendants), 36 were Somali as indicated by
school records; all attended schools in an urban area of
Northern New England, US. Eighty percent of students from
thedistrictswhereparticipantsweredrawnwereWhiteand
the offspring of parents born in the United States.
Design
The study employed a 3 (intervention condition: control,
dual identity, common group) 32 (grade: second, third)
between-participants design. Approximately two (T1) and
nine (T2) weeks after the intervention, a different group of
researcher assistants assessed attitudinal outcomes and a
potential mediator.
Procedure
All second and third grade children in the target schools par-
ticipated in an extended contact intervention. To avoid
potential confounds between classroom and condition, chil-
dren in each class were individually and randomly assigned
to one of the conditions (common group, dual identity, and
control) and exposed to six (weekly) storybook sessions led
by researcher assistants. To ensure that dual identity, com-
mon group or control messages were directly embedded in
the stories themselves, a set of (6) stories were written by the
researchers in collaboration with two local children’s book
authors, and illustrated by a local children’s book illustrator
(stories available upon request to first author). Each story
included the same class and four protagonists (teo boys and
two girls) engaging in normal school activities: a writing
competition, a field trip, mural painting, a class play, a rainy
week, and a pizza making competition (stories available from
first author on request). The class was used as a superordi-
nate category in both Common group and Dual identity
conditions. Three versions of each story were written: one
control version, in which the four protagonists were Cauca-
sian, as indicated by their names and through illustration;
and, two experimental versions. One, the common group
version, was identical to the control version except that two
of the four protagonists were of obvious Somali descent (one
male and one female). Although themes of group cohesion
ran throughout these stories, no additional mention was
made to cultural background, consistent with the theoretical
conceptualization of a common group identity (Gaertner
et al., 1998). The second, the dual identity version, was iden-
tical to the common group version with respect to the incor-
poration of male and female Somali and non-Somali
children, but these stories also included cultural information,
232 Telling tales in school
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
designed to help students think about the ways in which each
childwasbothconnectedwithaspecificculture(Somalior
non-Somali) as well as a common group (the class). For
instance, Somali children in the dual identity stories shared
information about religious holidays, henna or Somali food.
Non-Somali children were all depicted as Caucasian in these
stories and shared information about nonreligious holidays
like Thanksgiving and culturally American sports like
baseball.
A trained research assistant read stories to children in
small groups (3–5 children in a quiet area outside of the
classroom). All readings were followed by discussion
designed to enhance reading comprehension and further
support the primary message of each condition. Reading ses-
sions lasted approximately 15–20 minutes and took place
once a week for 6 weeks, with a different story being read
each week.
Children were recruited for program evaluation with the
support and endorsement of their local school district and
local school administration. Consent forms were sent home
with all children in their homework folder. These forms
described the nature of the research, including sample ques-
tions and parents were asked to return the form if they did
not want their child to participate. Because English and writ-
ten Somali are not well understood within the Somali com-
munity, a cultural broker was hired to contact all Somali
families either by phone or in person to offer the consent
form orally. One hundred percent of Somali families and
85% of non-Somali parents granted consent. Child assent
was obtained prior to questionnaire completion. Question-
naires were counterbalanced and administered by trained
research assistants blind to each child’s experimental condi-
tion in two 30-minute one-on-one sessions: once (T1) 2–3
weeks following intervention completion; and, a second time
(T2) 9–10 weeks following intervention completion
(N5194). As a thank you, children were offered a small
($3.00 value) toy. All procedures were approved by the insti-
tutional review board at the first author’s institution, by indi-
vidual schools, and by school superintendents.
Measures
Prior to the completion of any measures children were given
instructions to help them understand the two target groups,
Somali and non-Somali, they would be asked to consider.
Primary measures included intended behaviors and outgroup
attitudes. Intergroup anxiety was also assessed as a potential
mediator of contact-prejudice relationships (Brown &
Hewstone, 2005; Pettigrew & Tropp, 2008). Prior contact was
included as a potentially moderating variable, in view of the
well-established link between outgroup friendships and inter-
group attitudes (Pettigrew & Tropp, 2006). Finally, grade was
also included as an additional (potentially moderating) vari-
able given established patterns between age and outgroup
attitudes among children (Doyle, Beaudet, & Aboud, 1988;
Williams, Best, Boswell, Mattson, & Graves, 1975).
Intended Behaviors were assessed using a series of short
scenarios regarding future interactions with Somali or non-
Somali children. If they were children of non-Somali descent
they were told about a same sex Somali or non-Somali child
named Fatima or Ismail; if they were of Somali descent, they
were told about a same sex non-Somali child named Sophie
or Max. Then, they were asked to indicate how much they
would want to engage in six separate activities with the target
child, ranging from playing with them on the playground, to
having them over for dinner. Children were asked to indicate
their interest using a five point Likert type scale, in which
possible responses were indicated by smiley faces ranging
from “not at all” (deep frown) to “very much” (big smile).
One score representing the average response for intended
behavior items was calculated for each child. Scores ranged
from 1 to 5, with higher scores on this scale indicated more
positive intentions (T1a5.76; T2a5.82).
To a s se s s outgroup attitudes, children were presented with
seven positive and seven negative adjectives and asked to
indicate how many Somali or non-Somali people were that
way. Like intended behaviors, children were asked to respond
regarding an outgroup child (e.g., Somali if they were non-
Somali). The positive words were kind, hard working, polite,
clean, likeable, good and friendly. The negative words were
unfriendly, rude, lazy, sloppy, dislikable, naughty/bad, and
unkind. These adjectives were taken from the Preschool
Racial Attitude Measure—II (PRAM—II) Series A (Williams
et al., 1975). Using a four point Likert scale children were
asked to respond by choosing from an array of circles con-
taining different numbers of stick people (see Abrams et al.,
2003) corresponding with the words, “none,” “some,”
“most,” and “all,” which were written underneath each circle.
The order of the adjectives was randomized for each child.
Positive and negative outgroup attitude scores were calcu-
lated for each child. These scores ranged at Time 1 from 7 to
28 and 6 to 25 and at Time 2 from 11 to 28 and 7 to 23,
respectively. Cronbach’s alphas for positive (T1a5.84;
T2a5.75) and negative adjectives (T1a5.77; T2a5.81)
indicate reasonable reliability. An outgroup attitude score for
each child was calculated by subtracting positive attitude
scores from negative attitude scores. Thus, higher scores on
this scale indicate more positive attitudes.
To assess intergroup anxiety children were presented with a
picture of several, same sex, Somali or non-Somali children
that they did not know and asked to imagine that they were
with them. As with the intended behaviors and outgroup
attitudes measures, they were asked to respond regarding an
outgroup child. Using a five point likert scale for which pos-
sible responses were indicated by the presence of increasingly
large circles ranging from “not at all” (red X through circle)
Aronson et al. 233
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
to “very much” (large green circle), children were presented
with six adjectives (well, happy, afraid, nervous, calm, and
worried) and asked to indicate how they would feel. Scores
ranged from 1 to 4 (T1a5.71; T2a5.74).
Prior Contact with Somali or non-Somali children was
assessed using one item, “how many Somali/non-Somali
friends do you have?,” for which children were asked to indi-
cate their response by choosing from one of five squares con-
taining a set of dots corresponding with the words, “none,”
“one,” “two or three,” “four or five,” and “more than five.”
Scores ranged from 0 to 5 with a median score of 4.0.
Results
Analyses were first conducted with child grade and sex and
ethnic group membership (Somali, non-Somali) as inde-
pendent variables. All main and interaction effects involving
these control variables were non-significant.
1
Thus, only the
analyses conducted with the primary independent variables,
school grade, experimental condition and prior contact will
be presented. Variable interrelationships at T1 and T2 can be
found in Table 1.
Time 1
Intended behaviors
Hierarchical regression analyses were utilized to explore the
effects of condition on intended behavior. Step 1 incorpo-
rated the control variables grade and prior contact. Step 2
incorporated a set of planned, orthogonal contrasts: C1
tested the control against the two extended contact condi-
tions (weights 22, 11, 11); C2 tested the dual identity con-
dition against the common group condition (weights 0, 21,
11). The third step incorporated possible interactions
between contrasts, grade and prior contact (see Table 2).
Steps 1 and 2 revealed a significant main effect for prior
contact: more contact was associated with higher intended
behavior scores (b5.18, p5.01). In step 3 a C1 X grade
interaction was detected (b52.45, p5.04) as well as a C2
X grade interaction (b52.42, p5.05). To decompose these
interactions the regression was conducted for each grade sep-
arately. Findings indicate a significant main effect for C1
among younger children, with younger children in the con-
trol group reporting significantly lower levels of intended
behavior than in the two experimental groups (b5.22,
p5.02). In addition, findings from C2 reveal that the
intended behaviors among second grade children in the dual
identity condition were more positive than in the common
group identity condition (b5.22, p5.02; see Figure 1).
Outgroup attitudes
Hierarchical regression analyses were also used to explore the
effects of condition on outgroup attitudes using the same
model as described for intended behaviors. As is apparent
from Table 2, prior contact was the only significant predictor.
Intergroup anxiety
Hierarchical regression analyses were also used to explore the
effects of condition on intergroup anxiety attitudes using the
Table 1 Study 1 Variable Interrelationships at T1 and T2
Measure 1 2 3 4 5
1. Intended behaviors .45** 2.45** .18** .06
2. Outgroup attitudes .34*** 2.38** .32** 2.09
3. Intergroup anxiety 2.28*** 2.20** 2.20** 2.06
4. Prior contact .27** .32** 2.15* .10
5. Grade .01 .09 2.01 .11
Note.T1(N5213) values above the diagonal, T2 (N5194) below
*p<.05, **p<.01, ***p<.001.
Table 2 Study 1 (T1) Intended Behaviors, Outgroup Attitudes, and
Intergroup Anxiety as a Function of Condition in Hierarchical Regression
Analyses
Intended
behavior
Outgroup
attitudes
Intergroup
anxiety
bR
2
bR
2
bR
2
Step 1 .04* .11*** .04**
Grade .06 .07 2.03
Prior Contact .18* .32*** 2.20**
Step2 .06* .11*** .05*
Grade .06 .07 2.32
Prior Contact .19* .32*** 2.20**
C1 .10 .02 2.07
C2 .11 .04 .00
Step 3 .10** .12** .07
Age .05 .06 2.03
Prior Contact .18* .31*** 2.19**
C1 .41 .39 2.31
C2 .30 .34 2.37
Grade X C1 2.45* 2.22 .32
Grade X C2 2.42* 2.18 .40
a
Prior Contact X C1 .12 2.18 2.06
Prior Contact X C2 .22 2.14 2.02
a
p5.06.
*p<.05. **p<.01. ***p<.001.
1
When Study 1 analyses are run separately for Somali and non-Somali partici-
pants significant findings disappear, although the pattern of Means mimics
those for the full sample. Given the pattern ofresults we believe the loss of sig-
nificance can be attributed to the removal of participants.
234 Telling tales in school
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
same model as described for intended behaviors. Steps 1 and
2 revealed a significant main effect for prior contact:
(b52.20, p5.004). Step 3 revealed a near significant C2
X grade interaction (b5.40, p5.06). The effect was exam-
ined in each grade separately but neither proved reliable (C2
grade 2 b52.22, p5.40, R
2
5.09); C2 grade 3 b5.32,
p5.31, R
2
5.36). Nevertheless, inspection of the means sug-
gested a similar pattern as was observed for intended behav-
ior—namely, the least intergroup anxiety for second grade
children in the dual identity condition.
Intergroup anxiety as a mediator
In view of the above marginal interaction, additional analyses
were conducted to explore intergroup anxiety as a potential
mediator of the relationship between condition and intended
behavior among the second grade children. Multiple regres-
sion analyses were employed, with the condition variable
dummy coded to mimic the condition comparisons of inter-
est (C1: 22, 1,1; C2: 0, 1, 21).
Grade X contrast effects were significant for C1 (b52.45,
p5.03) and for C2 (b52.43, p5.05). The addition of
intergroup anxiety to the model reduced both effects to non-
significance (C1: b511, p5.09; C2: b52.34, p5.08).
However, further bootstrapped analyses did not support
mediation.
Durability of effects
To explore the potential durability of effects across T1, T2,
three mixed-design ANOVAS were conducted with time as a
within-subjects factor (T1, T2) and condition (dual identity,
common group, control), age (second, third) and prior con-
tact (low, high) as between-subjects factors. Findings revealed
no significant main effects or interactions for intended
behaviors or intergroup anxiety (all pvalues >.13). A main
effect for time on outgroup attitudes emerged, F(1,
184) 5217.95, p5.000, eta squared.72. Children reported
significantly more positive attitudes across conditions. This
relationship was qualified by an interaction between
time and prior contact, F(1, 184) 514.83, p5.001, eta
squared.27: children who reported more friends reported a
larger increase in positive attitudes than those with fewer
outgroup friends.
Further analysis of Time 2
Intended behavior
Closer examination was given to T2 intended behavior scores
using the same hierarchical regression model as employed for
T1. Steps 1, 2, and 3 revealed a significant main effect for
prior contact (b5.27, p5.000). No significant main effects
or interactions for condition at T2 (all pvalues >.13).
Outgroup attitudes
Analysis of T2 outgroup attitudes revealed a main effect for
prior contact (b5.30, p5.000) with no other significant
results (all pvalues >.41).
Intergroup anxiety
Analysis of T2 intergroup anxiety scores indicated a main effect
for prior contact in steps 1 and 2 (b52.15, p5.04). A signif-
icant C2 X grade interaction also emerged at T2 (b5.53,
p5.02). Findings isolate this effect among younger children,
meaning that the second grade children in the dual identity
condition were less anxious than those in the common group
identity condition (b52.25, p5.02;seeFigure2).
Brief discussion
In line with Hypothesis 1, findings indicate that story-telling
interventions depicting positive intergroup relationships
between majority and minority children can improve inter-
group attitudes for younger children; the intervention had
Figure 2 Study 1 (T2) intergroup anxiety as a function of condition and
grade. Second grade children in the dual identity condition were less anx-
ious than those in the common group identity condition.
Figure 1 Study 1 (T1) intended behaviors as a function of condition
and grade. Intended behaviors among second grade children in the dual
identity condition were significantly more positive than in the common
group identity condition.
Aronson et al. 235
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
no significant effect among older children in the current sam-
ple. Second graders in the extended contact (common group
and dual identity) conditions reported significantly more
positive behavioral intentions than controls at T1. Moreover,
children in the dual identity condition at T1 reported signifi-
cantly more positive behavioral intentions than those in the
common group condition. No significant findings emerged
for outgroup attitudes at T1 or T2. Although intergroup anx-
iety seemed to be lower among second graders in the dual
identity condition as compared to those in the common
group condition (C2; with such an effect being stable across
time), this factor was not found to be responsible for the
observed effects of the interventions on intergroup attitudes
(see Hypothesis 2).
These findings are consistent with previous research sup-
porting the benefits of extended contact, and identifying dual
identity messages as the most effective among elementary
school children (Cameron et al., 2006), especially amongst
younger rather than older children (Cameron, Rutland, Hos-
sain, et al., 2011a; Cameron, Rutland, Turner, et al., 2011b).
Although extended contact has been found to be more effica-
cious in changing attitudes than intended behaviors in the
past (Cameron et al., 2006), this was not the case here.
Findings from within participant analyses indicate that
mean intended behavior and intergroup anxiety scores
remained stable across T1 and T2. However, although signifi-
cantly different behavioral intentions emerged for younger
children in the extended contact conditions at T1, they did
not emerge at T2. This suggests that the positive effects of
extended contact on behavioral intentions may diminish
over time. In contrast, the pattern of results suggests more
lasting results, and possibly gains, for intergroup anxiety.
Specifically, although significant grade by condition findings
emergent at T1 were not reliable, they were reliable at T2,
with younger children in the dual identity group reporting
significantly lower levels of intergroup anxiety than children
in the common group condition at this time point. Thus,
findings offer mixed support for the durability of positive
effects of extended contact three months post intervention.
Study 2
Method
Participants
Participants were 222 elementary aged school children (49%
male; mean age 9.05 years, range 5–12 years). The majority
of participants 171 were non-immigrant Italian nationals, 51
were of immigrant descent (parents born in Italy). The
schools in which testing was carried out were extremely
diverse (approximately 35% minority students), with over 30
different countries of origin represented, modally by a single
child.
Design
The study employed a three group (intervention condition:
control, dual identity, common group) between-participants
design. Dependent variables were assessed 2 weeks after the
intervention. Participant birth date information was available
for the Italian sample. Although grade level in school can be
usedasaproxyforageintheUS,inItalythisislesstrue.
Thus, age is used as a continuous variable in this sample.
Procedure
The same approach to enacting control, common group and
dual identity manipulations used in Study 1 was undertaken in
Study 2. However, Study 2 stories—five in all—featured Afri-
can and Chinese immigrant children as they are more relevant
to the context where the study was conducted (Firenze). Each
story included 5 protagonists (3 boys; 1 African; 2 Italian) and
2 girls (1 Italian; 1 Chinese) engaging in normal school activ-
ities (e.g., a writing competition, trying to locate a missing
pen). Three versions of each story were written: a control ver-
sion, in which the 5 protagonists were Italian, as indicated by
their names; a common group version, which was identical to
the control version except that two of the five protagonists
were of obvious African or Chinese descent (one male and one
female) and there were continual references to the school class
as a common ingroup identity; and, a dual identity condition,
which was similar to the common group version with respect
to the incorporation of immigrant children, but in which the
stories also included cultural information, designed to help
students think about the ways in which each child was both
connected with a specific cultural as well as a common group
identity. Each story was followed by a short discussion session,
designed to emphasize the main elements of the story. Follow-
up questionnaires were counterbalanced and administered in
one, 30-minute one-on-one sessions 2 weeks following the
intervention (N5222).
Measures
As in Study 1, children were administered a set of question-
naires designed to assess two attitudinal outcomes, intended
behaviors and outgroup attitudes, and intergroup anxiety. All
of the measures and response scales used in Study 2 were
identical to those used in Study 1 except that they were pre-
sented in Italian.
Children were first given instructions to help them under-
stand the two target groups, immigrant and non-immigrant,
they would be asked to consider. Next, the researcher said:
“Now, can you think about immigrants and can you also
236 Telling tales in school
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
think about Italian people? I want to ask you some questions
about immigrants and Italian people.” This introduction was
reviewed prior to the administration of each measure and
provided a comparison context for each question. Pertinent
psychometric information for Study 2 can be found below.
Please see the Study 1 methods section for a detailed descrip-
tion of each measure.
Intended Behaviors
Scores ranged from 1 to 5, with higher scores on this scale
indicating more positive intentions (a5.72). Positive and
negative outgroup attitude scores were calculated for each
child. These scores ranged from 11 to 30 and 6 to 28, respec-
tively. Cronbach’s alphas for positive (a5.69) and negative
adjectives (a5.72) indicate reasonable reliability. An out-
group attitude score for each child was calculated by sub-
tracting positive attitude scores from negative attitude scores.
Thus, higher scores on this scale indicate more positive atti-
tudes. Intergroup anxiety scores ranged from 1 to 5 (a5.73).
Prior Contact scores ranged from 1 to 5 with a median score
of 3.0.
Results
Analyses were first conducted with child school, sex, and eth-
nic group membership (immigrant, nonimmigrant) as inde-
pendent variables. All main and interaction effects involving
these control variables were nonsignificant. Thus, only the
analyses conducted with the primary independent variables,
child age, experimental condition, and prior contact will be
presented.
Intended behavior
Variable relationships can be found in Table 3. Hierarchical
regression analyses were utilized to explore the effects of con-
dition on intended behavior (see Table 4). The first step of
this model incorporated the control variables age and prior
contact. Step 2 incorporated a set of planned, orthogonal
contrasts: (C1) tested the control against the two extended
contact conditions (weights 22, 11, 11); (C2) tested the
dual identity condition against the common group condition
(weights 0, 21, 11). The third step incorporated possible
interactions between contrasts, age and prior contact.
Steps 1 and 2 revealed only a significant main effect for
Prior Contact: more contact was associated with higher
scores on intended behavior (b5.24, p5.00). In step 3, a
C1 X age interaction was detected (b52.92, p5.05; see
Table 4). To decompose this interaction, the regression was
conducted for each age group separately using the mean sam-
ple age as the grouping cut off. These ages mimic the
younger/older dichotomy within the U.S. sample (5–8 and
9–12 years). Findings indicate a significant main effect for C1
among younger children, with younger children in the con-
trol group reporting significantly lower levels of intended
behavior than in the two experimental groups (b5.25,
p5.03;seeFigure3).
Outgroup attitudes
Hierarchical regression analyses were also used to explore the
effects of condition on outgroup attitudes, using the same
modelasabove.AsisapparentfromTable4,priorcontact
was the only significant predictor, with children reporting
more prior contact reporting significantly more positive atti-
tudes (b5.24, p5.00).
Table 3 Study 2 Variable Interrelationships (N 5222)
Measure 1 2 3 4 5
1. Intended behaviors .32** .40** .24** 2.01
2. Outgroup attitudes .32** .23** 2.05
3. Intergroup anxiety .27** .11
4. Prior contact .09
5. Age
*p<.05. **p<.01. ***p<.001.
Table 4 Study 2 Intended Behaviors, Outgroup Attitudes, and Inter-
group Anxiety as a Function of Condition in Hierarchical Regression
Analyses
Intended
behavior
Outgroup
attitudes
Intergroup
anxiety
bR
2
bR
2
bR
2
Step 1 .06** .06** .08***
Age 2.30 2.07 .08
Prior contact .24*** .24*** .26***
Step2 .07** .06** .09***
Age 2.30 2.08 .08
Prior contact .24*** .24*** .26***
C1 .09 2.08 2.06
C2 .03 .01 .03
Step 3 .09** .07 .11**
Age 2.03 2.08 .07
Prior contact .22** .22*** .24***
C1 1.06* 2.08 .84
C2 .17 .14 .53
Age X C1 21.01* .11 2.82
a
Age X C2 .01 2.20 2.20
Prior contact X C1 .03 2.12 2.09
Prior contact X C2 2.15 2.08 2.31
a
p5.06.
*p<.05. **p<.01. ***p<.001.
Aronson et al. 237
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
Intergroup anxiety
The same analysis was used to examine the effects of experi-
mental condition on intergroup anxiety. In this analysis,
prior contact was again a significant predictor (b5.26,
p5.00). Step 3 revealed a nearly significant C1 X age interac-
tion (b521.76, p5.08). To further probe this result, the
regression was conducted for each age group separately. Find-
ings indicate a significant main effect for C1 among younger
children in the experimental groups reporting significantly
lower levels of anxiety than in the control (b5.65, p5.02;
seeFigure4).
Intergroup anxiety as a mediator
Additional analyses were conducted to explore intergroup anx-
iety as a potential mediator of the relationship between condi-
tion and intended behavior. Contrast 1 was significant within
the model (b51.35, p<.001), as was the C1 X age interaction
(b521.00, p5.03). While the addition of intergroup anxiety
to the model did not reduce C1 (C1: ß 51.01, p5.02), it did
reduce the interaction term (C2: b521.52, p5.13). How-
ever, further bootstrapped analyses did not support mediation.
Brief discussion
Like Study 1, Study 2 findings indicate that younger children
in the extended contact (common group and dual identity)
conditions reported significantly more positive behavioral
intentions. Like Study 1, a significant condition by age interac-
tion emerged for intergroup anxiety within this sample, but in
the inverse direction: younger children in the experimental
conditions reported more anxiety. Also, differences between
the two extended contact conditions themselves did not
emerge in this sample. Findings do not support intergroup
anxiety as a potential mediator of the relationship between
condition and intended behavior. As in Study 1, no significant
findings emerged for outgroup attitudes within this sample.
General discussion
Taken as a whole, findings support the original hypothesis
that story-telling interventions depicting positive intergroup
relationships between majority and minority children can
improve some intergroup attitudes, and that these effects
seem to be particularly evident in younger age children (fol-
lowing Cameron, Rutland, Hossain, et al., 2011a).
However, the current findings are equivocal regarding
whether one extended contact message is more effective than
another: dual identity was associated with the highest behav-
ioral intention scores in the United States; but, dual identity
and common group messages were associated with equally
positive outcomes within the Italian sample. These results
somewhat mirror the extended contact research landscape in
which some research links dual identity messages with the
most positive outcomes (Cameron et al., 2006), and the
endorsement of acculturative attitudes that emphasize both
contact with mainstream culture and cultural maintenance
(integration) with prosocial intergroup attitudes and behav-
ioral intentions among majority children (Aronson & Brown,
2013), whereas other findings do not distinguish between
common group or dual identity messages (Cameron, Rutland,
Turner, et al., 2011b). For instance, Cameron Rutland, Turner,
et al. (2011b) speculate that common group messages may be
equally effective in environments where the minority group
category is already salient for children, predisposing them to
consider both superordinate and subordinate categories dur-
ing story reading. Although it was not directly assessed within
the current project, the American and Italian school settings
were slightly different in their demographic make-up (major-
ity/minority), with 65% majority children in the Italian
schools as compared to 80% in the U.S. schools, so this expla-
nation is somewhat consistent with the current data. Ulti-
mately, it may be an empirical question to resolve which type
of intervention is most effective where.
With respect to age, the current study did find the interven-
tion to be most effective at changing intended behaviors among
Figure 4 Study 2 intergroup anxiety as a function of condition and age
group. Younger children in the experimental groups reported signifi-
cantly lower levels of anxiety than those in the control group.
Figure 3 Study 2 intended behaviors as a function of condition and age
group. Younger children in the control group reported significantly lower
levels of intended behavior than those in the two experimental groups.
238 Telling tales in school
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
younger children in both the United States and Italy, thus con-
firming this result in two countries in addition to the original
study site, England (Cameron et al., 2006; Cameron, Rutland,
Hossain, et al., 2011a). The most likely explanation for this is
that older children have had more opportunities for direct con-
tact, which has proven to exert a more powerful influence on
outgroup attitudes among youth (Beelmann & Heinemann,
2014). Supporting this assertion, extended contact demon-
stratessignificantlylessofanimpactwhenchildrenhadprior
high quality direct contact with the target outgroup (Cameron,
Rutland, Turner, et al., 2011b). However, intergroup anxiety
yielded mixed results, with the general trend indicating lower
intergroup anxiety among younger children in the United
States. and but not in Italy, even if this outcome did not seem
to impede the behavioral intentions of Italian children.
Also, although extended contact has been found to be more
efficacious in changing outgroup attitudes than intended
behaviors in the past (Cameron et al., 2006), this was not the
case within Study 1 or Study 2. This may be a measurement
artifact, as children within both study settings were asked to
respond to outgroup attitude questions with respect to immi-
grants in general, whereas behavioral intention questions
focused on immigrant children, making it possible that the
current interventions altered participant attitudes towards
children but not adults. Future research should pursue this by
focusing on outgroup attitudes and behavioral intentions
towards children and adults to elucidate whether extended
contact interventions affect participant attitudes across age.
With respect to Hypothesis 2, although previous research
with adults indicates that the relationship between indirect
group contact and attitudesis mediated by intergroup anxiety
(Pettigrew & Tropp, 2008), no clear support was found for
intergroup anxiety as a mediator within the current study.
This suggests the possibility that reduced anxiety may not be
the mechanism by which extended contact works among chil-
dren. As previous extended contact work with children indi-
cates, inclusion of other in the self or outgroup norms may
operateasmorelikelymechanismsandshouldbefurther
explored (Cameron et al., 2006; Cameron, Rutland, Hossain,
et al., 2011a; Cameron, Rutland, Turner, et al., 2011b).
With respect to Hypothesis 3, findings indicate no signifi-
cant differences between minority and majority participants.
This contradicts previous research indicating that direct inter-
ventions designed to affect minority attitudes towards the
majority are less effective (Beelman & Heinemann, 2014).
Rather, it provides some support for findings from extended
contact work with older children, which indicates no differen-
ces between minority and majority outcomes (Liebkind et al.,
2014). Also, in the U.S. sample, children expressed great joy
about seeing themselves depicted in the picture books. To our
knowledge, no published books in the United States depict
Somali and non-Somali children in fun, cross-group situa-
tions. So, for the U.S. students it was their first time seeing
themselves and their normal daily activities depicted in print
at school. This may have influenced the study outcomes by
countering perceived discrimination in the school setting. The
current sample of minority children was relatively small in
both studies, future work should continue to explore the
effects of extended contact among minority children to
deepen our understanding of its impact.
Finally, within the U.S. sample, children were evaluated at
two time points post intervention. As there were no significant
differences between Time 1 and Time 2 scores, findings some-
what support stability in outcomes over time, which is consist-
ent with previous research indicating the stability of positive
outcomes from direct contact interventions (Beelman & Heine-
mann, 2014). However, although significantly different behav-
ioral intentions emerged for younger children in the extended
contactconditionsatT1,theydidnotemergeatT2.Thissug-
gests that the positive effects of extended contact on behavioral
intentions may diminish over time. Conversely, T2 intergroup
anxiety scores were significantly lower than T1 among younger
children in extended contact conditions, suggesting gains in
anxiety reduction over time. Future research should continue
to incorporate longer-term outcome evaluations. Doing so has
the potential to deepen our understanding of the longer-term
impact of vicarious contact, its etiology and tangible interper-
sonal manifestations. Like most prejudice reduction interven-
tions, the current study relied on paper and pencil outcome
measures. The incorporation of informant, sociometric or
observational assessments would only benefit the field.
The current study brings extended contact interventions
one step closer to the classroom through the creation and
testing of purpose written stories that can be used to affect
children’s attitudes. Previous research has relied on researcher
led interventions in which existing messages children received
regarding difference were modified by the research in post-
reading discussions. Although this approach has been fruit-
ful, the lack of control over story content made it difficult to
predict outcomes in more broad based use. It is our hope
that continued work in this area will enable teachers to use
extended contact stories on their own, solidly translating this
important work from the academy to the classroom. Never-
theless, from the studies presented here, which also included
post-story discussion to be consistent with previous research,
it is not possible to be definite about whether it is the stories
themselves, the discussion, or their combination, which is
responsiblefortheobservedeffects.Futureresearchcould
usefully disentangle the two components in a design that pre-
sented stories with and without discussion.
Taken as a whole, the current findings suggest that schools
utilizing story based extended contact interventions should
do so with younger children, utilizing either common group
or dual identity stories. In doing so, teachers and administra-
tors can expect an increased willingness to interact across dif-
ference among participating children. Recognizing that these
Aronson et al. 239
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
results may not last without additional support, they should
consider following up extended contact work with additional
interventions. Doing so several weeks post extended contact
intervention may capitalize on child willingness and
decreased intergroup anxiety to affect even greater and more
long lasting change. Future research should investigate the
effects of extended contact when combined with other inter-
vention techniques.
References
Aboud, F. (1988). Children and prejudice.
Oxford: Basil Blackwell.
Aboud,F.E.,Tredoux,C.,Tropp,L.R.,
Brown,C.S.,Niens,U.,&Noor,N.M.
(2012). Interventions to reduce prejudice
and enhance inclusion and respect for eth-
nic differences in early childhood: A sys-
tematic review. Developmental Review,32,
307–336. doi: 10.1016/j.dr.2012.05.001
Abrams, D., Rutland, A., & Cameron, L.
(2003). The development of subjective
group dynamics: Children’s judgments of
normative and deviant in-group and
out-group individuals. Child Develop-
ment,74, 1840–1856. http://dx.doi.org/
10.1046/j. 1467-8624.2003.00641.x
Allport, G. W.(1954). Thenatureofpreju-
dice. Oxford, England: Addison-Wesley.
Aronson, K., & Brown, R. (2013). Accultur-
ation and social attitudes among major-
ity children. International Journal of
Intercultural Relations,37,313–322.doi:
10.1016/j.ijintrel.2013.02.004
Asher, S. R., & Allen, V. L. (1969). Racial
preference and social comparison proc-
esses. Journal of Social Issues,25, 157–
166.
Awad, G. H. (2010). The impact of accul-
turation and religious identification on
perceived discrimination for Arab/Mid-
dle Eastern Americans. Cultural Diversity
and Ethnic Minority Psychology,16,59–
67. doi: 10.1037/a0016675
Binder, J., Zagefka, H., Brown, R., Funke,
F., Kessler, T., Mummendey, A., ...
Leyens, J.-P. (2009). Does contact reduce
prejudice or does prejudice reduce con-
tact? A longitudinal test of the contact
hypothesis among majority and minority
groups in three european countries. Jour-
nal of Personality and Social Psychology,
96, 843–856. doi: 10.1037/a0013470
Beelmann, A., & Heinemann, K. (2014).
Preventing prejudice and improving
intergroup attitudes: A meta-analysis
of child and adolescent training pro-
grams. Journal of Applied Developmen-
tal Psycholog y,35, 10–24. doi:10.1016/
j.appdev.2013.11.002
Brown, R. (2010). Prejudice: Its social psy-
chology,2nded.Chichester:Wiley-
Blackwell.
Brown, R., & Hewstone, M. (2005). An
integrative theory of intergroup contact.
In M. P. Zanna (Ed.), Advances in experi-
mental social psychology (pp. 255–343).
San Diego, CA: Elsevier Academic Press.
Cameron, L., Rutland, A., Brown, R., &
Douch, R. (2006). Changing children’s
intergroup attitudes toward refugees:
Testing different models of extended
contact. Child Development,77, 1208–
1219. http://dx.doi.org/10.1111/j. 1467-
8624.2006.00929.x
Cameron, L., Rutland, A., Hossain, R., &
Petley, R. (2011a). When and why does
extended contact work? The role of high
quality direct contact and group norms in
the development of positive ethnic inter-
group attitudes amongst children. Group
Processes & Intergroup Relations,14(2),
193–206. doi:10.1177/1368430210390535
Cameron, L., Rutland, A., Turner, R.,
Holman-Nicolas, R., & Powell, C.
(2011b). ’Changing attitudes with a little
imagination’: Imagined contact effects
on young children’s intergroup bias.
Anales de Psicolog
ıa,27(3), 708–717.
Cascio, E., Gordon, N., Lewis, E., & Reber
S. (2008). From Brown to busing. Journal
of Urban Economics. 64, 296–325.
doi.org/10.1016/j.jue.2008.02.001.
Castelli,L.,Zogmaister,C.,&Tomelleri,
S. (2009). The transmission of racial
attitudes within the family. Develop-
mental Psychology,45, 586–591. doi:
10.1037/a0014619 10.1037/a0014619.
supp (Supplemental)
Christ,O.,Hewstone,M.,Tausch,N.,
Wagner, U., Voci, A., Hughes, J., & Cairns,
E. (2010). Direct contact as a moderator
of extended contact effects: Cross-
sectional and longitudinal impact on out-
group attitudes, behavioral intentions,
and attitude certainty. Personality And
Social Psychology Bulletin,36, 1662–1674.
Cramer, P., & Steinwert, T. (1998). Thin is
good, fat is bad: How early does it begin?
Journal of Applied Developmental Psychol-
ogy,19, 429–451. doi: 10.1016/S0193-
3973(99)80049-5
De Tezanos-Pinto, P., Bratt, C., & Brown, R.
(2010). What will the others think? In-
group norms as a mediator of the effects
of intergroup contact. British Journal of
Social Psychology,49, 507–523. doi:
10.1348/014466609X471020
Dixon, J., & Durrheim, K. (2003). Contact
and the ecology of racial division: Some
varieties of informal segregation. British
Journal of Social Psychology,42,1–23.
doi: 10.1348/014466603763276090
Dunham,Y.,Chen,E.E.,&Banaji,M.R.
(2013). Two signatures of implicit
intergroup attitudes: Developmental
invariance and early enculturation. Psy-
chological Science,24, 860–868. doi:
10.1177/0956797612463081
Doyle, A., Beaudet, J., & Aboud, F. (1988).
Developmental patterns in the flexibility
of children’s ethnic attitudes. Journal of
Cross-Cultural Psychology,19,3–18.doi:
10.1177/0022002188019001001
Feddes,A.R.,Noack,P.,&Rutland,A.
(2009). Direct and extended friendship
effects on minority and majority childrens?
Interethnic attitudes: A longitudinal study.
Child Development,80, 377–390. doi:
10.1111/j.1467-8624.2009.01266.x
Gaertner, S., & Dovidio, J. (2000). Reducing
Intergroup Bias: The common ingroup
identity model. Hove: Psychology Press.
Gaertner, S. L., & Dovidio, J. F. (2012). The
common ingroup identity model. In P. M.
VanLange,A.W.Kruglanski,E.T.Higgins,
P.M.VanLange,A.W.Kruglanski,E.T.
Higgins (Eds.), Handbook of theories of social
psychology (Vol. 2, pp. 439–457). Thousand
Oaks, CA: Sage Publications Ltd.
Gaertner, S. L., Rust, M. C., Dovidio, J. F.,
&Bachman,B.A.(1994).Thecontact
hypothesis: The role of a common
ingroup identity on reducing intergroup
bias.Small Group Research,25,224–249.
doi: 10.1177/1046496494252005
240 Telling tales in school
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
Gaertner, L., & Schopler, J. (1998). Per-
ceived ingroup entitativity and inter-
group bias: An interconnection of self
and others. European Journal Of Social
Psychology,28(6), 963–980. doi: 10.1002/
(SICI)1099-0992(1998110)28:6<963::
AID-EJSP905>3.0.CO;2-S.
Gonz
alez, R., & Brown, R. (2006). Dual
identities in intergroup contact: Group
status and size moderate the generaliza-
tion of positive attitude change. Journal
of Experimental Social Psychology,42,
753–767. doi: 10.1016/j.jesp.2005.11.008
Halperin,E.,Crisp,R.J.,Husnu,S.,
Trzesniewski,K.H.,Dweck,C.S.,&
Gross, J. J. (2012). Promoting intergroup
contact by changing beliefs: Group malle-
ability, intergroup anxiety, and contact
motivation. Emotion,12, 1192–1195. doi:
10.1037/a0028620
Hornsey, M. J., & Hogg, M. A. (2000). Sub-
group relations: A comparison of mutual
intergroup differentiation and common
ingroup identity models of prejudice
reduction. Personality and Social Psychol-
ogy Bulletin,26,242–256.doi:10.1177/
0146167200264010
Killen, M., & Rutland, A. (2011). Children
and social exclusion: Morality, prejudice
and group identity.NY:Wiley-Blackwell.
LaFreniere, P., Strayer, F. F., & Gauthier, R.
(1984). The emergence of same-sex affili-
ative preferences among preschool peers:
A developmental/ethological perspective.
Child Development,55, 1958–1965. doi:
10.2307/1129942
Levin, S., van Laar, C., & Sidanius, J.
(2003). The effects of ingroup and out-
group friendship on ethnic attitudes in
college: A longitudinal study. Group
Processes & Intergroup Relations,6,76–
92. doi: 10.1177/1368430203006001013
Liebkind, K., & McAlister, A. L. (1999).
Extended contact through peer modelling
to promote tolerance in Finland. European
JournalofSocialPsychology,29, 765–780.
doi: 10.1002/(SICI)1099-0992(199908/
09)29:5/6 <765::AID-EJSP958 >3.0.CO;
2-J
Liebkind, K., M€
ah€
onen, T. A., Solares, E.,
Solheim, E., & Jasinskaja Lahti, I. (2014).
Prejudice reduction in culturally mixed
classrooms: The development and assess-
ment of a theory driven intervention
among majority and minority youth in
Finland. Journal of Community &
Applied Social Psychology,24,325–339.
doi:10.1002/casp.2168
Maccoby, E. E., & Jacklin, C. N. (1987).
Gendersegregationinchildhood.InH.
W. Reese (Ed.), Advances in child devel-
opment and behaviour,(Vol.20, pp. 239–
287). San Diego, CA: Academic Press.
Nesdale, D., Maass, A., Durkin, K., &
Griffiths, J. (2005). Group norms, threat,
and children’s racial prejudice. Child
Development,76, pp. 652–663.
Paolini,S.,Hewstone,M.,Cairns,E.,&
Voci, A. (2004). Effects of direct and indi-
rect cross-group friendships on judg-
ments of catholics and protestants in
Northern Ireland: The mediating role of
an anxiety-reduction mechanism. Person -
ality and Social Psychology Bulletin,30,
770–786. doi: 10.1177/0146167203262848
Pettigrew, T. F. (2004). Justice Deferred A
Half Century After Brown v. Board of
Education. Amer ican Psychologist,59, 521–
529. doi: 10.1037/0003-066X.59.6.521
Pettigrew, T. F., Christ, O., Wagner, U., &
Stellmacher, J. (2007). Direct and indirect
intergroup contact effects on prejudice: A
normative interpretation.International
Journal of Intercultural Relations,31, 411–
425. doi: 10.1016/j.ijintrel.2006.11.003
Pettigrew, T. F., & Tropp, L. R. (2006). A
meta-analytic test of intergroup contact
theory. Journal of Personality and Social
Psychology,90, 751–783. doi: 10.1037/
0022-3514.90.5.751
Pettigrew, T. F., & Tropp, L. R. (2008). How
does intergroup contact reduce prejudice?
Meta-analytic tests of three mediators.
European Journal of Social Psychology,38,
922–934. doi: 10.1002/ejsp.504
Raabe, T., & Beelmann, A. (2011). Develop-
ment of ethnic, racial, and national prej-
udice in childhood and adolescence: A
multinational meta-analysis of age differ-
ences. Child Development,82,1715–
1737. doi:10.1111/j.1467-
8624.2011.01668.x
Reardon,S.F.,Brewal,E.T.,Kalogrides,D.,&
Greenberg, E. (2012). Brown fades: The
end of court-ordered school desegrega-
tion and the resegregation of American
public schools. Journal of Policy Analysts
and Management,31, 876–904. doi:
10.1002/pam.21649.
Stephan, W. G., & Stephan, C. W. (1985).
Intergroup anxiety. Journal Of Social
Issues,41(3), 157–175. doi: 10.1111/
j.1540-4560.1985.tb01134.x
Swart, H., Hewstone, M., Christ, O., &
Voci, A. (2011). Affective mediators of
intergroup contact: A three-wave longi-
tudinal study in South Africa. Journal Of
Personality And Social Psychology,101(6),
1221–1238. doi: 10.1037/a0024450
Tropp, L. R., & Pettigrew, T. F. (2005). Rela-
tionships between intergroup contact
and prejudice among minority and
majority status groups. Psychological Sci-
ence,16, 951–957. doi: 10.1111/j.1467-
9280.2005.01643.x
Turner,R.N.,Hewstone,M.,Voci,A.,
Paolini, S., & Christ, O. (2007). Reducing
prejudice via direct and extended cross-
group friendship. European Review of
Social Psychology,18,212–255.doi:
10.1080/10463280701680297
Turner, R . N . , Tam , T. , H e w s t o ne , M . ,
Kenworthy, J., & Cairns, E. (2013).
Contact between Catholic and Protes-
tant schoolchildren in Northern Ire-
land. Journal of Applied Social
Psychology,43(Suppl 2), E216–E228.
doi: 10.1111/jasp.12018
Williams, J. E., Best, D. L., Boswell, D. A.,
Mattson, L. A., & Graves, D. J. (1975).
Preschool racial attitude measure II. Edu-
cational and Psychological Measurement,
35, 3–18. http://dx.doi.org/10.1177/
001316447503500101
Wright,S.C.,Aron,A.,McLaughlin-
Volpe,T.,&Ropp,S.A.(1997).The
extended contact effect: Knowledge of
cross-group friendships and prejudice.
Journal of Personality and Social Psy-
chology,73, 73–90. doi:10.1037/0022-
3514.73.1.73
Yee, M., & Brown, R. (1994). The develop-
ment of gender differentiation in young
children. British Journal of Social Psychol-
ogy,33, 183–196. doi: 10.1111/j.2044-
8309.1994.tb01017.x
Aronson et al. 241
V
C2015 Wiley Periodicals, Inc Journal of Applied Social Psychology, 2016, 46, pp. 229–241
